Highest Achievable Win %?

[OCF]

Just for kicks, what do you think would be the highest achievable win % over a one calendar year period, picking at least 100 horses over the course of that year, and with total disregard for ROI?

Please humor me, I already know that many of you think the ability to pick winners is secondary in importance to finding value.

I think it's an interesting question, and I'm also hoping the answers to this question can provide some insight as to how much of this game is "pure luck", i.e. seemingly inferior horses beating seemingly superior horses for unforeseeable or unknown reasons.

[davew]

65% should be fairly easy with 1 race every 3 days

[Dave Schwartz]

If you bet nothing but 1/9 and above, I would say that you'd be hitting 85% or higher.

(And losing money.)

[BELMONT 6-6-09]

Even the most selective spot plays over a 100 race sample would be hard pressed to top 45% playing with an eye for value. With a program of selecting winners of any odds maybe 50% disregarding playing 1/9 opportunities.

[CincyHorseplayer]

I think the win % can be equally absurd in both directions.As Dave said you could massage your ego by adding in 1-9 shots to jack up the win %.In the other direction some jack up their ego by lowering the win % to 10-15% at larger odds and really get off on that too.Personally I bet a horse less than 2-1 about 1 of 100 bets.My win % id middle of the road and I'm comfortable with that.I guess this isn't a fun answer!

[OCF]

It might help if you imagine a handicapping contest with the winner being the contestant with the highest win % rather than the highest ROI.

It may seem silly, but as I tried to get at in the original post there may be some valuable insight in considering the inverse, i.e. the lowest possible losing %.

[DeltaLover]

As far as predicting the outcome of the race the game is not 'pure luck'.
It can be seen as a chaotic event that can be handicapped sufficiently well, to the point of picking way more winners than a random process would do.

As a bettor though you are not getting paid by picking winners. The mutual nature of the game, is shifting its purpose from forecasting the winner, to judging the validity of the expectations of the crowd (who is very good as a 'pure' race predictor). Your job as a bettor is not to find the winner, but to detect the mistakes the crowd is making in its estimates and try to capitalize on them. Just knowing that this horse is way superior based in figures and class and will win 80% of time will not make any good to you as a bettor if this is already known to everyone at the track, same thing about ruling out a horse who has never finished better than 8th coming from a long layoff and ridden by a 1% jock if this horse happens to be 200-1.

I cannot count the times I have heard my fellow horse players, claiming that the point of the game is not to throw their tickets to the floor (especially after they hit some chalky combination like a $10.0 exacta). I am convinced that this is the worst misconception a horse bettor can hold. There is absolutely nothing wrong with throwing all the tickets of the week except one, to the floor, as long as the winning one is enough to cover the others and show some profitability.

[OCF]

As far as predicting the outcome of the race the game is not 'pure luck'.
It can be seen as a chaotic event that can be handicapped sufficiently well, to the point of picking way more winners than a random process would do.

As a bettor though you are not getting paid by picking winners. The mutual nature of the game, is shifting its purpose from forecasting the winner, to judging the validity of the expectations of the crowd (who is very good as a 'pure' race predictor). Your job as a bettor is not to find the winner, but to detect the mistakes the crowd is making in its estimates and try to capitalize on them. Just knowing that this horse is way superior based in figures and class and will win 80% of time will not make any good to you as a bettor if this is already known to everyone at the track, same thing about ruling out a horse who has never finished better than 8th coming from a long layoff and ridden by a 1% jock if this horse happens to be 200-1.

I cannot count the times I have heard my fellow horse players, claiming that the point of the game is not to throw their tickets to the floor (especially after they hit some chalky combination like a $10.0 exacta). I am convinced that this is the worst misconception a horse bettor can hold. There is absolutely nothing wrong with throwing all the tickets of the week except one, to the floor, as long as the winning one is enough to cover the others and show some profitability.

Well said, maybe a better question is: What is the highest estimate you could imagine putting on a horse's chances of winning a race? Assume a field of seven, which I believe is a little less than the average.

