Win wagering using an odds line...

[thaskalos]

I have drifted away from win wagering in recent years, and have been thinking as of late of getting back into it...seeing that today's short fields are not the ideal battleground for the super-exotic wagering I am naturally attracted to.

And I know that this will require creating an odds line...which is something I am not really looking forward to.

There are a few questions I have about win-betting using an odds line, and I would like to direct them at those who have more experience than me in wagering of this type:

1.) Do you confine your win wagers to those horses who are legitimate contenders in the race...or do you make serious bets on rank outsiders, just because they are offered at overlaid prices?

Would you make a serious wager on a 25-1 longshot...just because it is offered at 40-1?

2.) Do you bet less on those horses who don't really figure to win the race...or do you wager the same amount, since your odds line tells you that you still have your expected edge on the wager?

It occurs to me that the player should reduce the amount of his wagers when he bets on these overlaid outsiders...to account for the longer losing streaks he is sure to encounter when using wagering of this type.

3.) When there is more than one overlay in the race, is it better to dutch those horse...or should the wager be placed on the horse offering the biggest edge?

Thank you in advance for your replies; your help is greatly appreciated.

[Overlay]

I take a look at the whole field, since the highest-paying horses are likely to come from those that aren't regarded as traditional "contenders".

My wager sizes are governed by a percentage of the disparity between a horse's actual odds and my fair odds (what might be called "fractional hyper-Kelly"), which serves to place a limit on the amount that will be wagered on horses that I assign high fair odds to, while still assuring that I won't end up kicking myself if the horse wins and I had bet nothing on it.

Since I use handicapping categories that have "open-ended" bottom ranks, I stay away from horses (even if they are overlaid) that rank in the bottom category on all of my handicapping factors, no matter how large an overlay they might be.

If there are multiple overlays in a race, I use an "Optimal Group Dutch" strategy that can even allow for including horses with a high actual probability of winning that are slightly underlaid in the wager, while actually receiving a higher rate of return than if the underlaid horse is omitted. (If I were to bet only one horse in such a race, I would, as you say, select the one with the biggest edge.)

[Dave Schwartz]

1.) Do you confine your win wagers to those horses who are legitimate contenders in the race...or do you make serious bets on rank outsiders, just because they are offered at overlaid prices?

Would you make a serious wager on a 25-1 longshot...just because it is offered at 40-1?

the central focus of all my handicapping and wagering is my contenders. So the answer to this question is a resounding "yes."

Outsiders, as you call them, would never come up as overlays. They could, however, figure in exotics (but not in the win hole).


2.) Do you bet less on those horses who don't really figure to win the race...or do you wager the same amount, since your odds line tells you that you still have your expected edge on the wager?

I wager completely according to Kelly. Therefore, I am wagering money based upon a combination of profitability and hit rate.



3.) When there is more than one overlay in the race, is it better to dutch those horse...or should the wager be placed on the horse offering the biggest edge?

Let's say you have a race with three profitable contenders. If you apply the Kelly Criterion to each individual wager it might tell you to bet (say) 8% of bankroll. However, if you Dutch the three horses together, that is you treat the entire race as a single bet, the optimum that is liable to double or triple.

This is the one reason I am able to bet such a high percentage of bank in each race.


Regards,
Dave Schwartz

[pondman]

First and foremost a player must have confidence in a horse, somewhere in the neighborhood of 40%. I prefer to sit out until a horse comes along which meets my requirements, rather than force a bet.

This puts my line at 9-5, with a requirement of +4-1.

And at that point a player should worry about the enemy and where the enemy is placing their money. I don't think most players should try and scalp the odds by trying to have multiple win bets and accelerated betting schemes. Rather the player should condition himself/herself to take 60% of the bets as losses and keep the betting units consistent.

The industry doesn't like people to say it, but defending you money should be a priority. This means passing until you've got a live one, that the enemy doesn't see. If this requires changing your views and methods, then do it, or else you'll do a lot of bleeding.

[maddog42]

I have drifted away from win wagering in recent years, and have been thinking as of late of getting back into it...seeing that today's short fields are not the ideal battleground for the super-exotic wagering I am naturally attracted to.

And I know that this will require creating an odds line...which is something I am not really looking forward to.

There are a few questions I have about win-betting using an odds line, and I would like to direct them at those who have more experience than me in wagering of this type:

1.) Do you confine your win wagers to those horses who are legitimate contenders in the race...or do you make serious bets on rank outsiders, just because they are offered at overlaid prices?

Would you make a serious wager on a 25-1 longshot...just because it is offered at 40-1?

2.) Do you bet less on those horses who don't really figure to win the race...or do you wager the same amount, since your odds line tells you that you still have your expected edge on the wager?

It occurs to me that the player should reduce the amount of his wagers when he bets on these overlaid outsiders...to account for the longer losing streaks he is sure to encounter when using wagering of this type.

3.) When there is more than one overlay in the race, is it better to dutch those horse...or should the wager be placed on the horse offering the biggest edge?

Thank you in advance for your replies; your help is greatly appreciated.

I am a big believer in the 4 horse contender line. In 6 horse fields that is a bit dicey, but usually I still do it. My contenders usually consist of the top rated bsr horses and the pace setter or so called fulcrum horse. The exceptions I make to this are class drops. I give them the benefit of the doubt, and I will fudge a few points on pace and speed to get these horses in my top 4. Class drops win over 50% of races so you have to respect them. You are going to have to whittle this down to 4 or 5. If I have 5 horses then those last 2 will probably be saddled with 7-1 or 8-1 .

After class/speed/pace contenders are gotten I eliminate horses with very low trainer stats. 5% is my standard minimum, but the other day I bet a 3% trainer. I know this will turn a few heads but the book Speed to Spare got me to thinking of an exception guideline.I used to use 10% trainer minimum, but it was costing me too much money.
I am primarily a numbers guy, and the class thing is my weakest area. Still struggling here. If I am in doubt, about the A horse, and I often am, I will use the tote board to assign odds to it. If I hate the favorite so be it. I bet against him.If I like the favorite and he is 3-2 or lower and I think this is correct I pass the race. Yeah I know I was skeptical at first too, but Fierro
has shown me the light. Those other horses floating up to 7-1, because of that legitimate favorite deserve to be 7-1. If you have a legitimate favorite of 3-2 or lower you had better get some very good odds on a very legitimate horse that is at least 12-1. I am still experimenting with this last 12-1 part, but I know that my other contenders in the 4-1 5-1 7-1 range, with a legit favorite are money burners.False overlays. I don't bet any horse generally that is less than 5-1, and it has served me well. This enables me to pass races and save some money.

If a horse is not one of my top 4 i pass on it.

[Harvhorse]

After handicapping the race and narrowing it down to 3 or four contenders, I would bet my top 2 if they were 4/1 or higher. I would bet my 3rd and 4th choice if they were 8/1 or better. I will dutch if the odds are right because 10% roi in the long run is what I`m looking for.


"Life is 6/5 against"

[Inglewood Flamingo]

thaskalos, having personally been on the inside and privy to the wagering approach of one of the biggest successful whales in existence (talked about by name countless times on this forum) I can tell first hand that they do, and will continue to exploit and wager on any horse in a given race if it is overlaid according to their odds line. They have no limit as to how many they will invest in either. If it is an overlay, it will be wagered on accordingly. Hope this helps.

[MNslappy]

one of the biggest successful whales in existence (talked about by name countless times on this forum) I can tell first hand that they do, and will continue to exploit and wager on any horse in a given race if it is overlaid according to their odds line. They have no limit as to how many they will invest in either. If it is an overlay, it will be wagered on accordingly.


the key to this, because it's a whale, must be fat rebates, correct?

[Inglewood Flamingo]

Yes, but not necessarily. The whale I had access to certainly had the ability to beat the game straight up without rebates although they were getting generous ones.

[Dave Schwartz]

The whale business model is, basically, if you are flat-bet profitable, you aren't betting enough money.

[Golf and Horses]

Thask,


Firstly, I've really enjoyed reading your posts. They are informative and thought provoking.

As to the topic, IMHO, you must develop an accurate oddsline. This may take awhile. The idea is to have horses you assign 2-1 win more than horses you assign 5-2, 3-1...etc. Then you will feel confident in your linemaking.

I am mostly a win bet and exacta player, and as I've gotten older, I don't wager nearly as often as I used to...I mean a horse has to practically grab me by the collar and scream in my face..."BET ME!!!!", and he must be overlaid on my line. Except when I am at Saratoga...then I bet like my former self!!:D

I do make a line for horses I don't think are contenders, but even if they are overlaid on my line I do not wager on them

[Inglewood Flamingo]

The whale business model is, basically, if you are flat-bet profitable, you aren't betting enough money.

Absolutely correct Dave.

[kingfin66]

This is an interesting topic. One of the best and most eye opening pieces of information I have ever seen regarding the realities of betting an odds line comes from Ken Massa. Ken is very scientific in his approach to handicapping and betting and did a study regarding the inherent difficulties of winning by betting based on value alone. I would love to post a copy of the newsletter with the article here, but it is clearly copyrighted. Instead, I invite you or any other interested handicapper to visit the library at Ken's site and look in the Jan/Feb 2010 newsletter.

http://homebased2.com/km/library.htm#newsletters

[Dave Schwartz]

King,

which month newsletter would that be?

Dave

[iRNA]

Dave I think King said Jan/Feb 2010

[Robert Goren]

The main problem with value betting an odds line is that if you make a mistake in constructing the odds line, you will always bet that mistake. You have to be spot on all the time on all of the horses in a race.

Then there is problem of how sure you are of the odds lines. There are plenty of horses that could be anything. Horses with SRs earned on drying tracks are but one problem encounter because you aren't sure how reliable that SR is.

[Inglewood Flamingo]

HTR Newsletter January2010 pages 7-9. Funny Kingfin I almost referenced the identical article from HTR as I have been a subscriber for almost a decade and it really threw me for a loop when it was published. Ken and Dave and others on this board are among the sharpest minds the public has access to and I am certainly no stranger to the accuracy and validity of Ken's K-Line. I have been relying on since it's inception in 2003.

However, having said that while investing in single overlays may no longer be the path to riches as Barry Meadow and Mark Cramer advocated in the pre-whale days, nobody ever suggested the axiom of wagering on every overlay in every race and that is what I witnessed first hand via the whale.

[kingfin66]

Dave I think King said Jan/Feb 2010

Correct. It is a bi-monthly newsletter. It is restricted to HTR subscribers for the first year after release and is available to all thereafter.

[Inglewood Flamingo]

The main problem with value betting an odds line is that if you make a mistake in constructing the odds line, you will always bet that mistake. You have to be spot on all the time on all of the horses in a race.

Then there is problem of how sure you are of the odds lines. There are plenty of horses that could be anything. Horses with SRs earned on drying tracks are but one problem encounter because you aren't sure how reliable that SR is.

Robert, as I am sure that Dave and others on this board can attest to, the lines these whales create and rely on are that accurate and constantly being refined and tested. Just as quantitative and algorithmic traders are being wooed from the top 1% at MIT for successful hedge fund analysis, successful teams and whales employ the same strategies. Their lines are that good.

[Inglewood Flamingo]

Here is a link to a free eBook a friend of mine recently authored that actually talks quite a bit about the mathematical edge quant traders and fund managers have in the market.

It is a short read but I found it entertaining and informative as it often parallels our racing related thoughts. It is a tad technical but I feel most on this board can gain some insight from it and comprehend the material.

If you enjoy it please post your feedback here and I will let him know as he is not a horsey type of guy.

http://unbouncepages.com/free-book-give-away-1/

[Robert Goren]

I don't know that I would want to apply the hedge fund business model to betting horse racing considering how many of them end up broke. I am not sure what the success rate is for hedge funds, but I am sure it is a long way from 100%. It was the "quants" who sent the banks and AIG to the verge of bankruptcy in 2008. despite claims to the contrary, this is long way from an exact science. The best odds line in the world is still nothing more than an educated guess. When that guess is wrong, it will force the user to make a bad bet every time.
There are too many variables that the odds line makers don't have access to for their lines to be accurate all the time. How do you know whether a trainer is giving a horse an illegal drug or not? That has a huge effect on a horse's chances of winning. There are a lot of other unknowns as well. All the fancy number manipulation in the world has a hard time overcoming these unknowns.

[Bullet Plane]

Great question,


I will play a overlay on my odds line who is a contender and is 9-2 or less on my odds line. For a 9-2 shot, I would need 7-1 to play that horse.

If I see more than one overlay, I generally don't play the race. More than one overlay is a warning sign to me that I might not being seeing the race clearly.

I don't bet every overlay I come up with. I need to have a certain "feeling" that the horse will win. If I have gone a long time without getting that feeling, sometimes I will go ahead and play the overlay. Though, generally, I'm less successful that way.

The best race to put in an odds line spread sheet is the Kentucky Derby. You will see that many of your perceived "overlays" are not really overlays. You will also see how random the race is.... that is, with contenders 6-1 and above... the entire field being above random.

Best books out there on the subject:

Barry Meadow: Money Secrets at the Racetrack
Mark Cramer: The Odds on Your Side
Commonsense Betting: Dick Mitchell
The Four Quarters of Horse Investing: Steve Fierro

[davew]

Having your odds-line is good for 'relative ranking' as well as specific overlays

when your top 2 horses should be close in odds and they are not, the longer is an overlay

when your 2nd, 3rd, 4th horse are not being bet - they usually are overlay

you can use your relative ranking in exotics




disadvantage in parimutual, is after race started and all off-site locations get collected - overlays can become underlays

[Gapfire]

Robert, as I am sure that Dave and others on this board can attest to, the lines these whales create and rely on are that accurate and constantly being refined and tested. Just as quantitative and algorithmic traders are being wooed from the top 1% at MIT for successful hedge fund analysis, successful teams and whales employ the same strategies. Their lines are that good.

How does a whale overcome the problem of late odds changes?

I usually place my wagers within 30 seconds of post time, and normally see quite a drastic change when the race is over.

[Dave Schwartz]

by using months of tote board data to predict the final odds based on the recent movement in the odds.

[Gapfire]

by using months of tote board data to predict the final odds based on the recent movement in the odds.

The fluctuations in late tote action vary from day to day, and race to race for that matter. Getting the final tote right, is as big a part of the equation, as getting the probabilities of the horses.

[Dave Schwartz]

you must not underestimate the resources available to a group that has 50 or more employees to study things.

[TexasDolly]

Dave, the large Canadian whale stated on 60 minutes (300 million a year) only had 20 employees.
TD

[Dave Schwartz]

My sources say it is more like 45 employees.

He also said started with him and the laptop, but, again, my sources say he jumped in with a staff of over 20 to start with and invested over $400,000 in capital to get started.

[TexasDolly]

I suppose he could have more employees then what he stated but I can't see how he would gain by giving out a lower number then he actually had. Either way though he still makes wages it would appear. :)
TD

[maddog42]

The main problem with value betting an odds line is that if you make a mistake in constructing the odds line, you will always bet that mistake. You have to be spot on all the time on all of the horses in a race.

Then there is problem of how sure you are of the odds lines. There are plenty of horses that could be anything. Horses with SRs earned on drying tracks are but one problem encounter because you aren't sure how reliable that SR is.

"The main problem with( fill in the blank) is that if you make a mistake in (fill in the blank), you will always bet that mistake." Robert this pretty much applies to any endeavour of gambling. I am not being disrespectful here. I find I am making fewer mistakes and finding more value in most races. I am passing more races.
I really suspect that most line makers, even those whales that are being mentioned probably make mistakes more than half the time. Thats ok. You can make a line that is dead solid perfect and still get beat. You can be right less than half the time and still make money if your mutuel is high enough.The main purpose of an odds line is to make you bet your opinion more without getting sidetracked.

[Robert Goren]

It is not size of the mutual covers your mistakes but the profit margin you have at those odds. If you are operating with a 2% profit margin at 20/1 odds, you can't afford any more mistakes than if you operating at profit margin of 2 % at even money. Some people can and do make money value betting their lines. It is just that sounds so easy to do when in practice it is very complicated and very hard to do.

It seems to the method du jour on this board right now and the people who advocate it get a pass. If a system player started advocating that style he is meant with a barrage. And don't even dare to bring up tote board handicapping. Value line handicapping has a lot of inherit problems to overcome such as odds drops after the start. The whales may or may not do it, I have no idea. But most of the posters here are not whales and have neither the bankroll or the resources to try copy their methods. A 2% ROI and 10k bankroll is a tough way to pay the bills every month. And that doesn't include the costs to develop a good value line. The information for large data bases is not cheap. Luckily the programs to study the data are pretty reasonable. Of course you still have learn how to use them. An online course in basic statistics is not going cut to cut it. It doesn't take too long to get the point in your quest for knowledge that everything starts to read like Greek to a Roman. If you are asking about where to find papers by Bolton, etc you are light-years away from developing something profitable by going this route.

[Dave Schwartz]

Goren,

What you are saying sounds a lot like "It's too hard," and you are right: it is very difficult. In fact, it is TOO difficult for a "normal," human being with less than a PhD in mathematics to accomplish.

But what if there was a way that normal people could accomplish this?

What if there was a way that was just a little difficult but still do-able without any advanced math skills? (Nothing beyond a little multiplication and division) What if you could do it in a spreadsheet?

After all, anything worth doing is worth doing right? Yes?

[Overlay]

Goren,

What you are saying sounds a lot like "It's too hard," and you are right: it is very difficult. In fact, it is TOO difficult for a "normal," human being with less than a PhD in mathematics to accomplish.

But what if there was a way that normal people could accomplish this?

What if there was a way that was just a little difficult but still do-able without any advanced math skills? (Nothing beyond a little multiplication and division.)
Dave,

Overlay Publications (http://www.overlaypublications.com) thanks you for the plug! ;)

[Dave Schwartz]

Overlay,

I am not sure I understand.

Please elaborate.


Dave

[Dave Schwartz]

Oh, I get it now.

You think I am referring to using Impact Value Tables to make a line.

Sorry, but, IMHO, that simply does not work.

The idea that we can just multiply a few factors together and come up with probabilities accurate enough to beat the game on a regular basis is not something I can agree with. In fact, I would say that the primary lesson from this thread is that the make-a-line-from-factor-values only works at the highest level of mathematics.

This does not mean that IV Tables do not have, uh... value (no pun intended). They do. They offer a player a method of determining who the best horses in a race are.

But, IMHO, producing profit directly from the normalized output of an IV system is highly unlikely. It is going to take another step to make it work.

It just so happens I know of someone who has created that step. <G>


Regards,
Dave Schwartz

[Robert Goren]

Goren,

What you are saying sounds a lot like "It's too hard," and you are right: it is very difficult. In fact, it is TOO difficult for a "normal," human being with less than a PhD in mathematics to accomplish.

But what if there was a way that normal people could accomplish this?

What if there was a way that was just a little difficult but still do-able without any advanced math skills? (Nothing beyond a little multiplication and division) What if you could do it in a spreadsheet?

After all, anything worth doing is worth doing right? Yes?If you can develop an easy way to do it, more power to you. If you have formula to develop a good odds line, things get pretty easy. The work is in developing the formula. It involves more running a multi correlation program with 4 or 5 variables. Just take a look at days since the last race. Is being off for 90 days mean more if the horse won a cheap claimer than a horse that ran third in stakes. What it was in the week of the season for a circuit? What if the trainer wintered at Tampa but left the horse up north? All of these things could effect a horse's chances of winning in a huge way and therefore effect the other horses in the races chances of winning. When you operating on a razor thin ROI, you need to know the answers to those questions. If a person has the right software and a large enough data base, I suppose you run the those parameters on horse and get a reason good answer to those question. If I was doing it, that what I would do. I suppose you could set it up so it would automatically run all but the most extreme parameters. There are still some issues to deal with but it is probably doable. I am not exactly sure what your software does or doesn't do, but I suspect with a large enough database, a person could probably get a pretty good line on most races if they went about it right. The trick would be knowing what questions to ask.

[maddog42]

Robert, although I have used databases in the past(long time ago), I don't use them now. I am not trying to sell anything. I am not trying to help anybody sell anything. I noticed that you do advocate value in the handicapping process.
At least we agree on that. I don't know about the 2% edge. I am a notebook (paper)handicapper. I get the winner in the top 4 75%-80% of the time, depending on the category. I don't seem to be very good on turf sprints. I am certainly weak in some categories so I try to stay away from those.
One of the things I do with my betting line is try to find value in the place hole. I can't seem to do anything in the conventional way. This is a blessing and a curse. And while I can't claim as good a percentage as the database guys, they don't have my esoteric skill set either.

PS I am not one of the high rollers on this board, and I have only been using my new oddsline approach for a couple of months. That is not very long. It could unravel anytime. But I have noticed a marked improvement, in my bottom line. I will continue to tweak the damn thing.

[Dave Schwartz]

If you can develop an easy way to do it, more power to you. If you have formula to develop a good odds line, things get pretty easy. The work is in developing the formula. It involves more running a multi correlation program with 4 or 5 variables. Just take a look at days since the last race.

Goren,

50-60 variables would be more like it. (Or more.)

No, that would be more of the same: beyond most mortal men.

I said:
What if there was a way that was just a little difficult but still do-able without any advanced math skills? (Nothing beyond a little multiplication and division) What if you could do it in a spreadsheet?

What I have in mind is relatively simple, but amazingly powerful. It is so simple that even a paper-and-pencil guy like Maddog could do it in about 20 seconds with a spreadsheet. And it has the power to make a "good handicapper" into a profitable one with just a simple tweak.

In fact, the method would probably be perfect for him.


Regards,
Dave Schwartz

[maddog42]

Goren,

50-60 variables would be more like it. (Or more.)

No, that would be more of the same: beyond most mortal men.

I said:


What I have in mind is relatively simple, but amazingly powerful. It is so simple that even a paper-and-pencil guy like Maddog could do it in about 20 seconds with a spreadsheet. And it has the power to make a "good handicapper" into a profitable one with just a simple tweak.

In fact, the method would probably be perfect for him.


Regards,
Dave Schwartz

Don't follow leaders, watch the Parking meters - Bob Dylan

[Robert Goren]

Robert, although I have used databases in the past(long time ago), I don't use them now. I am not trying to sell anything. I am not trying to help anybody sell anything. I noticed that you do advocate value in the handicapping process.
At least we agree on that. I don't know about the 2% edge. I am a notebook (paper)handicapper. I get the winner in the top 4 75%-80% of the time, depending on the category. I don't seem to be very good on turf sprints. I am certainly weak in some categories so I try to stay away from those.
One of the things I do with my betting line is try to find value in the place hole. I can't seem to do anything in the conventional way. This is a blessing and a curse. And while I can't claim as good a percentage as the database guys, they don't have my esoteric skill set either.

PS I am not one of the high rollers on this board, and I have only been using my new oddsline approach for a couple of months. That is not very long. It could unravel anytime. But I have noticed a marked improvement, in my bottom line. I will continue to tweak the damn thing. I am not a value line bettor. I am basically a handicapper trying to pick to pick the most likely winner. I have also used systems. I am using one right now that is kicking a pretty good ROI. The problem for me is that once I develop a system, I get bored using it. Horse racing to me is a hobby. I find the idea of trying make a living gambing unappealing. I have known people who do it. It seems like a horible way to live. But that is just me.

[Dave Schwartz]

What I have in mind is absolutely independent of your handicapping approach. It almost does not matter how you handicap. (Obviously, the better the handicapping approach, the better the results.)

It doesn't matter whether you are a recreational player or a professional. Either way, it is more fun to win than to lose.

At least that's how I see it.


Regards,
Dave Schwartz

[therussmeister]

How does a whale overcome the problem of late odds changes?

I usually place my wagers within 30 seconds of post time, and normally see quite a drastic change when the race is over.
The whale is the one making the late odds change. The real problem is how does a whale overcome the possibility of another whale making the same bets?

[Overlay]

The whale is the one making the late odds change. The real problem is how does a whale overcome the possibility of another whale making the same bets?
And another problem would be how the whales guarantee that the horse that they are betting will win. That's why it's important to look at value over the entire field, since, as the odds on the whale's horse drop, the odds on other horses in the race (one or more of which may have been offering greater value than the whale's horse to begin with) will rise even higher.

[Dave Schwartz]

Guaranteed to win?

How is that done?

[thaskalos]

The whale is the one making the late odds change. The real problem is how does a whale overcome the possibility of another whale making the same bets?

This is the question I have been asking myself ever since the "whale-effect" started presenting itself.

We have these huge bettors placing the bulk of their wagers as late as possible...using sophisticated, hi-tech handicapping methods which we must assume are similar in nature.

How do they avoid betting on the same horses, and driving their prices far below their acceptable levels?

[Robert Goren]

Guaranteed to win?

How is that done?Frank Costello took that secret to his grave. ;)

[Overlay]

Guaranteed to win?

How is that done?
My point exactly. I was not stating "guaranteed to win" as a fact, but indicating that, even though a whale is loading up on a particular horse, it does not give that horse a 100% probability of winning. And the whale's activity potentially creates value on other horses in the race (or even more value, in the case of horses that may have already been overlays before the whale made his wager).

[thaskalos]

My point exactly. I was not stating "guaranteed to win" as a fact, but indicating that, even though a whale is loading up on a particular horse, it does not give that horse a 100% probability of winning. And the whale's activity potentially creates value on other horses in the race that may have even already been overlays to begin with.

How does the whale's activity create value on other horses in the race...when we don't see the effects of this activity until the race is already under way?

[Overlay]

How does the whale's activity create value on other horses in the race...when we don't see the effects of this activity until the race is already under way?
Because if you make a full-field odds line, you can see which horses are offering value over the full range of competitors in the race before the whale makes his wager. The odds on all of those horses (except for the one that the whale is betting) will rise even further as a result of the whale's activity.

[Dave Schwartz]

Overlay,

IMHO, your logic is faulty.



Regards,
Dave Schwartz

PS: Thaskalos, your mailbox is full and you are apparently not getting my emails. (Or you are avoiding me. <G>)

[thaskalos]

Because if you make a full-field odds line, you can see which horses are offering value over the full range of competitors in the race before the whale makes his wager. The odds on all of those horses (except for the one that the whale is betting) will rise even further as a result of the whale's activity.

Yes, Overlay...but this does not address the player's main concern when using an odds line.

We must assume that the whale is operating with an odds line much more soplisticated than own own...otherwise, he would last long in this game.

Now...if the whale's odds line is better than ours, and the legitimate overlays get pounded late as a result of the whale's betting activities...where does that place us "normal" bettors -- who have placed are wagers BEFORE the whales, taking the toteboard odds at face value?

[thaskalos]

Overlay,

IMHO, your logic is faulty.



Regards,
Dave Schwartz

PS: Thaskalos, your mailbox is full and you are apparently not getting my emails. (Or you are avoiding me. <G>)

Sorry, Dave...

I failed to notice that my mailbox was full.

I seem to be more popular than I thought I was...:)

[Dave Schwartz]

Check your messages now, please.

[Overlay]

Yes, Overlay...but this does not address the player's main concern when using an odds line.

We must assume that the whale is operating with an odds line much more soplisticated than own own...otherwise, he would last long in this game.

Now...if the whale's odds line is better than ours, and the legitimate overlays get pounded late as a result of the whale's betting activities...where does that place us "normal" bettors -- who have placed are wagers BEFORE the whales, taking the toteboard odds at face value?
Then why do we even continue to play this game? And why do I (and, I am sure, others such as Dave Schwartz) continue to receive positive comments about the performance of our approaches? And what alternative is there, when playing "pick-the-winner", without regard to odds or value, inevitably results in obsolescence through mutuels that are below break-even level?

[thaskalos]

Then why do we even continue to play this game? And why do I (and, I am sure, others such as Dave Schwartz) continue to receive positive comments about the performance of our approaches? And what alternative is there, when playing "pick-the-winner", without regard to odds or value, inevitably results in obsolescence through mutuels that are below break-even level?.

I am not doubting the effectiveness of your (or Dave's) handicapping approach. I am just stating a problem as I see it...which is the main reason why I have stayed away from odds lines in recent years.

I am just not seeing the logic in your comment that the whale's activity creates value on the board, which the "normal" player can exploit.

[TrifectaMike]

A little gift for all of you from me....

P(N,R) = 1/N (1/R + ... + 1/N) ... A probability generating function... Dave is correct you probability need a Ph.D in math to figure this out (Don'r even try...just accept the equation)

N -> Number of horses in the race
R -> Rank of horse's odds on the board

Example:

N -> 10 horse field
R -> Horse A ranked (by odds) 7th on the board

P(10,7) = 1/10 (1/7 + 1/8 + 1/9 + 1/10) ... Probability of Horse A winning as ranked by odds by the public.

You need to think carefully about what the equation is modeling. You might say, but the probability is defined by the odds...What is this probability, P(N,R) saying to me.

Mike (Dr Beav)

[Overlay]

I am not doubting the effectiveness of your (or Dave's) handicapping approach. I am just stating a problem as I see it...which is the main reason why I have stayed away from odds lines in recent years.

I am just not seeing the logic in your comment that the whale's activity creates value on the board, which the "normal" player can exploit.
The normal player exploits it (as I was trying to say earlier) by creating a full-field fair-odds line without regard to what the whale might or might not do, and thus benefiting when horses that are already overlays according to that line to begin with (i.e., before the whale bets) pay off at even higher prices on those occasions when the whale's horse does not perform as expected.

[AITrader]

How do they avoid betting on the same horses, and driving their prices far below their acceptable levels?

Supposedly the famous Hong Kong whales were down to a 2% edge at the end of the 1990's. Ranogajec is rumored to bet whenever he can gain a 2% edge. This is about the same edge an expert blackjack counter gets in a casino with highly favorable rules.

[Dave Schwartz]

The normal player exploits it (as I was trying to say earlier) by creating a full-field fair-odds line without regard to what the whale might or might not do, and thus benefiting when horses that are already overlays according to that line to begin with (i.e., before the whale bets) pay off at even higher prices on those occasions when the whale's horse does not perform as expected.

In my opinion this will not work.

...You need to think carefully about what the equation is modeling. You might say, but the probability is defined by the odds...What is this probability, P(N,R) saying to me.

In my opinion, for most players (without the required math degree), this won't work either.


There is a better way. When I return from our last graduation I will begin the process of building a seminar to explain it.

For those with an open mind, it will make life much simpler, easier, and possibly profitable. At the very least, you will always be able to tell a good race to bet from a race to pass.

[TexasDolly]

Overlay- I believe the "whales ' bet more than one in most situations ,leaving you restricted choices to find the overlays.
TD

[Overlay]

Overlay- I believe the "whales ' bet more than one in most situations ,leaving you restricted choices to find the overlays.
TD
So you're saying that the whales dominate the game today to such an extent that it is now an impossibility, in any race at any track, to systematically find and bet any horses that have a positive wagering expectation in light of their odds? Once again, why continue to play the game under those conditions?

[TexasDolly]

No, I didn't say it quite that way. Because of the large rebates their bet downs involve more than one horse most of the time and those horses are bet below a return of $ 2.00/horse. Of course, there are overlays but there are fewer and have smaller overlays than pre - whale.
TD

[TexasDolly]

TM, Does this mean that the actual value of the odds is of no consequence in calculating the probability ? I'm having some trouble with that thought.
Thank you,
TD

[thaskalos]

So you're saying that the whales dominate the game today to such an extent that it is now an impossibility, in any race at any track, to systematically find and bet any horses that have a positive wagering expectation in light of their odds? Once again, why continue to play the game under those conditions?
It is still possible to find overlays...but the whales are a big obstacle for us, for several reasons.

These whales are using much more sophisticated handicapping techniques than we are...and they spot these overlays more efficiently than we do. And then, they employ a wagering method which pounces on ALL these overlays, wherever they can be found...with the bulk of the wagering taking place at the last few seconds...after the average player has placed his wager.

So, we have a powerful opponent here...who is doing the exact same thing that we are trying to do (spot overlays), but his handicapping techniques allow him to be MUCH more proficient at it than we are...

AND...he gets to play his hand after we have played ours.

How can we effectively compete against him; because, the way he places his bets, there is no way for us to play "around" him.

We mere mortals are reduced to a guerrilla-type attack...striking the occasional overlay here and there...

No?

[Dave Schwartz]

I promise that there is a way. In fact, you can actually profit from it.

I have decided to explain it to Thaskalos. He has agreed to share his opinion of the idea (good or bad; agree or disagree) without divulging how it works.

In a couple of weeks I will begin designing a workshop.


Regards,
Dave Schwartz

[Overlay]

We mere mortals are reduced to a guerrilla-type attack...striking the occasional overlay here and there...

No?
Agreed, which all the more places a premium on being able to spot value wherever and whenever it occurs.

[Overlay]

TM, Does this mean that the actual value of the odds is of no consequence in calculating the probability ? I'm having some trouble with that thought.
Thank you,
TD
Are you referring to the concept that, if you're getting a big overlay on a horse according to your line, then the horse probably does not have as good a chance of winning as you believe that it does?

[TexasDolly]

TM had posted a formula for determining the probability for a horse to win and the formula was based on the rank of the horse's odds not the actual value of the odds. I find that
notion pretty difficult to grasp. I don't think all no.1 ranked
horses(same size fields) have the same probability of winning so I had written that note to him in hopes he will explain his thinking .
TD
TD

[TrifectaMike]

TM had posted a formula for determining the probability for a horse to win and the formula was based on the rank of the horse's odds not the actual value of the odds. I find that
notion pretty difficult to grasp. I don't think all no.1 ranked
horses(same size fields) have the same probability of winning so I had written that note to him in hopes he will explain his thinking .
TD
TD
Hi TD,

I'll do my best to explain.

Firstly, the probabilities associated with horse racing are not properties of the event itself, but rather a function of the information which we possess about the event.

There are NO objective probabilities. There is NO "true" probability for a horse to win.

To speak of the probability of a horse winning is fallacious. What we REALLY need is the "probability of ME being correct, if I forecast the outcome of a race to be horse 1 or horse 2, etc.

Secondly, the probabilities generated by the formula,

P(N,R) = 1/N (1/R + ... + 1/N)

(for those who need to know... the formula is derived strictly mathematically by the use of distributions of distributions, bijection and reduction).

are rank based and not derived from the actual values on the tote.

Assume, for the moment, you didn't know the actual odds offered. Does there exist in nature a way of assigning probabilities based soley on rank. The short answer, yes. Recall that probabilities for a horse race; those derived from the tote, regression, curvefitting, etc, etc are REALLY probabilities of the information which we possess about the event. The probabilities reflected by the tote are soley information driven and not implicit to the event. Remove actual values and let nature direct you to a subjective probability sans odds.

So, the differences between the probabilities generated by the formula (given above) and the probabilities derived from the actual tote values are due to information. Once again, the deltas are movements, based, on information... either positive or negative based on your position.

Another way to view this is that you are handicapping the total event, not a particular horse.

I hope this is clear. If not I can try again. Or someone else who does understand what I'm attempting to say can pitch in...all are welcome.

Mike (Dr Beav)

[Jeff P]

A little gift for all of you from me....

P(N,R) = 1/N (1/R + ... + 1/N) ... A probability generating function... Dave is correct you probability need a Ph.D in math to figure this out (Don'r even try...just accept the equation)

N -> Number of horses in the race
R -> Rank of horse's odds on the board

Example:

N -> 10 horse field
R -> Horse A ranked (by odds) 7th on the board

P(10,7) = 1/10 (1/7 + 1/8 + 1/9 + 1/10) ... Probability of Horse A winning as ranked by odds by the public.

You need to think carefully about what the equation is modeling. You might say, but the probability is defined by the odds...What is this probability, P(N,R) saying to me.

Mike (Dr Beav)
Testing the formula's resulting prob number against actual results pulled from a db, I arbitrarily picked: 3rd public choice in the odds, field size = 8.