[Johnny V]

I would say from 45% to 50% with 50% being very hard to go beyond for the long haul. That is not limiting yourself to any particular odds level (although most likely you will be on heavy favorites many times) and betting a wide variety of number of races that fit your particular handicapping method. This IMO opinion would be very hard to achieve for any length of time.

[DeltaLover]

Well, this is mostly a theoretical question that a practical one. Assuming a horse that has its company so outclassed to the point to miss only if something exceptional happens, like bleeding or any other accident, you probably can go close to 90%. Again this is a completely imaginary case that clearly contradicts real world situations. Realistically speaking it will be very rare the situation where you could be convinced that a horse has more than 60% chance to win on a field of 7.

[TrifectaMike]

As far as predicting the outcome of the race the game is not 'pure luck'.
It can be seen as a chaotic event that can be handicapped sufficiently well, to the point of picking way more winners than a random process would do.

As a bettor though you are not getting paid by picking winners. The mutual nature of the game, is shifting its purpose from forecasting the winner, to judging the validity of the expectations of the crowd (who is very good as a 'pure' race predictor). Your job as a bettor is not to find the winner, but to detect the mistakes the crowd is making in its estimates and try to capitalize on them. Just knowing that this horse is way superior based in figures and class and will win 80% of time will not make any good to you as a bettor if this is already known to everyone at the track, same thing about ruling out a horse who has never finished better than 8th coming from a long layoff and ridden by a 1% jock if this horse happens to be 200-1.

I cannot count the times I have heard my fellow horse players, claiming that the point of the game is not to throw their tickets to the floor (especially after they hit some chalky combination like a $10.0 exacta). I am convinced that this is the worst misconception a horse bettor can hold. There is absolutely nothing wrong with throwing all the tickets of the week except one, to the floor, as long as the winning one is enough to cover the others and show some profitability.

Many years ago, there was a fellow that posted under the name, John Messina. He posted a comprehensive ( Win, Place, Show) oddsline (supposedly based on a SINGLE factor, which was never disclosed nor was his methodology). Actually the oddsline was well calibrated and quite often sharp.

The interest to me was that in some races (2-4 races). he would list a darkhorse(s)...usually a single. He claimed this horse was selected mathematically and automatically based on his oddsline. In a sense claiming he could tell by his oddsline that in that race his oddsline pointed out errors that the public would make. Well, to make the story short... the darkhorses were phenomenal!! Some on this board tracked his line and darkhorses, when he posted.

Some people have suggested to me that he was a lead guy in a betting syndicate having some fun. I have my own opinion of who he was.

Mike

I no longer have access to his line. One day he stopped providing it.

[Overlay]

Well said, maybe a better question is: What is the highest estimate you could imagine putting on a horse's chances of winning a race? Assume a field of seven, which I believe is a little less than the average.
I think that the lowest fair odds that I can recall figuring for a horse were 2-5 (which would represent just over a 70% probability of winning). (My method assigns a winning probability to each horse in a race, which has the effects of putting a floor on acceptable odds for the most likely winner, and resulting in lower odds for such horses in smaller fields, where it would be more likely that they might dominate all their opponents.)

[DeltaLover]

The interest to me was that in some races (2-4 races). he would list a darkhorse(s)...usually a single. He claimed this horse was selected mathematically and automatically based on his oddsline. In a sense claiming he could tell by his oddsline that in that race his oddsline pointed out errors that the public would make. Well, to make the story short... the darkhorses were phenomenal!! Some on this board tracked his line and darkhorses, when he posted.

You certainly know more Doc, but I really cannot see how this can be done, especially hours before the pools are open.

In my case, I am doing something conceptually similar, meaning identifying this (single) dark horse which is usually close to natural odds and base my whole betting on it, but I am able to do so minutes before the race when it it clear how the public perceives each starter and using more than one criteria (few but not a single one).

More than this, as we have discussed many times in the past, what is probably more important are not the exact odds but the ranking that is implied by them. For newcomers I suggest reading Let's play a stupid game (http://www.paceadvantage.com/forum/showthread.php?t=92151) one of the most though provocative threads on this board, started by Mike.

[Robert Goren]

I think if you are very,very selective, you could hit in the low to middle 50s and still show a profit(maybe a good one). That means bet one or two horses a week with some weeks betting none. maybe a month or two a year with none.

[TrifectaMike]

I think if you are very,very selective, you could hit in the low to middle 50s and still show a profit(maybe a good one). That means bet one or two horses a week with some weeks betting none. maybe a month or two a year with none.

Talk to a poster named, SLIGG. He uses Bris data and has a win% in the mid 70's and a nice profit, but a low betting frequency. I know the method and it works as he has described it.

Mike

[Stillriledup]

In Baseball, they call a .200 batting average the "mendoza line" maybe we can call winning 50% the "Kirkdoza line"?

[Robert Goren]

Talk to a poster named, SLIGG. He uses Bris data and has a win% in the mid 70's and a nice profit, but a low betting frequency. I know the method and it works as he has described it.

MikeThe question then becomes how many of us are willing to settle for a very low betting frequency even if it shows a nice profit? I suspect very few of us could scan through the PPs of several tracks each day and only bet one or two races a week.

[davew]

If you bet nothing but 1/9 and above, I would say that you'd be hitting 85% or higher.

(And losing money.)


How many 1/9's in a typical year?

[TrifectaMike]

The question then becomes how many of us are willing to settle for a very low betting frequency even if it shows a nice profit? I suspect very few of us could scan through the PPs of several tracks each day and only bet one or two races a week.

If one is presently not showing a profit, I believe it would be a no-brainer.

Mike

[Robert Goren]

If one is presently not showing a profit, I believe it would be a no-brainer.

MikeRemember most horse players are gamblers, not businessmen. If they were businessmen they would be in another business.

[Dave Schwartz]

On tonight's handicapping show, I highlighted CS-N2L races. In a 2yr period, there were 5,500 such races. There were 15 horses that went off at 1/9 or below, with 14 winners.

So, one could probably guess a similar percentage: about 1 in 366 races.

http://www.horsestreet.com/BBSImages/CSN2L-LowOdds.PNG

[maddog42]

If you bet nothing but 1/9 and above, I would say that you'd be hitting 85% or higher.

(And losing money.)

My records show this to be pretty close. Barry Meadow came up with similar.
The fly in the ointment is many 1/9 shots don't become 1/9 until after the bell.

[flatstats]

Here's a method with 100% strike rate. And not surprisingly it has returned a profit in the past 6 years:

Lay every 200/1+ (SP) horse on a Betting Exchange.

Wins: 686
Runs: 686
Win% 100%
Profit: £651.70
ROI: 0.1% (of total stakes risked)

2008 160 wins from 160 runs
2009 170 wins from 170 runs
.
.
2013 33 wins from 33 runs

Of course this is not a sensible strategy at all because one day a 200/1+ horse is going to win and as 200/1 SP roughly translates to 999/1 on Exchange odds you are going to have to pay out on 999/1 when that horse does win. It will then take you another 3 years to break even again.

[OCF]

Here's a method with 100% strike rate. And not surprisingly it has returned a profit in the past 6 years:

Lay every 200/1+ (SP) horse on a Betting Exchange.

Wins: 686
Runs: 686
Win% 100%
Profit: £651.70
ROI: 0.1% (of total stakes risked)

2008 160 wins from 160 runs
2009 170 wins from 170 runs
.
.
2013 33 wins from 33 runs

Of course this is not a sensible strategy at all because one day a 200/1+ horse is going to win and as 200/1 SP roughly translates to 999/1 on Exchange odds you are going to have to pay out on 999/1 when that horse does win. It will then take you another 3 years to break even again.

Interesting. The "betting against" thing is a whole different world, to say the least. The information in your post is a good demonstration of that.

I don't know if I could get used to it to the point where I could move easily back and forth between it and what I've always done here in the U.S.. It would be nice to at least have the option on a widespread basis though.