If I understand the formula correctly, here's the calculation:

p(8,3) = (1/8) * (1/3 + 1/4 + 1/5 + 1/6 + 1/7 + 1/8)

p(8,3) = (.125) * (.3333 + .2500 + .2000 + .1667 + .1429 + .1250)

p(8,3) = 0.1522


Next, here's what I have in the db for all thoroughbred starters in North America 1/1/2011 through yesterday 5/11/2012 - 3rd choice in the odds, 8 horse fields:

query start: 5/12/2012 11:20:37 AM
query end: 5/12/2012 11:21:39 AM
elapsed time: 62 seconds

Data Window Settings:
Connected to: C:\JCapper\exe\JCapper2.mdb
999 Divisor Odds Cap: None

SQL: SELECT * FROM STARTERHISTORY
WHERE RANKODDS = 3
AND FIELDSIZE = 8


Data Summary Win Place Show
Mutuel Totals 19445.80 19034.40 20030.90
Bet -23786.00 -23786.00 -23786.00
Gain -4340.20 -4751.60 -3755.10

Wins 1789 3842 5880
Plays 11893 11893 11893
PCT .1504 .3230 .4944

ROI 0.8175 0.8002 0.8421
Avg Mut 10.87 4.95 3.41



Neat little formula. Thx Mike!


"Does this mean that the actual value of the odds is of no consequence in calculating the probability ?"


I say no. If I take the data from the above sample and break it out by odds range it looks like this:

By: Odds Range

>=Min < Max Gain Bet Roi Wins Plays Pct Impact
-999.00 0.00 0.00 0.00 0.0000 0 0 .0000 0.0000
0.00 0.50 0.00 0.00 0.0000 0 0 .0000 0.0000
0.50 1.00 0.00 0.00 0.0000 0 0 .0000 0.0000
1.00 1.50 6.50 8.00 1.8125 3 4 .7500 4.9859
1.50 2.00 -2.20 68.00 0.9676 12 34 .3529 2.3463
2.00 2.50 -24.00 240.00 0.9000 33 120 .2750 1.8282
2.50 3.00 -186.60 1050.00 0.8223 115 525 .2190 1.4562
3.00 3.50 -449.70 2680.00 0.8322 265 1340 .1978 1.3147
3.50 4.00 -751.80 3862.00 0.8053 331 1931 .1714 1.1395
4.00 4.50 -928.50 3800.00 0.7557 278 1900 .1463 0.9727
4.50 5.00 -371.80 3230.00 0.8849 251 1615 .1554 1.0332
5.00 5.50 -485.00 2452.00 0.8022 159 1226 .1297 0.8622
5.50 6.00 -495.50 1848.00 0.7319 101 924 .1093 0.7267
6.00 6.50 -109.30 1318.00 0.9171 84 659 .1275 0.8474
6.50 7.00 -112.80 1022.00 0.8896 59 511 .1155 0.7676
7.00 7.50 -23.20 614.00 0.9622 36 307 .1173 0.7796
7.50 8.00 -119.60 448.00 0.7330 19 224 .0848 0.5639
8.00 8.50 -17.90 312.00 0.9426 16 156 .1026 0.6818
8.50 9.00 -67.80 204.00 0.6676 7 102 .0686 0.4562
9.00 999999.00 -201.00 630.00 0.6810 20 315 .0635 0.4221
Win rates for each xth ranked public choice in the betting does vary based on the odds. However, that doesn't mean the formula Mike posted isn't useful.


-jp

.

[Light]

If I understand the formula correctly, here's the calculation:

p(8,3) = (1/8) * (1/3 + 1/4 + 1/5 + 1/6 + 1/7 + 1/8)

p(8,3) = (.125) * (.3333 + .2500 + .2000 + .1667 + .1429 + .1250)

p(8,3) = 0.1522

If it is an 8 horse field,why are only 6 horses being compared to the horse in question. Shouldn't there be 7?

[maddog42]

I worked out the oddsline/probabilities from the above formula.

8 horse field:

1 .339
2 .214
3 .1522
4 .1105
5 .079
6 .054
7 .033
8 .015

9 horse field

1 .314
2 .203
3 .147
4 .1106
5 .082
6 .060
7 .033
8 .026
9 .012

Notice that the top 3 choices have less chance of winning in the 9 horse field. No problem with me, but the 4 thru 8 choices have a greater chance of winning in the 9 horses field. HUH??
It is late and I am tired so maybe I punched the wrong numbers in, but I don't think so.

[Robert Fischer]

i do more of an intuitive thing.

If I love a horse to win and he is 2nd choice (or higher) he's a bet. However, if he is a ml favorite and "ice cold" on the board, then he could scare me away at higher odds.

If a horse is a lukewarm favorite and I feel he crushes the field, (or maybe I am strongly against the other favorite), he's a win play.

It's more a matter of seeing a mismatch and feeling the price is acceptable for me, than making an actual odds line.

[TrifectaMike]

I worked out the oddsline/probabilities from the above formula.

8 horse field:

1 .339
2 .214
3 .1522
4 .1105
5 .079
6 .054
7 .033
8 .015

9 horse field

1 .314
2 .203
3 .147
4 .1106
5 .082
6 .060
7 .033
8 .026
9 .012

Notice that the top 3 choices have less chance of winning in the 9 horse field. No problem with me, but the 4 thru 8 choices have a greater chance of winning in the 9 horses field. HUH??
It is late and I am tired so maybe I punched the wrong numbers in, but I don't think so.

You punched in the correct numbers. Odd isn't it, but the empirical data should valid that anomaly... nature isn't linear.

Mike (Dr Beav)

[johnhannibalsmith]

If it is an 8 horse field,why are only 6 horses being compared to the horse in question. Shouldn't there be 7?

If I understood it (which is doubtful) the formula begins with the third ranked through the eighth (ie 1/3 -> 1/8 in Jeff's example) derived from the (3,8)... aka 3,4,5,6,7,8.

[TrifectaMike]

Suppose one can take these probabilities based on soley on the rankings and map them into the actual odds (Tote odds):

P(Tote) = [Map]P(Rank)

[Map] would provide vital information as to what, which is driving the odds from their natural state to the P(Tote) state. In fact, [Map] could be computed at specific time intervals.

Mike (Dr Beav)

[TexasDolly]

Thanks for the very helpful replies . I think I am a little closer to understanding.
TD

[HUSKER55]

I am still on my first cup of coffee. Suppose your pick is third and the odds are 7/2 and the formula result is .1522 like jeff did.

what does that mean to the player?

thanks

h55

[Overlay]

I am still on my first cup of coffee. Suppose your pick is third and the odds are 7/2 and the formula result is .1522 like jeff did.

what does that mean to the player?

thanks

h55
I would say that, since the probability that the horse will win based on odds ranking and field size (.1522), corresponds to odds of 6-1 (or, more precisely, 5.57-1, which would be rounded up to 6-1, rather than down to 5-1, for purposes of determining value), and the horse is 7-2 on the board, then it is an underlay by that criterion, and should not be bet.

[teddy]

My sources say it is more like 45 employees.

He also said started with him and the laptop, but, again, my sources say he jumped in with a staff of over 20 to start with and invested over $400,000 in capital to get started.


My sources say there was a program built by such nerd that cost over 500k to look at all the variables, check the tote. ect.

[HUSKER55]

THANK YOU!

[teddy]

My exercise on picking first call leaders holds the one key to this type of wagering. Whether or not a whale can come up with a program to find the first call leaders is beyond me, but I think they probably could do a good job of it based on post position and ability of each horse, along with early lick. One thing that never changes is that the first call leader is golden.

On the thoughts of Dave and dealing with whales late bets... well Ken Massa already covered this with his projected slash favs. Horses that were projected to get huge late bets. At Mountaineer park the third fav is the safest bet. If a huge wager goes down early, then the retaliatory wagers in the last min will drive the 1-9 to even money, always on the next most likely winner. Driving odds up on the other horses, usually. With so many groups making retaliatory late bets, there is no way to know what your up against..

At mtn you would need a program that could factor in all the trainer moves, ones that you have to play there nightly to pick up on. The huge track bias, know what agent has what jockeys, know when assistant trainers are listed as trainers haha.. good luck whales

[Dave Schwartz]

At mtn you would need a program that could factor in all the trainer moves, ones that you have to play there nightly to pick up on. The huge track bias, know what agent has what jockeys, know when assistant trainers are listed as trainers haha.. good luck whales

I always have a chuckle over stuff like this:

The presumption that people using software cannot figure something out with an algorithm therefore it is somehow easier without one.

[Jeff P]

I am still on my first cup of coffee. Suppose your pick is third and the odds are 7/2 and the formula result is .1522 like jeff did.

what does that mean to the player?

thanks

h55

FWIW, I have a different take on this...

Situation #1. - If the only thing you know is that the horse ranks 3rd in a field of 8, then using the formula gives you estimated win prob for the horse at or near 15%.

Situation #2. - If you know that the horse ranks 3rd in a field of 8 - and you also know what the odds are (in this case 7/2) - Now you have more information than you did in the first situation.

Using all of the information at your disposal enables you to estimate win prob for the horse at somewhere close to 17% (the historical win prob in the odds chart that I posted for horses ranked 3rd in the odds in fields of 8 where the odds were at or near 7/2.)

17% is a more accurate prob estimate than 15% because you had better/more complete information to work with in situation #2 than you did in situation #1.

What's really happening here is that by setting the odds at 7/2, the public is giving you (additional) information about the horse - info you did not have before you knew what the odds were.

The public is telling you that the horse should win 17% of the time.

IMHO, the crux of the matter becomes:

How much movement in either direction - up or down - away from that 17% public probability can what YOU know about the horse reshape the outcome?

IMHO, odds line building done well means that you as a player know the performance of your "model" in terms of how much movement it produces away from the public probability inherent in the odds (given the situation.)

• If your "model" enables you to recognize when that 17% public horse is actually closer an 11% horse - It's likely that an edge can be had by betting against that 17% horse.

• If your "model" enables you to recognize when that 17% public choice is actually a 17% horse - Your model is telling you to look elsewhere.

• If your "model" enables you to recognize when that 17% public horse is actually closer to a 23% horse - It's likely that an edge can be had by betting on that 17% horse.

If I may make an educated guess... The "models" used by some of the whale teams are reasonably good at doing the above for every horse in each race.


-jp

.

[maddog42]

7 horse field:

1 .370 or 9/5
2 .227 or 7/2
3 .156 or 6-1
4 .108 or 9-1
5 .072 or 15-1
6 .044 or 25-1
7 .02 or 50-1

6 horse field

1 .408 or 3-2
2 .241 or 7-2
3 .158 or 6-1
4 .102 or 9-1
5 .06 or 16-1
6 .027 or 50-1

5 horse field

1 .456 or 6-5
2 .256 or 3-1
3 .156 or 6-1
4 .090 or 10-1
5 .040 or 25-1

4 horse field

1 .520 or 1-1
2 .279 or 3-1
3 .145 or 6-1
4 .062 or 15-1

First off, I apologize to Thask, for hijacking the thread. This "pure" mathematical line might have some applications (indeed it does) even though I wouldn't run out to the track and start betting. I would also be very surprised if the numbers I posted are without error. I will probably do a 20 horse line for the Derby even though my wife has threatened me with death.
Thanks Trifecta for this "Satanic" formula.

[teddy]

I always have a chuckle over stuff like this:

The presumption that people using software cannot figure something out with an algorithm therefore it is somehow easier without one.

I chuckle when someone sees interest in a topic an decides to try to do an workshop on it to profit from it. One after another.. workshop after workshop...if you were making a dime with all this great info you have then why would you waste your time doing workshops.....I know some guys making 500 k last year betting certain races that I wont talk about on here. They dont have the time to sell workshops.

[maddog42]

It is still possible to find overlays...but the whales are a big obstacle for us, for several reasons.

These whales are using much more sophisticated handicapping techniques than we are...and they spot these overlays more efficiently than we do. And then, they employ a wagering method which pounces on ALL these overlays, wherever they can be found...with the bulk of the wagering taking place at the last few seconds...after the average player has placed his wager.

So, we have a powerful opponent here...who is doing the exact same thing that we are trying to do (spot overlays), but his handicapping techniques allow him to be MUCH more proficient at it than we are...

AND...he gets to play his hand after we have played ours.

How can we effectively compete against him; because, the way he places his bets, there is no way for us to play "around" him.

We mere mortals are reduced to a guerrilla-type attack...striking the occasional overlay here and there...

No?

Ok, so basically you are saying that the whales are making us their Bitch. Sorry ladies. Well I don't give them God-like wagering powers. They almost surely make better lines than we do, but that doesn't mean they don't make mistakes. You said yourself :more than one way to play this game and make a profit. 2% doesn't sound like they are in betting heaven yet. The truth is we don't know what kind of monster whales are swimming out there in the parumutuel pools waiting to gobble us up. We probably aren't privy to that kind of info.
I sure like your "guerrilla-type attack" analogy.

[setup]

I chuckle when someone sees interest in a topic an decides to try to do an workshop on it to profit from it. One after another.. workshop after workshop...if you were making a dime with all this great info you have then why would you waste your time doing workshops.....I know some guys making 500 k last year betting certain races that I wont talk about on here. They dont have the time to sell workshops.

You know, I was thinking the same thing the other day. I was reading in another thread, the one about winning $300k a year, where the poster you quoted wrote about having 1st hand knowledge, I believe is the way he phrased it, about winning $300k a year betting. And I thought to myself, here's a guy who's continually hawking his products, I mean, putting in a ton of constant effort -- including all the time he's spent rebutting detractors of his systems. Why would he do this if he were 'profitable'; significantly profitable? Guess he just wants to help other in better understanding the game.:rolleyes: Of course, having this dude get his hooks into you must not be a very good feeling overall. Can't wait for the next announcement of a big NEW PACE hit. Funny how those only show up when there's a good hit. Guess there's no action between hits.

[teddy]

It goes on and on and on year after year... The only ones I know of that are making real monies are specialists and found and edge against the public. Not systems.. One thing they have in common is they play only certain tracks... no more than 2 or 3. Watch every video ,know every horse, trainer ect. Forgive me Mountainman but you are one of those, I don't pretend to know your roi but I know since I specialized in OUR track, I have a flat bet profit there. On all wagers in all catagories.. thats proof to me that specialization is very profitable. I watch all the races at MTN, have learned all the trainers angles.. Know all the nightly bias thanks to Mountainman. I also know of a guy that bets just HOOSIER trots. Same story, he knows all of the horses. Never had a losing year. There is no way an outsider can switch over to MTN and bet against those that specialize and win. They are bettng off bad info. false beyers ect.

[TexasDolly]

Jeff, That was a nice summary of the odds considerations.
Thank you for taking the time to present it.
Thread drift sure does tend to close down a productive thread doesn't it ?
TD

[thaskalos]

Ok, so basically you are saying that the whales are making us their Bitch. Sorry ladies. Well I don't give them God-like wagering powers. They almost surely make better lines than we do, but that doesn't mean they don't make mistakes. You said yourself :more than one way to play this game and make a profit. 2% doesn't sound like they are in betting heaven yet. The truth is we don't know what kind of monster whales are swimming out there in the parumutuel pools waiting to gobble us up. We probably aren't privy to that kind of info.
I sure like your "guerrilla-type attack" analogy.
Maddog...I still think that the game is worth playing, but the competition is getting tougher.

I have been around this game for a long time, and so have you...and we have both seen the gambling crowd at the racetracks and the OTBs reduced to a small fraction of what it once was.

The unsophisticated player is being forced out of this game...and this will soon become a battle between the "sharks" and the "whales".

Allow me to use another analogy...this one from the poker table:

Let's say that I have been playing in a regular poker game for years now, against the same nucleus of players...and a new, ultra-aggressive, super-sophisticated player suddenly shows up at my table.

He starts attacking every situation he deems profitable...and pounces on every conceivable edge he can find. He is so active, there is no way I can play around him. And his bankroll is getting bigger and bigger!

He has become more than an annoyance; he is a legitimate threat to my survival as a winning player.

And if I cannot find some way to counter-attack his playing style...then I will inevitably be forced to look for another game to play.

This is where the winning horseplayer finds himself today...IMO.

There is no doubt about it...the winning horseplayers make their profits from those less sophisticated (or less disciplined...or less committed) than themselves. The winners take the money away from the losers.

As the "losers" are filtered out of the game, with little "new blood" coming in to replace them...the game soon becomes a battle among WINNERS.

And, with the takeouts being what they are...this type if battle in unwinnable, IMO.

That's why I am still holding on to poker, as a side...:)

[TexasDolly]

Thask, Your opponent in poker still pays the same rake as you do and that's not the case in horse racing. He can't make your best play or situation a negative expectation.
I guess though, he could bust all the weaker players before you get a chance at them.:)

TD

[jerry-g]

First of all, I feel I am the last person to give advice to you as I am sure you are way beyond me in this game. I have said in previous posts that I have been lousy at picking winners. My problem is not in picking winners but in playing too many races and therefor loosing in the long run. So I refined myself like you are trying to do now with win wagering and have some new insights I'd like to share with you.

As an example, today in Churchill race 3, I had decided that the 11 horse was going to win if he ran as he was an also eligible. My second pici was to be the 3 horse and then the 4 but he got scratched and my 9 horse was up for third on my handicapping. They came in 11-3-9 and I did not make a bet. The reason is because I am not picking winners now but looking for overlays. The 11 went off at 6/5 and the 3 was 3/1. My game plan now is to bet the winner I pick if an overlay and if not, go with my second choice at 12-1 or better for a place bet. The 3 went off at 3-1 so no bet. I still feel like I made money because I kept my bankroll in my Pocket.

This way of playing is very boring to me because I seldom get to make a bet. I dry up waiting for that nice overlay. But in the long run I feel I will be profitable. Some days, on $2 bets I may only make $30 but if you increase your flat bet to $10 that would translate into $150 on those days.

I'm not suggesting anyone adopt my way of playing, I offer this as food for thought only. You do have to have a fair handicapping system for getting the ITM horses in at least your top 4 picks. Then it takes some skill to decide who is the stronger out of that bunch. For now, I do not play exotic combinations.

Best of luck to all.

[TrifectaMike]

[QUOTE=teddy]My exercise on picking first call leaders holds the one key to this type of wagering. Whether or not a whale can come up with a program to find the first call leaders is beyond me, but I think they probably could do a good job of it based on post position and ability of each horse, along with early lick. One thing that never changes is that the first call leader is golden.

Teddy,

If you have empirical evidence that justifies your first call premise, I will suggest you treat the race to the first call as the race in total. Create an oddsline as you would if you were attempting to determine the odds for the race.

Once you have either a probability line or odds line for the first call seek overlays by comparing your first call oddsline to the the tote odds.

For example:

First Call Odds for Horse A = 5/2
Tote Odds for Horse A = 4-1

In your world this would constitute an overlay. Obviously, it would require some development and test, but it is an approach that can work.

Mike (Dr Beav)

[TexasDolly]

First call horses were tremendously profitable in the late 80's and early 90's. I played them religiously while playing cards when I lived in LV. I had a 27/28 bet losing streak at one point and only lost back about half the profit. The word got out and people also improved their handicapping skills. The software developers shouted out that they had found Mecca and the big edge went away. While I am very aware of it's importance I can no longer isolate the 1st call leader often enough to make money.
Good luck Teddy,
TD

[overthehill]

what i find to be a major difficulty so far for me anyway is that when i go to bet i see a price but that price can change significantly between the time i bet on it and post time especially if it is either of the first two favorites in the ML. also frequently horse from outside posts take very late play. Today I passed on a horse that had the outside 8 post in a turf race at monmouth park it was 5-1 morning line 12-1 withone minute to go. because of the post i thought i could see a 3-4 point drop by post time so i passed the race. instead the odds ballooned to 19-1 and the the horse lost by a bob.

at fairgrounds a few months ago I tried some horse from an inside post position who i thought was an overlay at 17-1. I figure i was ok to be becauase of the inside post. that one went fom 17-1 to 7-1 in the last flash or so and ran second to an odds on favorite.
Many of my theoretical wagers have less margin for error than the last minute fluctuations i perceive in the tote unfortunately.

I have pretty much given up looking at horses with ml of 3-1 or less because they seem to almost always get spanked down in the last minute or so of play. One case i remember it being 9/2 in new york with 0 minutes i went to bet it . came back they loaded into the gate it dropped to 7/2. they took off it was 5/2.

[jerry-g]

As a post script to my above post, I had a dud of a day today. I handicapped 7 races and found only two to bet at odds I believed overlay. The first bet the horse came in 3rd and my second bet was the 9th race at Arlington Park the mothers day finale. I bet the 7 horse to win and he came in 2nd. He went off at 15-1 odds and duh...my own advise is to bet a place on that type for an insurance against loss. I think he paid $12 to place. Would have been better than my $2 loss. I think I've found a way to beat myself. I will now return you to your other lives and you are free to roam about the countryside. Myself, I'm returning to my Tahitian Sunset to make sure it sits well and ponder great races of the future.

[Fox]

A little gift for all of you from me....

P(N,R) = 1/N (1/R + ... + 1/N) ... A probability generating function... Dave is correct you probability need a Ph.D in math to figure this out (Don'r even try...just accept the equation)

N -> Number of horses in the race
R -> Rank of horse's odds on the board

Example:

N -> 10 horse field
R -> Horse A ranked (by odds) 7th on the board

P(10,7) = 1/10 (1/7 + 1/8 + 1/9 + 1/10) ... Probability of Horse A winning as ranked by odds by the public.

You need to think carefully about what the equation is modeling. You might say, but the probability is defined by the odds...What is this probability, P(N,R) saying to me.

Mike (Dr Beav)

Any chance there is a similar formula yielding a horse's probability to get 2nd place given field size and odds rank?

Thanks

[PaceAdvantage]

It goes on and on and on year after year... The only ones I know of that are making real monies are specialists and found and edge against the public. Not systems.. One thing they have in common is they play only certain tracks... no more than 2 or 3. Watch every video ,know every horse, trainer ect. Forgive me Mountainman but you are one of those, I don't pretend to know your roi but I know since I specialized in OUR track, I have a flat bet profit there. On all wagers in all catagories.. thats proof to me that specialization is very profitable. I watch all the races at MTN, have learned all the trainers angles.. Know all the nightly bias thanks to Mountainman. I also know of a guy that bets just HOOSIER trots. Same story, he knows all of the horses. Never had a losing year. There is no way an outsider can switch over to MTN and bet against those that specialize and win. They are bettng off bad info. false beyers ect.You and setup are full of it. First off, New Pace isn't a "system" for picking a winner. It's more of a contender selection method.

Second, I suppose it's your contention (along with Mr. "Setup") that NOBODY has ever produced ANYTHING for sale that was worthwhile. This is undeniably false.

Third, who's to say you can't do both? Wager profitably AND sell your knowledge? What person wouldn't want to do both if they could? Beyer certainly falls under this camp, as does every other purveyor of numbers. But of course, I'm sure you'll tell me nobody is winning with numbers... :lol:

Are you operating under some false notion that if you sell profitable knowledge, your edge will disappear? That might be true if you sell a magic black box that says "BET THIS HORSE" and nothing else, but as far as I can tell, Dave doesn't have any such product (although I'm sure he's constantly working on one...lol)

And before you come back at me and say "Yeah, well, you're just sticking up for Dave because he advertises here..."

WRONG. I don't charge Dave a single cent to advertise here. Yes, he'll send me a token check for a couple of bucks once in a blue moon out of guilt ( :lol: ), but the reason I don't charge him is because I've been a fan of his since the BRAIN WAVES newsletter days, when I was cutting my teeth in this game back in the early 90s. It's out of pure respect for the man and his products and his work ethic that I don't charge him to advertise here. He's the only one who enjoys such a privilege on this website.

You and setup are basically clueless when it comes to Dave, so you should really pipe down. Ignorance is a terrible thing to put on full display.

[davew]

I suspect the whales find races which allow them to dutch their top contenders - whether it means skipping a favorite with too much bet on it, or narrowing the field to 2-4 horses that will win the race a high percentage of the time.

Since it is difficult to predict how much money will come in and on whom after the race has started - if they have the winner in their cluster, some will have increased in odds while others will decrease.

I have been having some success with this but still refining - it works better for me using my oddsline for dutching rather than tote - and decide to pass or play

[Dave Schwartz]

I suspect the whales find races which allow them to dutch their top contenders - whether it means skipping a favorite with too much bet on it, or narrowing the field to 2-4 horses that will win the race a high percentage of the time.

1. They play almost every race on almost every card.

2. They are typically not dutch bettors.

3. They use "Kelly" to size their wagers.

[teddy]

OKay sure you have your opinion, but I know hes got special status here and everyone else does. I really have to disagree with you cause I dont see anyone working that hard for pennies when they are making bank. Its your board and you can cut off anyone thats not in your club. Wont stop anyone else from thinking its a joke for him to keep coming on here selling stuff to newbies. My question is why you get so mad?? Why the quick jump to defend?

[teddy]

[QUOTE=teddy]My exercise on picking first call leaders holds the one key to this type of wagering. Whether or not a whale can come up with a program to find the first call leaders is beyond me, but I think they probably could do a good job of it based on post position and ability of each horse, along with early lick. One thing that never changes is that the first call leader is golden.

Teddy,

If you have empirical evidence that justifies your first call premise, I will suggest you treat the race to the first call as the race in total. Create an oddsline as you would if you were attempting to determine the odds for the race.

Once you have either a probability line or odds line for the first call seek overlays by comparing your first call oddsline to the the tote odds.

For example:

First Call Odds for Horse A = 5/2
Tote Odds for Horse A = 4-1

In your world this would constitute an overlay. Obviously, it would require some development and test, but it is an approach that can work.

Mike (Dr Beav)

Darn, Mike thats an incredibly good idea. I have exact evidence that the first call leader is easily profitable.. Well over 1.20 or more if you can flat bet them.

[Dave Schwartz]

OKay sure you have your opinion, but I know hes got special status here and everyone else does. I really have to disagree with you cause I dont see anyone working that hard for pennies when they are making bank. Its your board and you can cut off anyone thats not in your club. Wont stop anyone else from thinking its a joke for him to keep coming on here selling stuff to newbies. My question is why you get so mad?? Why the quick jump to defend?

You are a really short man, right?

[raybo]

A little gift for all of you from me....

P(N,R) = 1/N (1/R + ... + 1/N) ... A probability generating function... Dave is correct you probability need a Ph.D in math to figure this out (Don'r even try...just accept the equation)

N -> Number of horses in the race
R -> Rank of horse's odds on the board

Example:

N -> 10 horse field
R -> Horse A ranked (by odds) 7th on the board

P(10,7) = 1/10 (1/7 + 1/8 + 1/9 + 1/10) ... Probability of Horse A winning as ranked by odds by the public.

You need to think carefully about what the equation is modeling. You might say, but the probability is defined by the odds...What is this probability, P(N,R) saying to me.

Mike (Dr Beav)

Trifecta Mike,

Just checking my math. Is the answer 4.79% probability?

By the way, is "P(N,R)" saying that's the probability of the odds "rank", in a certain field size, reflecting the public's confidence in each of the horses in the field to win, not anything else about the horse's actual odds or merit? The "Large Numbers" thing is, of course, of some merit.

[PaceAdvantage]

OKay sure you have your opinion, but I know hes got special status here and everyone else does. I really have to disagree with you cause I dont see anyone working that hard for pennies when they are making bank. Its your board and you can cut off anyone thats not in your club. Wont stop anyone else from thinking its a joke for him to keep coming on here selling stuff to newbies. My question is why you get so mad?? Why the quick jump to defend?Oh God, give me a break with this "club" crap and "special status."

I've already told you, unlike everyone else, I don't charge him to advertise here. I get no percentage of his sales. If he has "special status" here, why do you think that is? Maybe a bright boy like you can figure it out.

I'm quick to defend, because unlike you, I actually know the man and his products and have respect for the work he does. It's quite simple really. There, I figured it out for you in case you're not too bright after all...

[MPRanger]

I have drifted away from win wagering in recent years, and have been thinking as of late of getting back into it...seeing that today's short fields are not the ideal battleground for the super-exotic wagering I am naturally attracted to.

And I know that this will require creating an odds line...which is something I am not really looking forward to.

There are a few questions I have about win-betting using an odds line, and I would like to direct them at those who have more experience than me in wagering of this type:



thaskalos,

I would not presume to lecture you. You are such a knowledgable
handicapper. But for fun I'd like to take a shot it. I feel strong in
this area.


1.) Do you confine your win wagers to those horses who are legitimate contenders in the race...or do you make serious bets on rank outsiders, just because they are offered at overlaid prices?

Would you make a serious wager on a 25-1 longshot...just because it is offered at 40-1?


First, I would never ever bet a 25-1 shot to win
even if it was offered at 40-1 and I saw the trainer
give him the injection.

Never mind the longshot overbet bias . There is only
a 1.41% decimal percentage difference between 25-1 and 40-1.
They are virtually indistinquishable in true chances. From this
you should be able to see the futility of calling the 40-1
shot an overlay. Then there's the takeout but why bother?




2.) Do you bet less on those horses who don't really figure to win the race...or do you wager the same amount, since your odds line tells you that you still have your expected edge on the wager?

It occurs to me that the player should reduce the amount of his wagers when he bets on these overlaid outsiders...to account for the longer losing streaks he is sure to encounter when using wagering of this type.



Technically, your oddsline won't tell you what your edge is. It only
tells you the presumed true chances of winning .

You only know your edge after you know how much you are betting
to win - against however much a win would pay.

If you flip a coin and heads wins and tails loses then you have a 50% chance
to win and a 50% chance to lose the contest. If you are betting even money
on either then there is no edge. But if you lose $1 when tails comes up but
win $1.10 when heads comes up then you have a 10% edge. Heads or tails
equals 1-1 on the toteboard. Though it pays even money, the chances aren't
truly 50-50 due to the takeout. Your edge can come in the form of better
chances to win or more than break even payback. Odds = money. Money
=odds. Either the money you win or the chances you get must overcome
the takeout first. Then you need even more of a premium to over come
final odds which you never know what they will be. Just expect the worst.
You have to decide how much of a premium you want to play for. Barry
Meadow say's 50%. Dick Mitchell says 20 or 25%.

But your first task will be to make your odds line. It's about the simplist
thing in the world to do. I'm certain you will enjoy doing it. Somehow, I
assumed you already were.

But as far as how much to bet ...... depends on a couple of things but
I believe the Kelly criterion is absolutely the way to go. When I get
situated and setup I will use it. Right now I don't have a bankroll.
I have a sailboat instead. :cool:

The two experts I mentioned above say to never bet anything over 6-1.
I personally won't take anything over 4-1. I don't care if they are overlaid
or not. I prefer a short priced overlay but that's my own superstition.
Plus, it has to do with the size of your bankroll. If you have a large bankroll
and you want to put it into play properly then it probably is the correct thing
to bet a 6-1 (on your line) overlay.



3.) When there is more than one overlay in the race, is it better to dutch those horse...or should the wager be placed on the horse offering the biggest edge?

Thank you in advance for your replies; your help is greatly appreciated.

I'll take the one with the shortest price every time.
But if they are close to being equal on my line, I'll use proportional betting.

[thaskalos]

thaskalos,

I would not presume to lecture you. You are such a knowledgable
handicapper. But for fun I'd like to take a shot it. I feel strong in
this area.



First, I would never ever bet a 25-1 shot to win
even if it was offered at 40-1 and I saw the trainer
give him the injection.

Never mind the longshot overbet bias . There is only
a 1.41% decimal percentage difference between 25-1 and 40-1.
They are virtually indistinquishable in true chances. From this
you should be able to see the futility of calling the 40-1
shot an overlay. Then there's the takeout but why bother?




Technically, your oddsline won't tell you what your edge is. It only
tells you the presumed true chances of winning .

You only know your edge after you know how much you are betting
to win - against however much a win would pay.

If you flip a coin and heads wins and tails loses then you have a 50% chance
to win and a 50% chance to lose the contest. If you are betting even money
on either then there is no edge. But if you lose $1 when tails comes up but
win $1.10 when heads comes up then you have a 10% edge. Heads or tails
equals 1-1 on the toteboard. Though it pays even money, the chances aren't
truly 50-50 due to the takeout. Your edge can come in the form of better
chances to win or more than break even payback. Odds = money. Money
=odds. Either the money you win or the chances you get must overcome
the takeout first. Then you need even more of a premium to over come
final odds which you never know what they will be. Just expect the worst.
You have to decide how much of a premium you want to play for. Barry
Meadow say's 50%. Dick Mitchell says 20 or 25%.

But your first task will be to make your odds line. It's about the simplist
thing in the world to do. I'm certain you will enjoy doing it. Somehow, I
assumed you already were.

But as far as how much to bet ...... depends on a couple of things but
I believe the Kelly criterion is absolutely the way to go. When I get
situated and setup I will use it. Right now I don't have a bankroll.
I have a sailboat instead. :cool:

The two experts I mentioned above say to never bet anything over 6-1.
I personally won't take anything over 4-1. I don't care if they are overlaid
or not. I prefer a short priced overlay but that's my own superstition.
Plus, it has to do with the size of your bankroll. If you have a large bankroll
and you want to put it into play properly then it probably is the correct thing
to bet a 6-1 (on your line) overlay.



I'll take the one with the shortest price every time.
But if they are close to being equal on my line, I'll use proportional betting.

Hi MPRanger,

I have toyed with odds lines for years, but did not really enjoy the experience...probably because I could never create one which worked to my satisfaction.

I had no trouble finding "overlays" with them...but these overlays just would not live up to their potential when all was said and done. :)

It turned out that win-betting was not my cup of tea after all; my temperament seems much better suited to the ups and downs associated with the super-exotics.

But, with all the short fields that the regular player has to contend with on a daily basis...it appears that I need to expand my horizons by revisiting win-betting, and wrestling with odds lines once again...

Thanks for your advice...I greatly appreciate it; and this goes out to all the other posters who have responded to me as well...:ThmbUp:

[TrifectaMike]

Trifecta Mike,

Just checking my math. Is the answer 4.79% probability? Yes

By the way, is "P(N,R)" saying that's the probability of the odds "rank", in a certain field size, reflecting the public's confidence in each of the horses in the field to win, not anything else about the horse's actual odds or merit? The "Large Numbers" thing is, of course, of some merit. Yes, but read Post #70 and Jeff's Posts



Mike

[TrifectaMike]

Any chance there is a similar formula yielding a horse's probability to get 2nd place given field size and odds rank?

Thanks
I don't have one, but I do have a formula for ranking of ordered sets ( exactas ).... however, I doubt I'll make it public.

Mike (Dr Beav)

[MPRanger]

Let's say you have a race with three profitable contenders. If you apply the Kelly Criterion to each individual wager it might tell you to bet (say) 8% of bankroll. However, if you Dutch the three horses together, that is you treat the entire race as a single bet, the optimum that is liable to double or triple.


Word.

[teddy]

On the idea of an early speed line.. I have my 10 factors that lead horses have in common. Each factor I am only looking at the top horse, per say best beyer or fastest first fraction. Should I assign each horse in the race 2pts for having this factor and then have an idea of who has more common early speed factors. After which assign the likelyhood that they will make the lead at first call by giving them odds. Such as.

horse 3 12 pts 2-1
horse 6 8 pts 3-1
horse 9 2 pts thus odd requirment of 10-1


Something of this fashion?? Dont go too crazy with the math on me. Thanks

[TrifectaMike]

On the idea of an early speed line.. I have my 10 factors that lead horses have in common. Each factor I am only looking at the top horse, per say best beyer or fastest first fraction. Should I assign each horse in the race 2pts for having this factor and then have an idea of who has more common early speed factors. After which assign the likelyhood that they will make the lead at first call by giving them odds. Such as.

horse 3 12 pts 2-1
horse 6 8 pts 3-1
horse 9 2 pts thus odd requirment of 10-1


Something of this fashion?? Dont go too crazy with the math on me. Thanks

Teddy,

If you are going to assign points (intutively, I assume and lots of testing), I would sum all the points awarded to each horse and use that sum to assign a percentage allocation to each horse.