[Dave Schwartz]

Interesting. The "betting against" thing is a whole different world, to say the least. The information in your post is a good demonstration of that.

This is the key to my own profits!

It is all about seeing who you can bet against (that is, the percentage of pool that supports them), eliminating the ones that have a small chance of winning (i.e. like the 50/1 and up horses I talked in last night's show) and then playing the rest.

[OCF]

This is the key to my own profits!

It is all about seeing who you can bet against (that is, the percentage of pool that supports them), eliminating the ones that have a small chance of winning (i.e. like the 50/1 and up horses I talked in last night's show) and then playing the rest.

Your point is well-taken. I'm glad you posted because in relation to this thread I've been meaning to ask you something. I seem to remember a thread, maybe a year ago or so, where you had some back-and-forth regarding your willingness to bet on a specific horse that was the betting favorite in a minus pool. Or maybe not a minus pool but not far from it.

Do I have that right? My apologies in advance if I do not.

[Track Collector]

Just for kicks, what do you think would be the highest achievable win % over a one calendar year period, picking at least 100 horses over the course of that year, and with total disregard for ROI?

Please humor me, I already know that many of you think the ability to pick winners is secondary in importance to finding value.

I think it's an interesting question, and I'm also hoping the answers to this question can provide some insight as to how much of this game is "pure luck", i.e. seemingly inferior horses beating seemingly superior horses for unforeseeable or unknown reasons.

Using data from 2012 from virtually all North American thoroughbred races, and playing around with the handicapping tools available to me, I was able to find the following:

Total Plays = 101
Wins = 91
Win% = 90.1
ROI = 0.9847 (for reference)

The above is "fitting to the data", and 100% not possible to achieve in the real world because it involves the preciseness of the odds to $1 value, and one can not be 100% sure of that value prior to all the win pool monies being collected. I could only hazard a guess as to the real world achievable Win%, thinking it might be around 87%

In any event, a FUN theoretical question!


Chris

[pondman]

Just for kicks, what do you think would be the highest achievable win % over a one calendar year period, picking at least 100 horses over the course of that year, and with total disregard for ROI?


With 100 horses a good player should be able to get close to 50% with horses above +5-1.

I've run methods on Russell Baze and you could get nearly 70% wins on him in a year, but you won't make any money.

[OCF]

Using data from 2012 from virtually all North American thoroughbred races, and playing around with the handicapping tools available to me, I was able to find the following:

Total Plays = 101
Wins = 91
Win% = 90.1
ROI = 0.9847 (for reference)

The above is "fitting to the data", and 100% not possible to achieve in the real world because it involves the preciseness of the odds to $1 value, and one can not be 100% sure of that value prior to all the win pool monies being collected. I could only hazard a guess as to the real world achievable Win%, thinking it might be around 87%

In any event, a FUN theoretical question!


Chris

Thanks for playing along! I don't think I would have expected the 98+% ROI, would have expected something smaller.

How many filters did you apply to get to the 101 with a 90.1% win %?

[OCF]

With 100 horses a good player should be able to get close to 50% with horses above +5-1.

I don't doubt you and my hat is off to you if you've done it yourself. It's an interesting twist - when I think high strike rate I immediately think low odds. But if one is going to be that selective, i.e. bet that seldom, it's doesn't seem that far-fetched to think one could be picky about the odds also.

[pondman]

I don't doubt you and my hat is off to you if you've done it yourself. It's an interesting twist - when I think high strike rate I immediately think low odds. But if one is going to be that selective, i.e. bet that seldom, it's doesn't seem that far-fetched to think one could be picky about the odds also.

My plays don't exist under 1 umbrella. I'll go 40 horses on one method in S. Cal. I'll go 25 horses at CD. I'll play 75 horses in N. Cal. And then I'll keep my eyes out for crafty people looking to steal a race at places, such as Turf paradise and mountaineer. It's that kind of thing. I don't have 100 horses a year with 1 method at 1 track.

[Track Collector]

Thanks for playing along! I don't think I would have expected the 98+% ROI, would have expected something smaller.