Your example:

horse 3 12 pts, p=12/22 = .55 [ odds = 1/p -1] .83 to 1
horse 6 8 pts, p=8/22 = .36 [ odds = 1/p -1] 1.75 to 1
horse 9 2 pts, p=2/22 = .09 [ odds = 1/p -1] 10 to 1
sum = 22 points

Good luck,
Mike (Dr Beav)

[Light]

For me an odds line is useless for a myriad of obvious reasons.

The bottom line is you must have a "positive expectancy" if you want your ROI to be positive.Simple. A negative expectancy such as an odds line creates is not going to create a positive expectancy.(at least for me)

Mathematically expectancy is expressed as follows:

Expectancy=(probability of win*Ave win) - (probability of loss*Average loss)

Example: (25% hit rate*$9 ave win) - (75%*$2 per average loss)

Remember to take out the $2 invested from the Ave win which would make it a $7 profit per hit.

($175) -( $150) = $25 profit which would be converted to an ROI depending on the number of races.That is a +16% ROI.


Horses are not machines and handicappers are not robots. I don't go whipping out all these mathematical equations when I bet. But I do keep an internal "positive expectancy" when I bet.For example:

My last bet I was faced with playing a pk3 or a dd. I was sure of the dd. I felt a positive expectancy from the dd. The proceeding race from the dd was the start of the pk3 and I felt unsure of that race. I felt a negative expectancy from the pk3.

So its a no brainer right? Just play the dd. But I knew the pk3 was going to pay way better cause I had the chalk in the last.

The trick to making a profit was to accept my own limitations and go with the expectancy. Sure enough a bomb won the 1st leg of the pk3 which I wisely passed and I did hit the dd.


I could "see" the dd. I could not "see" the pk3. Expectancy was obvious and I ended up with a profit. Much smaller than the pk3 but I knew I made the right move by taking the positive expectancy rather than the negative expectancy.

[misscashalot]

....
The trick to making a profit was to accept my own limitations and go with the expectancy. Sure enough a bomb won the 1st leg of the pk3 which I wisely passed and I did hit the dd. ....

Ever hit the all button?

[Light]

Ever hit the all button?

Not the point.

[Light]

To expand on that. Since I could not "see" the 1st leg of the pk3, it did not automatically mean a bomb would win. It could have just as well been the favorite. Having no clue does not mean I don't like the favorites only. It means I don't like anything so I have no business playing the race and expecting "positive expectancy".

[TrifectaMike]

But your first task will be to make your odds line. It's about the simplist
thing in the world to do. I'm certain you will enjoy doing it. Somehow, I
assumed you already were.


Much of what you write is meaningless, but the above sentence is one of your best.

Mike (Dr Beav)

[HUSKER55]

My ability at making an odds line that works is a little weak. ( ok..ok...pretty bad.:D ) Suppose you have an 8 horse field and there are 3 horses that you are not going to bet come hell or high water.

Am I better off to just figure a five horse field, each horse has a 20% chance of winning, or would it be smarter to assign odds of 12:1 to each of the horses that I am not going to bet and then make the other 5 horses that I would consider less than 10:1.

[Overlay]

My ability at making an odds line that works is a little weak. ( ok..ok...pretty bad.:D ) Suppose you have an 8 horse field and there are 3 horses that you are not going to bet come hell or high water.

Am I better off to just figure a five horse field, each horse has a 20% chance of winning, or would it be smarter to assign odds of 12:1 to each of the horses that I am not going to bet and then make the other 5 horses that I would consider less than 10:1.
If I recall correctly, the late Dick Mitchell recommended allocating 20% of the total probability to the non-contenders (what he euphemistically called the "Stuff Happens" pile), without assigning individual odds to the horses in question, and then parceling out the remaining 80% among the horses that you consider contenders (but with fair odds based on a 100% line, rather than 80%). (For example, if you gave one of the contenders a 20% chance of winning, its fair odds would still be 4-1 (80-20), rather than 3-1 (60-20), as they would be if based on an 80% line.) (That's not how I personally proceed, but his recommendation sounds similar to what you're talking about.)

[Overlay]

The bottom line is you must have a "positive expectancy" if you want your ROI to be positive.Simple. A negative expectancy such as an odds line creates is not going to create a positive expectancy (at least for me).
I agree with your first sentence above, but I'm not sure that I understand the second sentence. To me, creating a fair-odds line is a key to determining which horses or combinations have actual odds or payoffs that offer the positive expectancy that you're looking for. How does the creation of that odds line by itself result in a negative expectancy?

[HUSKER55]

THANKS FOR THE INFO OVERLAY:)

[erikeepper]

If I recall correctly, the late Dick Mitchell recommended allocating 20% of the total probability to the non-contenders (what he euphemistically called the "Stuff Happens" pile), without assigning individual odds to the horses in question, and then parceling out the remaining 80% among the horses that you consider contenders (but with fair odds based on a 100% line, rather than 80%). (For example, if you gave one of the contenders a 20% chance of winning, its fair odds would still be 4-1 (80-20), rather than 3-1 (60-20), as they would be if based on an 80% line.) (That's not how I personally proceed, but his recommendation sounds similar to what you're talking about.)

This is a good strategy and totally adequate for most of us. Saves a ton of time also. Good idea is to first keep tabs on how your contenders actually do (win%, roi) and only when you know your contenders actually win 80 %+ and perform a lot better than the non-contenders, you proceed to make the line.

This way you usually make a line to your top 4-5 horses which is fine because it is very hard to make prob estimates on horses lower than that and the errors are huge (2 % vs 4 % -> 49-1 or 24-1 and can you really know the difference between the two?) and usually you can't bet a lot on those horses anyway without their odds dropping.

Orcourse the above work fine in win betting but if you play P4 or exacta or whatever, you can use your overlay contenders as the key horse.

[Dave Schwartz]

This is a good strategy and totally adequate for most of us. Saves a ton of time also. Good idea is to first keep tabs on how your contenders actually do (win%, roi) and only when you know your contenders actually win 80 %+ and perform a lot better than the non-contenders, you proceed to make the line.

IMHO, this is a very good start on a system for making a line.

So, you have 4 or 5 horses that get 80% between them. What now?

How do you decide who gets what cut of the 80%?


Dave

[ezpace]

It goes on and on and on year after year... The only ones I know of that are making real monies are specialists and found and edge against the public. Not systems.. One thing they have in common is they play only certain tracks... no more than 2 or 3. Watch every video ,know every horse, trainer ect. Forgive me Mountainman but you are one of those, I don't pretend to know your roi but I know since I specialized in OUR track, I have a flat bet profit there. On all wagers in all catagories.. thats proof to me that specialization is very profitable. I watch all the races at MTN, have learned all the trainers angles.. Know all the nightly bias thanks to Mountainman. I also know of a guy that bets just HOOSIER trots. Same story, he knows all of the horses. Never had a losing year. There is no way an outsider can switch over to MTN and bet against those that specialize and win. They are bettng off bad info. false beyers ect.
************************
Exaxtly ex. circuit like GP to KEE ,BEL,and SAR with an edge few have
done with puter or pencil n paper
hint : not ranking odds but something else compared to real time odds
complete method for exacta,tri,& super ONLY easily picks up chicanery ,best pace,
best contenders,best speed & Class. 3minute capping per race topz
proprietary*

[erikeepper]

IMHO, this is a very good start on a system for making a line.

So, you have 4 or 5 horses that get 80% between them. What now?

How do you decide who gets what cut of the 80%?


Dave

Hi Dave,

yes the first part (getting the contenders) is "easy" but the 2nd part is far more difficult :)

What I have done is following:
1) get contenders
2) rank them
3) and then just swing it, asking "at what odds would I play this horse?" and keep records how it went...

I don't have a strict mechanical system for this.

[maddog42]

Nobody said anything about validating or testing their oddsline. My oddsline has various weaknesses and strengths depending on the category. My oddsline at 7 furlongs seems very weak to me, despite starting off very strong. I think track bias makes this one of the toughest distances. At some tracks I am over 85% in the top 4. At others I am under 60%. This might be due to pure old randomness,
but I doubt it. The sample size is too small.
Use the top 4 percentages when you start out, and only when you have a good sample size go to the ranking percentages. Despite what some people have said,
Making a good Oddsline is hard. Why do you think so few people do it?

[maddog42]

The comment by Thask, about the field size, got me to thinking in a critical way about my own oddsline. In a 6 horse field having the winner in the top 4 80% of the time is not very impressive.In a 6 horse field, throwing darts at the form will get you 75%. In a 12 horse race it is very impressive. Thask is a damn troublemaker!!!

[Dave Schwartz]

What I have done is following:
1) get contenders
2) rank them
3) and then just swing it, asking "at what odds would I play this horse?" and keep records how it went...

I don't have a strict mechanical system for this.

I do. And in about a month I will be sharing it with a portion of our world. It will just blow your mind.

Stay tuned.


Dave

[thaskalos]

Thask is a damn troublemaker!!!

I've been called worse...:)

[Light]

I agree with your first sentence above, but I'm not sure that I understand the second sentence. To me, creating a fair-odds line is a key to determining which horses or combinations have actual odds or payoffs that offer the positive expectancy that you're looking for. How does the creation of that odds line by itself result in a negative expectancy?

If you look at TM's math,it showed that any horse in a field has a small fractional chance of winning and a negative ROI. That is a negative expectancy.

TM's math aside this system is ridiculous.This system has no rhyme or reason and has a built in negative expectancy because it spreads itself thin. Betting all the overlays in a race is going to produce a negative expectancy.

The ML guy may be wrong but you creating your own odds line have just as much of a chance of being wrong as the ML maker who usually knows the horses at his own track well.

As a simple solution,instead of being a self made ML maker of an entire field, I look at the one or two horses I may like in a field and decide if I feel OK with their odds. Why waste my energies creating a ML on horses I'm not interested to bet whether they are overlays or underlays and then reap the ramifications of a negative expectancy.It's a waste of time and mental energy better spent on more important factors.

[thaskalos]

I do. And in about a month I will be sharing it with a portion of our world. It will just blow your mind.

Stay tuned.


Dave

I must interject and make a comment here...

I have been a very vocal critic of the "true" effectiveness of odds lines for over two decades now...

It's not that I lack the sophistication to appreciate how valuable a good odds line can be for the player; I just didn't believe that it was possible to create an odds line which could be profitable against the toteboard...which, after all, is what we are all trying to do.

Well...I spent 3 hours on the phone with Dave Schwartz last monday...and now, I'm not so sure.

Dave has come up with a new way of creating an odds line...which is sure to be an eye-opening experience for the knowledgeable player. I know that it was for me...

I promised him that I wouldn't say much about this product yet...but I will say this:

I highly encourage you guys to take a look at this product when it comes out.

Sometimes...it's not necessary to reinvent the wheel; all we need to do is to look at the old game with a new set of eyes...

[maddog42]

Simple question. Simple answer: The best horse doesn't always win. Why? Everybody knows the obvious answers: trouble during the race: Bad Start, steadied, going wide, jockey brain farts, bad post, wrong time in form cycle, God, hurt during race, spills, horse goes crazy and rank. When one of these happens to the favorite(60% of the time), you have good paying horses. This chaotic aspect of horseracing is what makes it fun (and frustrating and causes mini break-downs among handicappers). I have a question for the genius line-makers out there in cyberspace.

What percentage of the time will the legitimate favorite, 2nd choice and 3rd choice all lose? I say it is at least 25%.

[erikeepper]

I do. And in about a month I will be sharing it with a portion of our world. It will just blow your mind.

Stay tuned.


Dave

Month is bit too long, can't wait! :)

Can it be done without HSH?

I absolutely loved your BOW-course because it was totally applicable to harness, thoroubreds, usa, uk, europe you name it. The ideas were universal... and profitable! I just "know" this is going to be another eye-opener :)

[thaskalos]

Simple question. Simple answer: The best horse doesn't always win. Why? Everybody knows the obvious answers: trouble during the race: Bad Start, steadied, going wide, jockey brain farts, bad post, wrong time in form cycle, God, hurt during race, spills, horse goes crazy and rank. When one of these happens to the favorite(60% of the time), you have good paying horses. This chaotic aspect of horseracing is what makes it fun (and frustrating and causes mini break-downs among handicappers). I have a question for the genius line-makers out there in cyberspace.

What percentage of the time will the legitimate favorite, 2nd choice and 3rd choice all lose? I say it is at least 25%.

You say that..."the best horse doesn't always win".

I'll take it a step further than that:

The horse that LOOKS to be the best, will lose MOST of the time!

Why?

Because the horse who LOOKS to be the best in the race, often is NOT the best in the race.

There are unknown elements in this game, which lie outside our field of knowledge.

Someone mentioned sports betting...and how you can't bet on sports without making a "line".

I agree; but line-making is much simpler and much more straight-forward in sports...because all the information we need to assess a team is "out there"...and we can get access to it.

How can I possibly know if the horse who ran a great race 4 weeks ago is in the same shape today, so he can repeat the effort?

28 days have passed since he last ran...and I have no knowledge AT ALL about what has happend to this horse since then.

And I am supposed to sit down and calculate an odds line...as if the horse's last race occurred yesterday?

The best horse doesn't just lose because of something that occurred in the race.

He often loses because he wasn't the best horse GOING IN to the race...

He just looked as if he was...

And that's why linemaking is such a tricky endeavor...

We are asked to create an odds line...without knowing what the condition of these horses really is.

We treat horses as if they are lines of print in a newspaper...or a computer screen.

[maddog42]

You say that..."the best horse doesn't always win".

I'll take it a step further than that:

The horse that LOOKS to be the best, will lose MOST of the time!

Why?

Because the horse who LOOKS to be the best in the race, often is NOT the best in the race.

There are unknown elements in this game, which lie outside our field of knowledge.

Someone mentioned sports betting...and how you can't bet on sports without making a "line".

I agree; but line-making is much simpler and much more straight-forward in sports...because all the information we need to assess a team is "out there"...and we can get access to it.

How can I possibly know if the horse who ran a great race 4 weeks ago is in the same shape today, so he can repeat the effort?

28 days have passed since he last ran...and I have no knowledge AT ALL about what has happend to this horse since then.

And I am supposed to sit down and calculate an odds line...as if the horse's last race occurred yesterday?

The best horse doesn't just lose because of something that occurred in the race.

He often loses because he wasn't the best horse GOING IN to the race...

He just looked as if he was...

And that's why linemaking is such a tricky endeavor...

We are asked to create an odds line...without knowing what the condition of these horses really is.

We treat horses as if they are lines of print in a newspaper...or a computer screen.

I certainly didn't mean to imply that the best horse doesn't win because of something that happens DURING the race, even though pace matchups can surely change things with a speed horse not firing. I can see that my wording would imply that. It usually is decided before he steps on the track. That is why I put wrong time of form cycle.

Those reasons you gave are a very good argument for an oddsline.
The whale doesn't usually know the condition of the horse either.

[Dave Schwartz]

Month is bit too long, can't wait!

Can it be done without HSH?

I absolutely loved your BOW-course because it was totally applicable to harness, thoroubreds, usa, uk, europe you name it. The ideas were universal... and profitable! I just "know" this is going to be another eye-opener

It can be done without HSH.

Sorry, it will be a month or so.


Dave

[Robert Goren]

Everything that Thaskalos said true. Furthermore it applies to every horse in the race. A mistake on any one horse will result in wrong numbers on all the horses when creating an odds line.

[Overlay]

You say that..."the best horse doesn't always win".

I'll take it a step further than that:

The horse that LOOKS to be the best, will lose MOST of the time!

Why?

Because the horse who LOOKS to be the best in the race, often is NOT the best in the race.

There are unknown elements in this game, which lie outside our field of knowledge.

Someone mentioned sports betting...and how you can't bet on sports without making a "line".

I agree; but line-making is much simpler and much more straight-forward in sports...because all the information we need to assess a team is "out there"...and we can get access to it.

How can I possibly know if the horse who ran a great race 4 weeks ago is in the same shape today, so he can repeat the effort?

28 days have passed since he last ran...and I have no knowledge AT ALL about what has happend to this horse since then.

And I am supposed to sit down and calculate an odds line...as if the horse's last race occurred yesterday?

The best horse doesn't just lose because of something that occurred in the race.

He often loses because he wasn't the best horse GOING IN to the race...

He just looked as if he was...

And that's why linemaking is such a tricky endeavor...

We are asked to create an odds line...without knowing what the condition of these horses really is.

We treat horses as if they are lines of print in a newspaper...or a computer screen.
Yes, but just as statistics can be used to handicap sports without knowing how each individual player is feeling or will perform on any given day, there are also established patterns of thoroughbred performance that have been repeated over and over as long as anyone has been keeping records, including patterns that produce results which lie outside the range of normally expected deviation, and constitute independent variables. While not being able to predict with certainty how a horse will perform in any given race on any given day (any more than it is possible to say with certainty that a baseball player will get a hit in his next at-bat), these patterns do allow for valid, bulk, long-term effectiveness in forecasting the probability of how a horse will perform against a particular set of competitors in a specific race on a given day. Is it a certainty that the horse will win or otherwise perform as forecast? No. But it is eminently possible to isolate cases where the horse has a better chance of doing so than the public gives the horse credit for, which gives the player an exploitable edge. Finding those opportunities is the purpose of odds lines -- particularly odds lines that are statistically based, rather than the result of subjective opinion.

[jasperson]

My oddline program which is usally pretty good a picking out the favorites predicts the odds on the :7: in the 3rd a FL should be 2-1. The morning line odds are 10-1. If I can get over 5-1 I will bet him win place and let's see what happens.

[Skip Away]

Wow.... That's what makes this game interesting.... I've got the #7 as the sixth choice at 22/1! Best of luck!

[thaskalos]

Yes, but just as statistics can be used to handicap sports without knowing how each individual player is feeling or will perform on any given day, there are also established patterns of thoroughbred performance that have been repeated over and over as long as anyone has been keeping records, including patterns that produce results which lie outside the range of normally expected deviation, and constitute independent variables. While not being able to predict with certainty how a horse will perform in any given race on any given day (any more than it is possible to say with certainty that a baseball player will get a hit in his next at-bat), these patterns do allow for valid, bulk, long-term effectiveness in forecasting the probability of how a horse will perform against a particular set of competitors in a specific race on a given day. Is it a certainty that the horse will win or otherwise perform as forecast? No. But it is eminently possible to isolate cases where the horse has a better chance of doing so than the public gives the horse credit for, which gives the player an exploitable edge. Finding those opportunities is the purpose of odds lines -- particularly odds lines that are statistically based, rather than the result of subjective opinion.
The comparison between sports and horse racing, as far as line-making is concerned, is like night and day. There is a world of difference between them.

The stats available to the sports bettor are more comprehensive and more trustworthy than the stats available to the horseplayer.

All the information is "out there" in sports...and we can get access to it rather easily.

But in horse racing...very relevant information about the horse remains hidden from all but the very few.

I'm sure that creating a profitable odds line is possible...and they are being used by some very smart guys right now. But it can't be the straight-forward, "mechanical" process you make it out to be.

These "impact value" lines have been around for many years now...and I have bought a few myself through the years...

The game is different today, IMO...and I don't mean that in a good way.

[maddog42]

It is mathematically impossible to win money in the long term unless a bet gets some type of fair value. 40% winners at $4.80 won't cut it. %10 winners at $19.00 won't hack it either. You are either making a 1 horse line or a 2 horse or (fill in the blank). That method,program, or Ouija board that is getting %30 winners at a $7.40 mutuel probably won't last long. Unless your spirit guide handicapper has a calculator you are in trouble.

[Turkoman]

Wow.... That's what makes this game interesting.... I've got the #7 as the sixth choice at 22/1! Best of luck!

Wow, very interesting! One guy has a horse at 2-1 and the other guy at 22-1. While possessing the unique ability to create odds must be helping some very sharp horseplayers, this example tells us that the odds-creating process must be terribly sinking others!

[thaskalos]

It is mathematically impossible to win money in the long term unless a bet gets some type of fair value. 40% winners at $4.80 won't cut it. %10 winners at $19.00 won't hack it either. You are either making a 1 horse line or a 2 horse or (fill in the blank). That method,program, or Ouija board that is getting %30 winners at a $7.40 mutuel probably won't last long. Unless your spirit guide handicapper has a calculator you are in trouble.
You are absolutely right; without "value", long-term profit is impossible. This is not the question...for me at least.

The question is:

Do we need an odds line in order to find this value?

Can't we create profitable situations without constructing an odds line?

Won't safely eliminating a short-priced horse or two accomplish the same thing?

[maddog42]

Wow, very interesting! One guy has a horse at 2-1 and the other guy at 22-1. While possessing the unique ability to create odds must be helping some very sharp horseplayers, this example tells us that the odds-creating process must be terribly sinking others!

No it does not. BOTH these guys may be very good line makers. Maybe they are both wrong. You guys seem to think there is only one way to do this, and I assure you there are many divergent and successful ways to do this. Everyone makes mistakes in the line making process, even the whales. An absolute correct line? I doubt it. Maybe these two gentlemen will both spot overlayed DIFFERENT
horses in the next race due to an overbet favorite that is going to bounce. One of the horses gets a bad trip and the other wins. One race doesn't prove anything, or 2 or 3. You must be correct enough to offset your mistakes.
One 16-1 shot overcomes a few mistakes.

You are partially correct in that "the odds-creating process must be terribly sinking others!" . I won't argue that the line-making process is sinking some people. So will speed figures or (name the current hot program).

[Turkoman]

First of all, I've never said that the ability to create a solid line is a bad thing. On the contrary, I sincerely applaud those who have that unique ability. It's definitely something I can admire. But when I see two individuals who are at both extremes with this process, it's obvious that there's plenty of confusion out there with all this stuff. I mean, seriously, from 2-1 to 22-1 is like from NYC to LA.

[jasperson]

Wow.... That's what makes this game interesting.... I've got the #7 as the sixth choice at 22/1! Best of luck!
I also have the 2 horse at 2-1 with the 5 3rd. I hope I get your 22-1.

[raybo]

In my opinion, creating an odds line is easy enough but creating a good odds line, one that properly assesses one horse's chances versus another's, is not easy.

Unless 2 odds line creators use the same factor(s) to rank/grade the horses, and use the same method for assigning odds, then the 2 lines will be different. That's not to say that both lines cannot be good ones, or bad lines. For one race, one could be good and the other bad, in another race just the opposite could be true, in another race both could be bad or good.

It's the overall, cumulative, value of the process that determines if the line is worth anything or not.

Much record keeping and conscientious analysis of the line's long term effectiveness is required in order to validate or improve a line and it's creation process.

So, unless you're prepared to do all the work to validate and /or improve the line/line making method, you're wasting your time, if you're goal is to stay, or become, profitable.

[Skip Away]

Hi jasperson.... Please know I intended no disrespect when I posted.... I have no real interest in this race (restricted $4K claimers who haven't won in over a year).... However, with that said, I might consider #5 at 7/1 or higher.

Regardless, I was hoping to see "others" put up their oddsline to spark some discussion as I find this a very interesting topic.... I hope you get your price and a winner on #7! Again, sincerely, best of luck to you!

[TrifectaMike]

For those of you who apply the Benter-Woods "odds smoothing" (second step regression) to their logistic regression tell me (us) which line is dominate in the second step regression (which of the two probability vectors has the greater influence).

Secondly are the coefficents for the second step regression static or dynamic? In other words, are they the same for all population of races?

Mike (Dr Beav)

[Light]

Wow.... That's what makes this game interesting.... I've got the #7 as the sixth choice at 22/1! Best of luck!

Just curious how you use your odds line? Do you bet all your overlays from your odds line?

[maddog42]

Yeah I thought that would get your attention. Better be conservative while starting out line-making. If you get wild with your contenders, then Robert Goren is right, you will bet on a lot of non-contenders that don't even have a
12-1 chance of winning. You have got to be conservative here. Being wild with contenders has cost me a lot of money.
What a sense of Deja-vu. Jim Bradshaw told me the same thing 20 years ago.

I know what is going to happen next. You are going to get very conservative with your line making and pass 8 races in a row, and 1 of those races will be won by a 10-1 shot that you might have had. Big deal. I probably just saved you money in the long run, and you didn't lose a damn thing.

Another mistake I still make, Don't leave out the favorite!! I know what you are saying: What kind of moron leaves out the favorite? I do. As you may have noticed I have an inflated ego, especially when I am on a roll. When I hit back to back $25 horses I have to physically restrain myself from prancing around the room/track like an idiot bragging to people. Don't do this. It is rude. And the next stupid thing I do is think that I am twice as smart as the crowd and I give the favorite 0 chance of winning. This is the best reason to do as Steve Fierro says and complete your line at home.

Even if you think the favorite is going to bounce like a rubber ball, or get locked in a speed duel, or break his leg, Make him one of your contenders.
Because lets face it, some favorites COULD break a leg and still win.
You don't have to make him the favorite in your line. Make him the D horse if you have to. The reason for this is it will make a more accurate line.

By the way, I am a liberal on most things.

[Dark Target]

For those of you who apply the Benter-Woods "odds smoothing" (second step regression) to their logistic regression tell me (us) which line is dominate in the second step regression (which of the two probability vectors has the greater influence).

Secondly are the coefficents for the second step regression static or dynamic? In other words, are they the same for all population of races?

Mike (Dr Beav)

I assume you are talking about the public probabilities estimates versus his (theirs) own?

[Light]

Horse ML Betline Off odds result Bet?

1 12 16 15 out 0
2 5/2 2 7/5 3rd 0
3 3 6.8 7 out -2
4 15 24 40 out -2
5 5 7/2 8 out -2
6 4 4 2 2nd 0
7 10 21 7/2 1st 0







The "Betline" is from "Skip's" line he posted. Not criticizing him or his handicapping. But this is a good illustration of how useless and counterproductive odds lines are.

The 3 overlaid horses according to the betline and ML ran out. While the 3 underlaid horses according to the betline and ML all hit the board.

You are going to be on the back end of the homeless food line before you make a profit with odds lines.

[TrifectaMike]

I assume you are talking about the public probabilities estimates versus his (theirs)(Yours?) own?

That is correct.

Mike

[jasperson]

Hi jasperson.... Please know I intended no disrespect when I posted.... I have no real interest in this race (restricted $4K claimers who haven't won in over a year).... However, with that said, I might consider #5 at 7/1 or higher.

Regardless, I was hoping to see "others" put up their oddsline to spark some discussion as I find this a very interesting topic.... I hope you get your price and a winner on #7! Again, sincerely, best of luck to you!
I know you didn't intend any disrespect when you posted and put up your oddsline for the race. I like you would like to see others post their oddslines. The reason I posted is because I thought the 7 would be a huge overlay anybody can post just favorites because most oddslines(including mine) just pick out the favorites. By the way the 7 won and paid 9.70 he went off at 7/2 also I didn't bet him because he wasn't 5/1 or better.
Jack

[Dark Target]

That is correct.

Mike

I would ne extremely surprised if anyone's is... besides, your objective isn't to build a model that rivals the public in predictability

[AITrader]

For those of you who apply the Benter-Woods "odds smoothing" (second step regression) to their logistic regression tell me (us) which line is dominate in the second step regression (which of the two probability vectors has the greater influence).

Secondly are the coefficents for the second step regression static or dynamic? In other words, are they the same for all population of races?

Mike (Dr Beav)

The coefficients aren't that meaningful without a frame of reference as they depend on how the two are scaled. The pseudo R2 (McFadden) is probably a better indicator. Currently in my system the public odds alone and the fundamental alone both have a pR2 of ~0.185 and the combined is model ~0.195.

I don't segregate by race type so they would be called static in this context.

[TrifectaMike]

I would ne extremely surprised if anyone's is... besides, your objective isn't to build a model that rivals the public in predictability

Thank you. I learn something new each day.

Mike (Dr Beav)

[TrifectaMike]

The coefficients aren't that meaningful without a frame of reference as they depend on how the two are scaled. The pseudo R2 (McFadden) is probably a better indicator. Currently in my system the public odds alone and the fundamental alone both have a pR2 of ~0.185 and the combined is model ~0.195.

I don't segregate by race type so they would be called static in this context.

Thank you Frank James.

Have you entertained the idea of using Bayesian Information Criteria instead of psuedo R2?

Mike (Dr Beav)

[Dark Target]

Thank you. I learn something new each day.

Mike (Dr Beav)
Without a sarcasm font I'm not sure if you are serious?

[Overlay]

These "impact value" lines have been around for many years now...and I have bought a few myself through the years...
When I mentioned statistics, I didn't intend to be referring solely to impact values.

[TrifectaMike]

Without a sarcasm font I'm not sure if you are serious?

Not scarcasm at all. You provided an answer for a question I didn't ask. However, the response was useful and noted as such.

Mike (Dr Beav)

[Dark Target]

I did give you an answer. You asked which line is dominant, I'm telling you the public's is and it always will be.

Let me ask you this, do you think the purpose of the Benter approach is to create a probability estimate that rivals the public in its predictabiity (as he says in his papers)?

[Dave Schwartz]

I did give you an answer. You asked which line is dominant, I'm telling you the public's is and it always will be.

If that were the case then nobody could ever beat the game.

[Dark Target]

Perhaps my response was a bit ambiguous, the public's model V's your own model (without the public probability estimates in it).

Do you think it is wise to try and build a model that "rivals" the public's probability estimates?

[jasperson]

Horse ML Betline Off odds result Bet?

1 12 16 15 out 0
2 5/2 2 7/5 3rd 0
3 3 6.8 7 out -2
4 15 24 40 out -2
5 5 7/2 8 out -2
6 4 4 2 2nd 0
7 10 21 7/2 1st 0







The "Betline" is from "Skip's" line he posted. Not criticizing him or his handicapping. But this is a good illustration of how useless and counterproductive odds lines are.

The 3 overlaid horses according to the betline and ML ran out. While the 3 underlaid horses according to the betline and ML all hit the board.

You are going to be on the back end of the homeless food line before you make a profit with odds lines.
I all depends on who's oddline you are using. Here is my oddline for that race
1 17/1
2 2/1
3 30/1
4 16/1
5 21/1
6 5/1
7 2/1

The only overlays were the 4 and 7 the 7 paid $9.70 - the $2 bet on the 4 a profit of $5.40 and I will take that all day long. Oh, by the way my voucher is showing a profit from the 1/1/12:jump:

[TrifectaMike]

Perhaps my response was a bit ambiguous, the public's model V's your own model (without the public probability estimates in it).

Do you think it is wise to try and build a model that "rivals" the public's probability estimates?

I think my question was a bit ambiguous. Benter-Wood uses an additional regression utilizing the public odds and the model odds as predictors to get a final vector. My question is;

If anyone is doing this step, do you know which vector has more weight in the regression?

Maybe now it's clear.

Mike (Dr Beav)

[Dark Target]

Yes, the public. Would be very interested to know if anyone has actually seen otherwise.

The point I'm trying to make is this, there is no point in comparing the 2, your primary objective with this type of approach is to design a model that adds as much predictability to the public model.

Why include in your own model the last time out speed ratings (for example) when clearly they are already a factor in the public estimates. When you combine yours with theirs its not going to add any predictability what so ever.

[TrifectaMike]

Yes, the public. Would be very interested to know if anyone has actually seen otherwise.

The point I'm trying to make is this, there is no point in comparing the 2, your primary objective with this type of approach is to design a model that adds as much predictability to the public model.

Why include in your own model the last time out speed ratings (for example) when clearly they are already a factor in the public estimates. When you combine yours with theirs its not going to add any predictability what so ever.

Which factors are NOT included in the publics estimates?

Mike (Dr Beav)

[TrifectaMike]

[QUOTE=Dark Target]Yes, the public. Would be very interested to know if anyone has actually seen otherwise. Thank you! That is an answer.

The point I'm trying to make is this, there is no point in comparing the 2, your primary objective with this type of approach is to design a model that adds as much predictability to the public model.

Why include in your own model the last time out speed ratings (for example) when clearly they are already a factor in the public estimates. When you combine yours with theirs its not going to add any predictability what so ever.[/QUOTE

]

[Dark Target]

Which factors are NOT included in the publics estimates?

Mike (Dr Beav)

A question which no one will be willing to answer.

[TrifectaMike]

A question which no one will be willing to answer.

So, what I need is a lightning bolt from Zeus to rearrange the publics vector!!

I must say that is an interesting POV.

Mike (Dr Beav)

[Dark Target]

Mike

I've said all that I'm willing to say on the subject. I must say though im having a hard time interpreting your questions though.

[TrifectaMike]

Mike

I've said all that I'm willing to say on the subject. I must say though im having a hard time interpreting your questions though.

You've helped a lot. What still confuses me is why "those" in the know repeatedly talk about using 50-70 regressors to arrive at a fundamental model and then use the public odds for smoothing.

It would seem to me that the 50-70 independent variables are already discounted by the public odds. What do you think?

Mike (Dr Beav)

[TrifectaMike]

I did give you an answer. You asked which line is dominant, I'm telling you the public's is and it always will be.

Let me ask you this, do you think the purpose of the Benter approach is to create a probability estimate that rivals the public in its predictabiity (as he says in his papers)?

Is it possible that Benter and company have mislead people into believing what he states in his paper?

That would be awful. I can't believe that to be true. That would be unprofessional.

Mike (Dr Beav)

Maybe he simply solved the problem backwards.

[Dark Target]

Far be it for me to say what Benter et al have done is not the best way of doing things, but yes, there is no point having variables in your model that are already accounted for in the publics (unless you interpret them different to the public).

Most of the highly predictive variables are already in the public model, that leaves you with producing variables the public doesn't consider (as hard as that is) and leaves your model far less predictive than theirs (on its own). But again, there is no point in producing a model as predictive as the publics if they contain the same variables, and when combined are no better than the public anyway.

That's enough from me.

[TrifectaMike]

Far be it for me to say what Benter et al have done is not the best way of doing things, but yes, there is no point having variables in your model that are already accounted for in the publics (unless you interpret them different to the public).

Most of the highly predictive variables are already in the public model, that leaves you with producing variables the public doesn't consider (as hard as that is) and leaves your model far less predictive than theirs (on its own). But again, there is no point in producing a model as predictive as the publics if they contain the same variables, and when combined are no better than the public anyway.

That's enough from me.
Just one more question... I'm beginning to get a better understanding.

In your opinion, is it fair to say that a little known, single metric can reshape the publics vector in a direction to make it more predictive?

Mike (Dr Beav)

I don't expect you to reveal any secrets.

[Dark Target]

Most definitely.

[TrifectaMike]

Most definitely.

Thank you very much. I have always suspected that a well directed, narrow piece of information would do the trick.... like pushing a ballon at the right spot to reshape it slightly, but not too much.

I dion't know you, but I do believe you. Are you one of those whales or part of a syndicate? You seem to know much about this subject. I'm certain you've been at this for a very long time.

Once again,thank you.

Mike (Dr Beav)

[Dark Target]

I have certainly been at it for a very long time, initially using Benters approach with great frustration.

Most certainly not a whale either and I work on my own (and not on US markets). Even once you have your probabilities, a similar challenge arises in getting many bets on at the last second (what I'm working on now)

[Light]

I all depends on who's oddline you are using. Here is my oddline for that race
1 17/1
2 2/1
3 30/1
4 16/1
5 21/1
6 5/1
7 2/1

The only overlays were the 4 and 7 the 7 paid $9.70 - the $2 bet on the 4 a profit of $5.40 and I will take that all day long. Oh, by the way my voucher is showing a profit from the 1/1/12:jump:

Sorry, doesn't count after the race. Show me one of your odds lines before the race like Skip did.

[TrifectaMike]

I have certainly been at it for a very long time, initially using Benters approach with great frustration.


I hear that often. So much wated time.

Let me run this by you. (This is fun)

If it is true that the publics probability is the dominate vector and more predictive than one's model, what if one can isolate "one" horse for "positive treatment" from the top (for example) 4 ranked horses by the public odds and assign to him all the probability mass scaled by a confidence factor. Then recompute the public's probability vector. In theory I would have a resultant vector which is more predictive.

Conversely, let's say I target a horse for "negative treatment". Once again, for example, a horse ranked in the top 4 by the public and assign to him a zero probability mass scaled by some confidence factor. Then recompute the public's probability. Again in theory, I would have a resultant vector which is more predicitive.