How many filters did you apply to get to the 101 with a 90.1% win %?

I did not save the query, but I think it was at least 3 filters, and perhaps a 4th. One might be able to keep searching and searching (but still requiring at least 100 plays per year) to get an even higher number, but the additional filter(s) would yield such small benefits that the 90.1% might go to something like 90.2% or 90.3%.

The major starting criteria was horses that went off at less than $0.20 to the dollar. Then it was refined by a couple of filters (in the JCapper handicapping software) to get to the 90.1%. The "refining" filters only improved things somewhere between 1.0% and 1.5%.

Not easily translated to to real world wagering activity when your main handicapping factor can not be known with 100% accuracy until no more wagers are allowed.


Chris

[OCF]

I did not save the query, but I think it was at least 3 filters, and perhaps a 4th. One might be able to keep searching and searching (but still requiring at least 100 plays per year) to get an even higher number, but the additional filter(s) would yield such small benefits that the 90.1% might go to something like 90.2% or 90.3%.

The major starting criteria was horses that went off at less than $0.20 to the dollar. Then it was refined by a couple of filters (in the JCapper handicapping software) to get to the 90.1%. The "refining" filters only improved things somewhere between 1.0% and 1.5%.

Not easily translated to to real world wagering activity when your main handicapping factor can not be known with 100% accuracy until no more wagers are allowed.


Chris

I gotta ask but I'll definitely understand if you don't want to answer - what were the filters after the first filter for odds?

[DeltaLover]

I gotta ask but I'll definitely understand if you don't want to answer - what were the filters after the first filter for odds?


I do not think it really matters.

These additional filters can be anything. The fact that they worked last year does not mean they will be repeated for the next one.

This approach is clearly backbiting several a postreriori rules with random behavior thus there is no betting value associated with it.

[OCF]

I do not think it really matters.

These additional filters can be anything. The fact that they worked last year does not mean they will be repeated for the next one.

This approach is clearly backbiting several a postreriori rules with random behavior thus there is no betting value associated with it.

Duly noted and I can't disagree with anything you said.

I'm still interested if TC wishes to share.

[OCF]

My plays don't exist under 1 umbrella. I'll go 40 horses on one method in S. Cal. I'll go 25 horses at CD. I'll play 75 horses in N. Cal. And then I'll keep my eyes out for crafty people looking to steal a race at places, such as Turf paradise and mountaineer. It's that kind of thing. I don't have 100 horses a year with 1 method at 1 track.

How big does the purse need to be at TP or Mnr to make it worthwhile to ship there from So CA or one of the high end East Coast tracks?

[Topcat]

I merely looked at horses listed on the morning line under 4/5 and found 100 bets and you hit 67 or 67% and still manage to lose money. (so dont' do it) These are rare, took almost a year to get 100 bets.

[castaway01]

I know this is not exactly answering the original question, but to me there's no point in worrying about winning at 70% if all the horses pay $2.40 and you're losing money anyway. I have done 40% at $7 a horse for stretches and think someone smarter than me (not a large accomplishment) could do so at longer stretches and be very successful. It would take someone with a solid plan who could adjust in a split second when their method went sour yet also somehow did not lose faith when the method took a temporary setback yet was destined to rebound. Not sure who has those skills, but if you do, horse race wagering is for you.

[Track Collector]

I gotta ask but I'll definitely understand if you don't want to answer - what were the filters after the first filter for odds?

Hi OCF,

No secrets here.

The first filter (wagering odds to $1) is information that is not secret and available to anyone. The other filters were based on factors within the JCapper software which are proprietary and for which the exact formula/definitions are not made known to the user. Thus, I could not share them with you even if I wanted to.


Chris

[OCF]

Hi OCF,

No secrets here.

The first filter (wagering odds to $1) is information that is not secret and available to anyone. The other filters were based on factors within the JCapper software which are proprietary and for which the exact formula/definitions are not made known to the user. Thus, I could not share them with you even if I wanted to.


Chris

Thanks again for playing along, it was interesting info nonetheless. :)