Of course, I have not defined what the "positive treatment" nor the "negative treatment". But I have defined a framework whereby I believe the public's probability vector is made more predictive. To what extent the public's probability vector is more predictive is also unknown.

Can we agree on this?

Mike (Dr Beav)

[Dark Target]

Trying to get my head around what your want to do here... think I understand.

Let me preface (or post in regards to my previous answers) this by saying, I am by no means the best person to ask on this forum complex mathematics questions, I've got to where I am buy understanding the problem I'm trying to solve, knowing how to make full use of the tools available to me, and solving the problem through actual simulations with the computer programs I write.

In saying that, I would be be cautious of re-modelling the public's odds in anyway, after all, they are the most predictive variable in my model. What I'd rather do, is work out a way to integrate what you are suggesting in the way of some other equation that can be integrated in the model...?

[Kevroc]

It is mathematically impossible to win money in the long term unless a bet gets some type of fair value. 40% winners at $4.80 won't cut it.


Sure it will. Anything over 5% rebate, churn the butter.

[erikeepper]

Btw, MarkGoldie (or something like that) wrote an excellent post about line-making and why you shouldnt concentrate a lot of time into it. And I think that guy knows a thing or two. Use the search-function. I will post it later if I can find it.

[Overlay]

Do you think it is wise to try and build a model that "rivals" the public's probability estimates?
The main problem with the public probability model (in my view) is not the factors that the public uses, or even (in the aggregate) the weights that the public assigns to those factors. But, on a race-to-race basis, the public can and does make significant misjudgments regarding the effect of individual factors on the outcome of the race, producing odds disparities that are capable of being systematically detected and exploited.

[jasperson]

Sorry, doesn't count after the race. Show me one of your odds lines before the race like Skip did.
When I lie about data it will be the first time. Attached is the oddline program for flx 5/17. It is in a txt format and can viewed best with dos editor the lines are so long that word will wrap them around. I did make one mistake of the 3 horse his odds were 32/1 instead of 30/1.

[bettheoverlay]

But, on a race-to-race basis, the public can and does make significant misjudgments regarding the effect of individual factors on the outcome of the race, producing odds disparities that are capable of being systematically detected and exploited.

Am I correct in assuming that with the odds line generated by your methods, the Rank 1 and/or Rank2 from this line will show a positive ROI over a long period of time when either of these horse's final odds are significant overlays, say, >= 6 and < 21 ?

[Light]

When I lie about data it will be the first time. Attached is the oddline program for flx 5/17. It is in a txt format and can viewed best with dos editor the lines are so long that word will wrap them around. I did make one mistake of the 3 horse his odds were 32/1 instead of 30/1.

Not accusing you of lying.I just asked you to show me an odds lines from before a race and you show me yesterday's. Oy Vey!

[Overlay]

Am I correct in assuming that with the odds line generated by your methods, the Rank 1 and/or Rank2 from this line will show a positive ROI over a long period of time when either of these horse's final odds are significant overlays, say, >= 6 and < 21 ?
I don't specifically require a certain degree of overlay (that is, a horse doesn't have to be at least 6-1 on the board (as in your example) to be offering value), but (because of the mixture and fundamental nature of the handicapping factors that I take into consideration) the lower the horse's fair odds, the higher would be the expected outcome as actual odds increase.

[jasperson]

Not accusing you of lying.I just asked you to show me an odds lines from before a race and you show me yesterday's. Oy Vey!
You don't understand how my oddsline program works. It uses bris past performance data in a single line coma delimited format and it doesn't change. If I run the program before the race after the race or next month the output will always be the same.

[Light]

You're reluctance and excuses is all I need to know about your odds lines.You obviously have no confidence in them before the race.

[teddy]

Personally the public makes its mistakes when it has little info to go on, falls back on stats like trainer win % or jockey trainer combos and then they hope for the best. Or times when they are faced with surface switches and go by pedigree because thats the only thing they can go by.

[raybo]

Sure glad I put Light on ignore.

Now if others would quit quoting his posts it would be great!

[BetHorses!]

Teddy,

If you are going to assign points (intutively, I assume and lots of testing), I would sum all the points awarded to each horse and use that sum to assign a percentage allocation to each horse.

Your example:

horse 3 12 pts, p=12/22 = .55 [ odds = 1/p -1] .83 to 1
horse 6 8 pts, p=8/22 = .36 [ odds = 1/p -1] 1.75 to 1
horse 9 2 pts, p=2/22 = .09 [ odds = 1/p -1] 10 to 1
sum = 22 points

Good luck,
Mike (Dr Beav)


Dr Beav,

I have two questions.

1) I assign points (not for solely first call) and lets say my 4th rank horse in an 11 horse field comes out to 16-1 without field size accounted for. In the formula you were so kind to share, 4th ranked should be roughly 9-1 in an 11 horse field. Should I just split the difference? What would be the math involved there, if there is any?

2) Do you feel the overlays in your top 4 produce a much greater edge than just playing the highest odds horse of same 4.

Thanks

[Dave Schwartz]

I must interject and make a comment here...

I have been a very vocal critic of the "true" effectiveness of odds lines for over two decades now...

It's not that I lack the sophistication to appreciate how valuable a good odds line can be for the player; I just didn't believe that it was possible to create an odds line which could be profitable against the toteboard...which, after all, is what we are all trying to do.

Well...I spent 3 hours on the phone with Dave Schwartz last monday...and now, I'm not so sure.

Dave has come up with a new way of creating an odds line...which is sure to be an eye-opening experience for the knowledgeable player. I know that it was for me...

I promised him that I wouldn't say much about this product yet...but I will say this:

I highly encourage you guys to take a look at this product when it comes out.

Sometimes...it's not necessary to reinvent the wheel; all we need to do is to look at the old game with a new set of eyes...

Thaskalos,

Thank you for the sterling recommendation.

I never noticed this post before or else I would have commented sooner.

I am going to get this short seminar on the calendar ASAP.

Regards,
Dave Schwartz

[ezpace]

I hear that often. So much wated time.

Let me run this by you. (This is fun)

If it is true that the publics probability is the dominate vector and more predictive than one's model, what if one can isolate "one" horse for "positive treatment" from the top (for example) 4 ranked horses by the public odds and assign to him all the probability mass scaled by a confidence factor. Then recompute the public's probability vector. In theory I would have a resultant vector which is more predictive.

Conversely, let's say I target a horse for "negative treatment". Once again, for example, a horse ranked in the top 4 by the public and assign to him a zero probability mass scaled by some confidence factor. Then recompute the public's probability. Again in theory, I would have a resultant vector which is more predicitive.

Of course, I have not defined what the "positive treatment" nor the "negative treatment". But I have defined a framework whereby I believe the public's probability vector is made more predictive. To what extent the public's probability vector is more predictive is also unknown.

Can we agree on this?

Mike (Dr Beav)

^^^^^^^^^^^^^^^^^^^^
THAT WORKS TriMike for years = proprietary, just find the *SIMPLE*
other 2 factors and presto. You will probably find better ones than i
have.

[theartfuldodger]

Dave I think King said Jan/Feb 2010

LOL. :rolleyes:

[theartfuldodger]

Goren,

50-60 variables would be more like it. (Or more.)

No, that would be more of the same: beyond most mortal men.

I said:


What I have in mind is relatively simple, but amazingly powerful. It is so simple that even a paper-and-pencil guy like Maddog could do it in about 20 seconds with a spreadsheet. And it has the power to make a "good handicapper" into a profitable one with just a simple tweak.

In fact, the method would probably be perfect for him.


Regards,
Dave Schwartz

Advertising???

[theartfuldodger]

You know, I was thinking the same thing the other day. I was reading in another thread, the one about winning $300k a year, where the poster you quoted wrote about having 1st hand knowledge, I believe is the way he phrased it, about winning $300k a year betting. And I thought to myself, here's a guy who's continually hawking his products, I mean, putting in a ton of constant effort -- including all the time he's spent rebutting detractors of his systems. Why would he do this if he were 'profitable'; significantly profitable? Guess he just wants to help other in better understanding the game.:rolleyes: Of course, having this dude get his hooks into you must not be a very good feeling overall. Can't wait for the next announcement of a big NEW PACE hit. Funny how those only show up when there's a good hit. Guess there's no action between hits.

AGREE 100%

[duncan04]

Advertising???

Wow you must really have something against Dave. :ThmbDown:

[theartfuldodger]

I do. And in about a month I will be sharing it with a portion of our world. It will just blow your mind.

Stay tuned.


Dave

Nice commercial. :sleeping:

[theartfuldodger]

All you need is a good correlation factor and then base your odds line off of that. I have one I use that has the winner 83% of the time within 3 horses. It's printed right on the past performances that I use. The top ranked horse gets 2/1. The next gets 3/1 and the last gets 3/1. This is the distribution of winners that I found in my sample.

Today I played BHP and AP using nothing but these odds and betting the overlays. I made $38 on $62 bet.

Oh yes I did. And I did nearly as well yesterday on 2 different tracks and on Wed not as well but still made significant money.

The winners it "picked" I would have never played on what I saw in the pps. They weren't horrible, but other horses looked better on class, form and speed. Yet, these horses won. Some handily.

Value betting and good spot plays are perfectly good ways of playing the game although in my opinion value will always reign supreme.

After all, that's all intelligent "gambling" is, placing wagers when you have a clear edge. Be it poker, horses or whatever.

Andy Beyer once wrote about a Harvard classmate who after obtaining his MD immediately put it away in a drawer and continued to play the ponies. His method? He used the win pool odds to spot place and show pool odds that were out of whack with the win pool and bet the overlays.

Any decent correlation factor can be used and I have a much better one than the public odds. Oh I still look for class and trainer moves but they don't manifest in every race. Value betting is a very easy and a very simple way to play the game.

And it isn't necessary to bet 4 horses in every race to have the winner. :D

[theartfuldodger]

Wow you must really have something against Dave. :ThmbDown:

He's blatantly advertising. Maybe he pays for the privilege. Whatever. I'm not into betting 4 horses a race and scraping by with a 2-3% ROI. Don't need the constant commercials.

[HUSKER55]

PLEASE CLARIFY,

you are betting 3 horses per race and you HAVE THE WINNER 7 races per race card. ( use 9 races as the base card )

thank you.

[raybo]

PLEASE CLARIFY,

you are betting 3 horses per race and you HAVE THE WINNER 7 races per race card. ( use 9 races as the base card )

thank you.

That's what he said, and to top it all off, he's betting value, not probability of winning. Amazing! Value betting will assuredly have a lower hit rate than betting the best horse, and if you're betting the best horse, you won't hit 83% with 3 contenders. If you can do that, you won't be wasting your time on a forum, you'll be managing your money, or out on your yacht, etc..

[Dave Schwartz]

All you need is a good correlation factor and then base your odds line off of that.

As soon as someone uses the phrase "All you need..." in reference to odds line making, I begin to laugh. To trivialize the odds-making process absolutely disqualifies you as a contender for "poster of the year."

The very first step in the upcoming commercial seminar, which I am clearly advertising here, is to explain why a "correlated factor" will not work.

The fact that you claim it does, tells me that you are not likely a qualified researcher.

And in case you need to hear it officially, I develop good handicapping ideas and sell them. It IS my business model. My ideas are good ones and the cost is reasonable.

In an age where horseplayers are clamoring for more good materials, it is difficult for me to understand how a person can fall back on that "anything good would never be sold," and then stop by the local bookstore to see if there are any new horse racing books.

[Dave Schwartz]

I have one I use that has the winner 83% of the time within 3 horses. It's printed right on the past performances that I use.

If you have such a factor, then you are a wealthy man.

Hard to believe that it is printed right there in the racing form and I have missed it all these years.

Please, oh, please, tell us what that magic factor that nobody but you has seen.

[HUSKER55]

when you guys are talking correlation factor are you talking about an item or a process?

[theartfuldodger]

PLEASE CLARIFY,

you are betting 3 horses per race and you HAVE THE WINNER 7 races per race card. ( use 9 races as the base card )

thank you.

POTENTIALLY betting 3 per race but only rarely does this happen. I am value betting OK? I am betting only overlays. That means 1 horse most of the time but occasionally 2 will qualify. Each bet is a separate bet that just happens to occur in the same race. I will explain...

I know how frequently the correlation factor wins for each of the top 3 "selections". Each bet therefore is an individual bet against the infinite string of all of these opportunities. Example: the top selection wins slightly more than 33% of the time. That's 2/1. I will accept 2/1 or better odds on those horses. Every time I see one come in at those odds I bet it. Makes no difference if I also bet another horse in the same race that also qualified as its particular odds overlay. That is a separate bet.

What I do not do is bet 3 or 4 horses per race trying to chase the winner or pass races because they are "poor betting opportunities". There is no such thing as a poor betting opportunity when you have an overlay.

Look at it this way. Why do you not bet the favorite every time to have a 33% win rate and cash a lot of tickets? Because you'll lose money right? Why will you lose money? Because if you always bet the favorite, then over the lifetime of such bets you will not win enough money to cover your losses. It's the same with this except here I know when I have a good bet and by making nothing but such bets money flows into your pocket.

[theartfuldodger]

That's what he said, and to top it all off, he's betting value, not probability of winning. Amazing! Value betting will assuredly have a lower hit rate than betting the best horse, and if you're betting the best horse, you won't hit 83% with 3 contenders. If you can do that, you won't be wasting your time on a forum, you'll be managing your money, or out on your yacht, etc..

Actually, that's not what I said at all. You might have reading comprehension issues.

Here, does this headline read well to you?

CURE DISLEXIA FOR FOUND

[thaskalos]

POTENTIALLY betting 3 per race but only rarely does this happen. I am value betting OK? I am betting only overlays. That means 1 horse most of the time but occasionally 2 will qualify. Each bet is a separate bet that just happens to occur in the same race. I will explain...

I know how frequently the correlation factor wins for each of the top 3 "selections". Each bet therefore is an individual bet against the infinite string of all of these opportunities. Example: the top selection wins slightly more than 33% of the time. That's 2/1. I will accept 2/1 or better odds on those horses. Every time I see one come in at those odds I bet it. Makes no difference if I also bet another horse in the same race that also qualified as its particular odds overlay. That is a separate bet.

What I do not do is bet 3 or 4 horses per race trying to chase the winner or pass races because they are "poor betting opportunities". There is no such thing as a poor betting opportunity when you have an overlay.

Look at it this way. Why do you not bet the favorite every time to have a 33% win rate and cash a lot of tickets? Because you'll lose money right? Why will you lose money? Because if you always bet the favorite, then over the lifetime of such bets you will not win enough money to cover your losses. It's the same with this except here I know when I have a good bet and by making nothing but such bets money flows into your pocket.

You say that you do not pass races because they are poor betting opportunities.

Do you find that your correlation factor functions equally well in all types of races?

[GameTheory]

Actually, that's not what I said at all. You might have reading comprehension issues.


I have one I use that has the winner 83% of the time within 3 horses.It sure sounds like you are saying you have identified a factor that contains the winner in the top 3 horses 83% of the time. And the response is that if you get the winner 83% of the time in the top 3 (which beats the top 3 favorites, which gets 66% or something), then you could just bet all 3 in each race without thinking and make money. But you say no, that's not what you are saying, even though you clearly said it.

So what does the 83% refer to?

[theartfuldodger]

All you need is a good correlation factor and then base your odds line off of that.

As soon as someone uses the phrase "All you need..." in reference to odds line making, I begin to laugh. To trivialize the odds-making process absolutely disqualifies you as a contender for "poster of the year."

The very first step in the upcoming commercial seminar, which I am clearly advertising here, is to explain why a "correlated factor" will not work.

The fact that you claim it does, tells me that you are not likely a qualified researcher.

And in case you need to hear it officially, I develop good handicapping ideas and sell them. It IS my business model. My ideas are good ones and the cost is reasonable.

In an age where horseplayers are clamoring for more good materials, it is difficult for me to understand how a person can fall back on that "anything good would never be sold," and then stop by the local bookstore to see if there are any new horse racing books.

So, to sum it all up, because YOU say you have been unsuccessful finding a correlation factor means that I couldn't have possibly found one and because I am saying this means I am not a "qualified researcher" (whatever that means).

What a lovely arrogant world you must live in Dave. Send me a postcard from time to time and let me know how you're doing. :rolleyes:

[theartfuldodger]

If you have such a factor, then you are a wealthy man.

Hard to believe that it is printed right there in the racing form and I have missed it all these years.

Please, oh, please, tell us what that magic factor that nobody but you has seen.

First he jumps down my throat on my very first post and rolls his eyes and now he's trolling for information. Sorry Dave, that ship has sailed. It's far off in the horizon.

[theartfuldodger]

when you guys are talking correlation factor are you talking about an item or a process?

Well, first of all WE aren't talking. Dave is pontificating and arrogantly dismissing new information solely because he himself didn't discover it.

A correlation factor can be one number or it could be a combination of factors. In my case it is one number. The power of that number is all in how you make use of it.

[theartfuldodger]

You say that you do not pass races because they are poor betting opportunities.

Do you find that your correlation factor functions equally well in all types of races?

It is more consistent in races for winners but hits higher payoffs in maidens so I play both. Maidens are usually won by the underlay. If too many first time starters or first time on turf I pass. That's gambling.

[theartfuldodger]

It sure sounds like you are saying you have identified a factor that contains the winner in the top 3 horses 83% of the time. And the response is that if you get the winner 83% of the time in the top 3 (which beats the top 3 favorites, which gets 66% or something), then you could just bet all 3 in each race without thinking and make money. But you say no, that's not what you are saying, even though you clearly said it.

So what does the 83% refer to?

Here is what HUSKER55 said "you are betting 3 horses per race and you HAVE THE WINNER 7 races per race card."

I never said that I bet all 3 horses per race. And I certainly don't win 7 out of 9. If I DID bet all 3 every race then I would hit 83% on average yes. But I don't. Why would I do that? That is throwing money away. Betting a horse going off at 4/5 with a 33% chance of winning? No!

Think of these 3 potential bets as occurring in separate betting pools. They each have their own probability of winning. I am essentially playing 3 games at the same time. OK?

[GameTheory]

If I DID bet all 3 every race then I would hit 83% on average yes. But I don't. Why would I do that? That is throwing money away. Betting a horse going off at 4/5 with a 33% chance of winning? No!
Yeah yeah, I get it about betting overlays, everyone does. What you are glossing over is that you are claiming a single factor being present in "your PPs" (which I assume are from a source all of us have heard of) gives you 83% winners in the top 3 horses. This is the claim that is getting the attention because it is outrageous -- such a factor would be performing (in terms of win%) FAR BETTER than the public odds, which doesn't even break 70%. Do you see why that is hard to swallow as it contradicts that experience of all handicappers and every study of factors ever done? If I knew of such a factor, I'd retire next week.

[Speed Figure]

Well, first of all WE aren't talking. Dave is pontificating and arrogantly dismissing new information solely because he himself didn't discover it.

A correlation factor can be one number or it could be a combination of factors. In my case it is one number. The power of that number is all in how you make use of it.
It must be Prime Power!

[theartfuldodger]

Yeah yeah, I get it about betting overlays, everyone does. What you are glossing over is that you are claiming a single factor being present in "your PPs" (which I assume are from a source all of us have heard of) gives you 83% winners in the top 3 horses. This is the claim that is getting the attention because it is outrageous -- such a factor would be performing (in terms of win%) FAR BETTER than the public odds, which doesn't even break 70%. Do you see why that is hard to swallow as it contradicts that experience of all handicappers and every study of factors ever done? If I knew of such a factor, I'd retire next week.

Well, first of all what YOU are glossing over is that YOU and others were claiming that I bet 3 horses per race and win 7 out of 9. So when you say everyone gets it, they clearly DIDN'T get it.

If you find what I said hard to swallow then fine, don't swallow. It's a moot point anyway as I have no intention of sharing and I am not selling anything.

And in case you haven't figured it out yet, anything for sale is by definition nearly worthless. If YOU had software that cranked out a 20%+ ROI would you be offering it to others for anything under say $10k? I sure as hell wouldn't.

[theartfuldodger]

It must be Prime Power!

It must be!

Or not. :(

[thaskalos]

Well, first of all what YOU are glossing over is that YOU and others were claiming that I bet 3 horses per race and win 7 out of 9. So when you say everyone gets it, they clearly DIDN'T get it.

If you find what I said hard to swallow then fine, don't swallow. It's a moot point anyway as I have no intention of sharing and I am not selling anything.

And in case you haven't figured it out yet, anything for sale is by definition nearly worthless. If YOU had software that cranked out a 20%+ ROI would you be offering it to others for anything under say $10k? I sure as hell wouldn't.

Why would you offer it for even $10K?

[GameTheory]

Well, first of all what YOU are glossing over is that YOU and others were claiming that I bet 3 horses per race and win 7 out of 9. So when you say everyone gets it, they clearly DIDN'T get it.

If you find what I said hard to swallow then fine, don't swallow. It's a moot point anyway as I have no intention of sharing and I am not selling anything.
Ok, you are living up to your name. Evasions.

I didn't claim anything about your betting. But others did point out that you could bet all 3 and win 7 of 9. This follows from your own math, and you confirmed it. And what we are all pointing out is that if you can get 83% from 3 horses each race, overlays are pretty much irrelevant because every 3-horse group is a group overlay in every race. So you could just bet those 3 in each race without worrying about the odds -- you could make all your bets in the morning after cashing your big pile of tickets from yesterday. You'd be wealthy in no time. Seriously, it really is a waste of your time to be odds-watching when you could just be playing golf as the winners roll in.

So, there are two choices:

A) You have made an error in your addition, because you do NOT have a single factor that is printed in any PPs that wins 83% of the time with 3 horses per race.

B) You are a liar and a troll.

[raybo]

Well, first of all what YOU are glossing over is that YOU and others were claiming that I bet 3 horses per race and win 7 out of 9. So when you say everyone gets it, they clearly DIDN'T get it.

If you find what I said hard to swallow then fine, don't swallow. It's a moot point anyway as I have no intention of sharing and I am not selling anything.

And in case you haven't figured it out yet, anything for sale is by definition nearly worthless. If YOU had software that cranked out a 20%+ ROI would you be offering it to others for anything under say $10k? I sure as hell wouldn't.

You said that the winner comes from your 3 selections 83% of the time. That is what you said, don't try to back-pedal now.

What we are saying is that if you have an 83% hit rate from your 3 selections then you could become very rich, very fast, because nobody anybody has ever heard of has accomplished that incredible feat. Basically, what we are saying is: You are either not saying what you mean or you are; "full of it".

[thaskalos]

All you need is a good correlation factor and then base your odds line off of that. I have one I use that has the winner 83% of the time within 3 horses. It's printed right on the past performances that I use. The top ranked horse gets 2/1. The next gets 3/1 and the last gets 3/1. This is the distribution of winners that I found in my sample.

Today I played BHP and AP using nothing but these odds and betting the overlays. I made $38 on $62 bet.

Oh yes I did. And I did nearly as well yesterday on 2 different tracks and on Wed not as well but still made significant money.

The winners it "picked" I would have never played on what I saw in the pps. They weren't horrible, but other horses looked better on class, form and speed. Yet, these horses won. Some handily.

Value betting and good spot plays are perfectly good ways of playing the game although in my opinion value will always reign supreme.

After all, that's all intelligent "gambling" is, placing wagers when you have a clear edge. Be it poker, horses or whatever.

Andy Beyer once wrote about a Harvard classmate who after obtaining his MD immediately put it away in a drawer and continued to play the ponies. His method? He used the win pool odds to spot place and show pool odds that were out of whack with the win pool and bet the overlays.

Any decent correlation factor can be used and I have a much better one than the public odds. Oh I still look for class and trainer moves but they don't manifest in every race. Value betting is a very easy and a very simple way to play the game.

And it isn't necessary to bet 4 horses in every race to have the winner. :D

And another thing, artfuldodger...

If we had a powerful correlation factor like this working for us...we might be persuaded to wager more than $62 on one day's action. :)

[raybo]

Did Light get back on the forum under another name? Sounds like awfully familiar crap.

Oh yeah, I will go out on the proverbial limb and state that Dave has forgotten more about gambling, and racing specifically, than you will ever know.

How do I know this, because there is no single number in the past performances that has ever produced the kind of hit rate you claim, even Prime Power, which is Brisnet's best hit rate single factor, and has been for years.

[TrifectaMike]

All you need is a good correlation factor and then base your odds line off of that. I have one I use that has the winner 83% of the time within 3 horses. It's printed right on the past performances that I use. The top ranked horse gets 2/1. The next gets 3/1 and the last gets 3/1. This is the distribution of winners that I found in my sample.


May I ask the following questions and hopefully you'll provide answers.

1. Whay is the average field size in races that you wager on.
2. The so-called correlation factor: Does it correlated to only 3 horse then it is minimized or discontintious? Isn't the rest of the field affected by the so called correlation factor. It seems rather unusual that over 80% of the correlation would fall upon exactly three horses exclusiively especially at those odds 2-1, 3-1, 3-1. There is nearly no difference between the three horses. They are basically undistinguishable from each other, but nearly totally distinquishable from the remaing field ( odd I'd say).
3. I know of no factor that would mathematically behave in this manner. Noy single or in combination.
4. I think you are making shit up...just my opinion.

Mike (Dr Beav)

[theartfuldodger]

Ok, you are living up to your name. Evasions.

I didn't claim anything about your betting. But others did point out that you could bet all 3 and win 7 of 9. This follows from your own math, and you confirmed it. And what we are all pointing out is that if you can get 83% from 3 horses each race, overlays are pretty much irrelevant because every 3-horse group is a group overlay in every race. So you could just bet those 3 in each race without worrying about the odds -- you could make all your bets in the morning after cashing your big pile of tickets from yesterday. You'd be wealthy in no time. Seriously, it really is a waste of your time to be odds-watching when you could just be playing golf as the winners roll in.

So, there are two choices:

A) You have made an error in your addition, because you do NOT have a single factor that is printed in any PPs that wins 83% of the time with 3 horses per race.

B) You are a liar and a troll.

Apparently you do not understand the distinction between win% and ROI. The statement you made "overlays are pretty much irrelevant because every 3-horse group is a group overlay in every race." is meaningless as far as making money goes.

You called me a liar. My conversation with YOU is concluded. Goodbye.

[theartfuldodger]

You said that the winner comes from your 3 selections 83% of the time. That is what you said, don't try to back-pedal now.

What we are saying is that if you have an 83% hit rate from your 3 selections then you could become very rich, very fast, because nobody anybody has ever heard of has accomplished that incredible feat. Basically, what we are saying is: You are either not saying what you mean or you are; "full of it".

I am not "back-pedaling". Somebody claimed I said I bet all 3 horses and win 7 out of 9 races and I simply said no, I do not. That is completely different from saying IF I bet all 3 I would have the winner 83% of the time. Do you not understand that?

Your assertion that WERE I to do that I would be rolling in cash is equally untrue. If the 2/1 or below wins the majority of the time then I'd probably lose money. Wouldn't I. You seem to be confusing odds with outcome. A PUBLIC 2/1 is not destined to win 33% of the time. MY 2/1 seems to hover around that percentage very closely. THAT'S why I bet it.

[theartfuldodger]

And another thing, artfuldodger...

If we had a powerful correlation factor like this working for us...we might be persuaded to wager more than $62 on one day's action. :)

Who said I bet only $62? Do you quote odds as 13,763 to 2,435?

[theartfuldodger]

Did Light get back on the forum under another name? Sounds like awfully familiar crap.

Oh yeah, I will go out on the proverbial limb and state that Dave has forgotten more about gambling, and racing specifically, than you will ever know.

How do I know this, because there is no single number in the past performances that has ever produced the kind of hit rate you claim, even Prime Power, which is Brisnet's best hit rate single factor, and has been for years.

Dave bets 4 horses per race for a 2-3% ROI. He don't know much.

[GameTheory]

Apparently you do not understand the distinction between win% and ROI.Priceless! I have a perverse affection for these kind of people. They will NEVER admit anything, no matter how stupid they look -- they just go round in circles. Anybody ever play that old computer psychologist game ELIZA? Reminds me of that.

You called me a liar. My conversation with YOU is concluded. Goodbye.I didn't call you a liar. You could have picked choice A: math error. But yeah, stupid or a liar, those are the only choices you've got at this point. Painted yourself into a corner. But are really are the artful dodger -- instead you try to paint me as a impudent name-caller so you can declare the conservation over and claim the high moral ground for yourself even though you are just trying to avoid admitting you a full of it.

[raybo]

Apparently you do not understand the distinction between win% and ROI. The statement you made "overlays are pretty much irrelevant because every 3-horse group is a group overlay in every race." is meaningless as far as making money goes.

You called me a liar. My conversation with YOU is concluded. Goodbye.

Every 3 horse group is a group overlay because of the preposterous hit rate you claim they have. Odds are irrelevant in such a case, because unless you have an average winning odd that is extremely low, then you will always make a profit, betting all 3 horses.

So, just bet them, regardless of odds, and collect your profit. Or, admit that you're lying about your 83% hit rate with just 3 horses, and probably everything else you've said.

[theartfuldodger]

May I ask the following questions and hopefully you'll provide answers.

1. Whay is the average field size in races that you wager on.
2. The so-called correlation factor: Does it correlated to only 3 horse then it is minimized or discontintious? Isn't the rest of the field affected by the so called correlation factor. It seems rather unusual that over 80% of the correlation would fall upon exactly three horses exclusiively especially at those odds 2-1, 3-1, 3-1. There is nearly no difference between the three horses. They are basically undistinguishable from each other, but nearly totally distinquishable from the remaing field ( odd I'd say).
3. I know of no factor that would mathematically behave in this manner. Noy single or in combination.
4. I think you are making shit up...just my opinion.

Mike (Dr Beav)

1. You are entitled to your opinion however if you "think I am just making shit up" then why do you bother to ask for more information?

2. You're drinking right now, aren't you. Have been since breakfast. Ask me how I know.

[raybo]

Dave bets 4 horses per race for a 2-3% ROI. He don't know much.

Well, since 98+% of all players are " long term losers", then that puts him in the top 2% or less of all players worldwide who make a long term profit. After reading your drivel, I am convinced you don't even come close to break-even, much less make 2-3% profit, unless, maybe, you're getting about a 25% or higher rebate on every dollar you bet. And I find it hard to even believe that you would be profitable, long term, with that kind of non-existent rebate.

[theartfuldodger]

Every 3 horse group is a group overlay because of the preposterous hit rate you claim they have. Odds are irrelevant in such a case, because unless you have an average winning odd that is extremely low, then you will always make a profit, betting all 3 horses.

So, just bet them, regardless of odds, and collect your profit. Or, admit that you're lying about your 83% hit rate with just 3 horses, and probably everything else you've said.

Here we go again. Next step, the censors will move in and close the thread.

With your minds closed like a steel trap how can you EVER hope to succeed?

This is the SECOND time I have literally handed you valuable information on a silver platter and yet you refuse to take it.

Hope to see YOU in the mutual pools early and often every day of the week pal.

[theartfuldodger]

Well, since 98+% of all players are " long term losers", then that puts him in the top 2% or less of all players worldwide who make a long term profit. After reading your drivel, I am convinced you don't even come close to break-even, much less make 2-3% profit, unless, maybe, you're getting about a 25% or higher rebate on every dollar you bet. And I find it hard to even believe that you would be profitable, long term, with that kind of non-existent rebate.

Oh, you're convinced. I see. That's nice.

[raybo]

1. You are entitled to your opinion however if you "think I am just making shit up" then why do you bother to ask for more information?

2. You're drinking right now, aren't you. Have been since breakfast. Ask me how I know.

He asked because he knows you can't answer it. You are now just trying to get out of a bunch of lies. Most liars just continue to lie to cover the other lies they've told.

Give it up, you're looking more pitiful with each word you type.

[GameTheory]

He asked because he knows you can't answer it. You are now just trying to get out of a bunch of lies. Most liars just continue to lie to cover the other lies they've told.

Give it up, you're looking more pitiful with each word you type.No -- it's valuable information and you're spitting on it! PEARLS BEFORE SWINE!!!!

[raybo]

Here we go again. Next step, the censors will move in and close the thread.

With your minds closed like a steel trap how can you EVER hope to succeed?

This is the SECOND time I have literally handed you valuable information on a silver platter and yet you refuse to take it.

Hope to see YOU in the mutual pools early and often every day of the week pal.

First of all, I am successful, and have proven it, without a doubt, on this very forum. That makes me one of the 2%, or less, long term profitable players worldwide. Since we already have at least 2 profitable players in this thread, your saying that you are also one of that extreme minority is more than highly unlikely. And then when you state the crap you state, it moves from unlikely to absolutely impossible.

[raybo]

No -- it's valuable information and you're spitting on it! PEARLS BEFORE SWINE!!!!

Oops, I must be so extremely ungrateful. I'll have to work on that. Thanks for the "headsup"!

[theartfuldodger]

May I ask the following questions and hopefully you'll provide answers.

1. Whay is the average field size in races that you wager on.
2. The so-called correlation factor: Does it correlated to only 3 horse then it is minimized or discontintious? Isn't the rest of the field affected by the so called correlation factor. It seems rather unusual that over 80% of the correlation would fall upon exactly three horses exclusiively especially at those odds 2-1, 3-1, 3-1. There is nearly no difference between the three horses. They are basically undistinguishable from each other, but nearly totally distinquishable from the remaing field ( odd I'd say).
3. I know of no factor that would mathematically behave in this manner. Noy single or in combination.
4. I think you are making shit up...just my opinion.

Mike (Dr Beav)

1. I play the major tracks. NY, LA, Churchill, GS, WO, Keeneland.]
2. It is clear that you don't understand. Being drunk isn't helping. ONE of THREE horses wins 83% of the time. The odds of EACH of the THREE winning are 33%, 25%, and 25%. That adds up to 83%. See how that works? :rolleyes: The #1 ranked horse will win 33% but at varying odds from 1/5 to whatever. Consequently, I only bet it when it is 2/1 or greater. Blindly betting all 3 might throw a profit but the point is to maximize the profit not accept pennies instead of dollars.
3. Irrelevant.
4. Don't care.

[theartfuldodger]

First of all, I am successful, and have proven it, without a doubt, on this very forum. That makes me one of the 2%, or less, long term profitable players worldwide. Since we already have at least 2 profitable players in this thread, your saying that you are also one of that extreme minority is more than highly unlikely. And then when you state the crap you state, it moves from unlikely to absolutely impossible.

Well, my IQ is 162 which puts me in the top 1% of the world population in intelligence. I have been playing the game for 20 years.

I think it highly unlikely that you are as intelligent as I am.

Now, go ahead and call me a liar again. I find it oh so amusing.

[theartfuldodger]

First of all, I am successful, and have proven it, without a doubt, on this very forum. That makes me one of the 2%, or less, long term profitable players worldwide. Since we already have at least 2 profitable players in this thread, your saying that you are also one of that extreme minority is more than highly unlikely. And then when you state the crap you state, it moves from unlikely to absolutely impossible.

Crap huh? How much do you pay for software and data files each month?

I don't even use a pencil. And my data is free. My costs are $0.

[GameTheory]

It is clear that you don't understand. Being drunk isn't helping. ONE of THREE horses wins 83% of the time. The odds of EACH of the THREE winning are 33%, 25%, and 25%. That adds up to 83%. See how that works? Yes, everyone knows how that works as we keep repeating. SO IF YOU BET ALL THREE, YOU'LL WIN 83% OF THE TIME, right? So if the average race winner pays at least $6 (it does), you'll be rich in no time even without looking at any odds, RIGHT?


"With these last words, the Dodger suffered himself to be led off by the collar; threatening, till he got into the yard, to make a parliamentary business of it; and then grinning in the officer's face, with great glee and self-approval."

-- Charles Dickens from Oliver Twist, the last scene of the Artful Dodger as he is dragged off to prison...

[duncan04]

Crap huh? How much do you pay for software and data files each month?

I don't even use a pencil. And my data is free. My costs are $0.

Guess you aren't going to reply to my pm to you. :rolleyes: