FINAL THREAD - PART III VALUE BETTING AT THE RACETRACK

[ValueBetting]

To Whom It May Concern::ThmbUp:

For those of you who have made "reasonable" queries or remarks about my
book:

I AGREE with you that:

My book does not present statistics on overlays to The Valuline odds line, because it was not intended to do so. I had assumed that everyone agreed that overlays are the profitable bets (value bets), and that a demonstration of the near perfection of The Valuline odds line (Table III) was sufficient by itself.

My book does not explain "how to" construct an odds line. This was done
years ago by Meadow and by Cramer, whom I cite in my recommended
reading (Selected Bibliography).

My book does not present win percentages and ROIs by betting strategy,
race type, or race track. (It does say (Table IV) that this is a work-in-
progress, and I have set for myself a completion date of May of next year (the Kentucky Derby).

I DISAGREE with those of you who have objected to the following:

Whereas I may have been too zealous in the use of the word "discovery,"
I KNOW that my computer program is a contribution to the knowledge of
horse race handicapping and betting. Why?

Because no other program matches the verified capability of my program
in providing the mathematical chances of winning (win probability)of the most likely win contenders in a race.

Therefore, this alone, is an original contribution to horse race scholarship.
(To question this assertion REQUIRES YOU naming another program that
matches The Valuline odds line in accuracy (Table III). ___________

__________________________

The verification of The Valuline prompted the writing of my book. But,
I set out also to keep young horseplayers on the right path, which is comprehensive handicapping (class,condition,pace,speed and connections).
I set out to explain that my personal betting experiences demonstrated that
angle-betting at worst will make them a loser, and at best will manipulate
their handicapping to their disadvantage. The proliferation of angle-sellers
in our "age of specialization" made the tellling of this, I thought, worthwhile.
(Perhaps some of you are either angle-sellers, or friends or relatives of
angle-sellers.)

It is unfortunate that the verification of my program's odds line is associated with the very program I am selling. This has made a number of you claim that my book is a commercial. But, it's not! It is meant to
announce and to educate.

If you don't believe me, I can't help it.

You should, however, note this: In my book, I encourage horseplayers
to handicap comprehensively, and not be deterred. I present a "Selected"
Bibliography of good handicapping and betting books, including Meadow's
and Cramer's how to calculate an odds line. I encourage them to "go for
it." (Sure I offer a substitute, but it is at a price. They need not use it
to make profit. In fact, I state that boldly in my book. The HEADING to
Chapter Five reads "I made a profit when I did Comprehensive Handicapping."
This was clearly comprehensive handicapping by hand, NOT by computer
program!

THEREFORE, I am disappointed at the comments by some of you, NOT
because of the effect it may (or may not) have on sales of my book. I
say again, for the final time, I wrote the book not for profit, but to share
and to educate.

I would be encouraged to read a comment from any of you who have re-considered your opinion after reading my book, or after what I have said here today.

Sincerely,
Dave Johnson

[melman]

What you have said here today is the same thing you have said in earlier posts. You just changed the pitch in the beginning then right back to the "I have done a great thing" mantra.

[the_fat_man]

Any relation to Davey Johnson, 2nd baseman and manager of the Mets?

:lol:

[sjk]

ValueBetting,

Most of the time the true pioneers of reseach do not get fully recognized for many years after their works are known so you should expect this sort of treatment.

Most of us are bettors so we do not care if your betting line shows the same winning probabilities as are borne out by the race results but rather whether the betting line can bring in profits.

I suggest that you bet real money using your line because the profits you make will seem much more important than the posts you read here.

The usual standard for the accuracy of an odds line is how profitable the bets are that the odds line would suggest. I proposed in an earlier thread that you create a table so that we could see how your line stacks up in this regard. That would be interesting information.

Otherwise bet away and let your growing bankroll (or history if you are really patient) be the judge of your contribution.

[ValueBetting]

To SJK:

Yours is a "fair enough" reply.

And, that's exactly what I am intending to provide with my current testing.

Thank you for your explanation.:)

Regards,
Dave

[ValueBetting]

To Melman:

Why you don't appreciate what I've done, I'd like to know. Isn't there
some desirability in knowing a computer can generate an almost perfect
betting line? (Perfect in the sense that it somehow matches win probabilities
to race results). ___________

It's certainly at least "one" point of beginning to
profitable betting, don't you think? ______________

But, I can't change my "pitch" any further, so I'll simply stop trying.

Dave:confused:

[GameTheory]

Isn't there
some desirability in knowing a computer can generate an almost perfect
betting line? (Perfect in the sense that it somehow matches win probabilities
to race results).
NO!!!!

It's no big deal unless the overlays make money!!!!

DUH!

[Steve 'StatMan']

NO!!!!

It's no big deal unless the overlays make money!!!!

DUH!

Exactly! If the odds line overall is more predictive of the true odds than a line I would make on my own, that may nice, congratulations - BUT if the 'overlays' on both of our lines are not profitable, then it doesn't matter, I shouldn't be betting with either one! Anyone can make a line where the 4/1 shots that one bets at 6/1 and above are unprofitable, and the ones going off at 8/5 balance out the overall win % to 20% but are also unprofitable. A method that identifies fair 4/1 (20% win) horses that are profitable when they go off at 6/1 and higher, now that would really be something!

I should admit that I am starting to sense some use for a line that can better predict (at least generally) more accurate odds for the horses I don't like. I could then adjust lines to factor in my own handicapping that isn't in the method, or lower my expectations due to possibly better 'non-contenders' than I perceive.

Yes, software would be more helpful, but a manual method might help as well, for making these determinations. Would need to know more about what Dr. Johnson has currently produced, and what might be to come, to see if anything would fit my needs in these regards.

Of course, with any method or well distributed software, if a significantly large number of bettors and/or heavy wagerers utilize the same method (in this case, the same fair odds) and they all bet very large amounts of money on the same 'overlays', in theory they would drive down the final odds and prices, possibly affecting the methods level of profitablity. There are plenty of bettors already frustrated now when their 'overlays' drop from 4/1 before the gate opens, down to 3/1 when the horses round the turn, and all wagers from the final 20-30 seconds of wagering finally hit the mutuel pools.

[chickenhead]

off topic a bit --

as most any odds line will get "pulled around" by the tote in terms of actual win %, suggesting blending...I was just curious if anyone has compared two good non-tote lines (like say your homegrown, and/or HTR, HSH, Jcapper etc) and observed the same behavior?

I'm just wondering what the result would be of blending multiple commercial lines together -- if they are all of more or less equal quality (but somewhat different as to emphasis) should the result theoretically be better than any of the constituent pieces alone?

[ValueBetting]

To Game Theory:

O.K.

SJK, Melman, and you are saying basically the same thing.

You don't really care about advances in scholarship, and you won't
accept what seems obvious (overlays=profit underlays=loss), you
want to see the specific stats.

Well, as I have said, that's next. I am working on it and have no
doubt the + ROIs will be there.

I still believe, as a past academic, that scholarship advances are beneficial
per se, which is what SJK kindly intimated.

Dave

[ValueBetting]

;) To Stat Man:

Yes, we are all communicating better now. See my combined replies to Melman, Game Theory, and SJK.

You have made a good new point. The Valuline helps point out some
"horses you have tossed."j In fact, I know perhaps the best comprehensive
handicapper (by hand) in Southern California who uses The Valuline to confirm his selections (and their value) but, more importantly perhaps, to re-assess horses he may have missed or tossed.

Dave

[headhawg]

SJK, Melman, and you are saying basically the same thing.

You don't really care about advances in scholarship, and you won't
accept what seems obvious (overlays=profit underlays=loss), you
want to see the specific stats.You can write all of the scholarly material you want, and I love to read new material myself, but it has to be useful. I have seen melman and sjk in the War Room -- live -- picking very nice priced winners. They don't need a book to tell them about overlays. Trust me they know a lot more about real world overlays than you give them credit for.

Perhaps they should be the ones writing the books.

[ValueBetting]

To Chicken Head:

GREAT QUESTION.

Hope you get some good answers. You have many serious horseplaying
viewers.

Before developing my computer program (1996) I searched far and wide.
As I continued to "tweek it" (my odds line) I continued my search.

To date, I know of no program which has as good as odds line as mine.
I would love to have someone tell me an alternative to test: ___________

To answer your specific question, I think the answer must be "NO". Your
consensus would be a hodge-podge or dilution of the oridinarily small
margin that a "fair odds" line provides.

Sure, you want to bet the "monster" who goes off at 10/1 when you
calculate the horse at 5/1. But, how often does this happen?

What about the horse you calculate at 2/1 that goes off at 5/2? Here
are the many potential bets you will lose by the consensus method above.

Dave

[ValueBetting]

To Headhaug:

No problem.

Perhaps they should write books.

I am getting the impression that there are some highly qualified handicappers
on this website, and that's great. Sure, they can pick winning longshots
using the Form without a computer.

You should remember, however, that my book is also a guide to young
horseplayers to keep them on the right path to comprehensive handicapping
however they do it.

Dave:)

[GameTheory]

You don't really care about advances in scholarshipI do care about advances in scholarship, but it is NOT an advance! Making a so-called "perfect" oddline AIN'T HARD if you don't care about profitability. In fact, in order to achieve profitability you usually have to make it "less perfect" -- you don't make that your goal, of course, but "less perfectness" usually occurs as a side-effect of making it less correlated with the odds but this less perfect line will nevertheless be MORE PROFITABLE.

and you won't accept what seems obvious (overlays=profit underlays=loss), you want to see the specific stats.You lost me here. To me an overlay is a horse with higher tote odds than the fair odds given by a personal line, or your line, or whatever line you're using. So overlays are subjective -- what may be an overlay according to your valuline may not be an overlay according to some other line. But GIVEN your valuline as our fair-odds generating/assigning process, we can objectively say this group of horses are overlays (tote odds > valuline odds), and that group of horses are underlays (tote odds < valuline odds). So, we've got overlays and underlays without ever looking at profit or loss. The overlays don't automatically make money! In fact, most oddlines, and now listen closely -- EVEN PERFECT ONES (by your definition of "perfect") -- LOSE MONEY. Usually they lose it at a greater rate than the take! On most oddlines, the OVERLAYS ARE BAD BETS. The underlays actually do better (whilst still being unprofitable).

How can this be? Because most lines are too correlated with the toteboard, and if we take a subset of horses that have all been assigned THE SAME FAIR ODDS, as the tote odds go up (i.e. the overlay gets bigger), the win percentage goes down. And it usually goes down at a rate faster than the "value" is going up due to the higher odds. Result? LOSING MONEY BIG-TIME. (If you want, you can assume we are only considering the top 2 or 3 horses on the line -- the contenders if you will -- or else you'll really lose money fast.)

So until you can show that your line makes money betting on overlays, or on "contenders only" that are overlays, the assumption has to be that it LOSES MONEY BIG-TIME. That's right -- the NORMAL, USUAL, AND EXPECTED BEHAVIOR of "perfect" lines like yours is to LOSE MONEY BIG-TIME. And since you don't seem to have any idea of TRUE PERFORMANCE of your own wonderful creation, my guess is that this is exactly how it behaves, and you are in for a rude awakening, JUST LIKE THE HUNDREDS (POSSIBLY THOUSANDS) OF OTHERS BEFORE YOU that have made a seemingly perfect oddsline and seen it bite the dust in actual practice all in glorious disregard of its AVERAGE correspondence with reality when comparing win percentage and expected win percentage according to that line.

Are you getting this? This is an OLD and WELL-KNOWN problem. Barry Meadow apparently brought it up to you when you sent him the book and you ignored him. You keep saying, "That's the second part, that's the second part I'm working on now." But the thing is, THE FIRST PART DOESN'T MEAN SQUAT if the performance is not borne out in the second part. The second part should be the first part because the second part is the important one! You have advanced scholarship by exactly ZERO by making a so-called perfect oddsline!

Here's another shocker! Most oddslines THAT DO MAKE MONEY are no where near perfect (by your definition, again). So you've chosen the wrong goal, gone down the wrong path, and now you want others to follow you when you have not shown that any reward is to be found at the end of the journey!

!!!!!!

[xtb]

You don't really care about advances in scholarship, and you won't
accept what seems obvious (overlays=profit underlays=loss), you
want to see the specific stats.

Well, as I have said, that's next. I am working on it and have no
doubt the + ROIs will be there.



Pons and Fleischmann had no doubts that they had discovered cold fusion in a glass jar.

[banacek]

I had assumed that everyone agreed that overlays are the profitable bets (value bets), and that a demonstration of the near perfection of The Valuline odds line (Table III) was sufficient by itself.


It isn't. That's the problem we have been trying to get you to address. The fact that you don't understand this elementary mathematical concept is what has turned me off buying your book.

[ryesteve]

In fact, most oddlines, and now listen closely -- EVEN PERFECT ONES (by your definition of "perfect") -- LOSE MONEY. Usually they lose it at a greater rate than the take!
Excellent summation!

[GameTheory]

Actually, I should have said ESPECIALLY perfect ones, because they are usually more correlated to the tote than ever. So such "perfection" is actually usually a NEGATIVE indicator if the line-maker put his effort into achieving such perfection instead of putting it into achieving profit...

[Steve 'StatMan']

There is the 'Perceived Overlay' when the odds is greater than either fair odds line we're using, or the price we feel we need, regarless of the method of handicapping, to feel we're getting a good bet, and perceived it as an 'Overlay'.

Then there the Real Overlays - did those horses really run better than their odds, and make a profit. Did they underperform, and end up only being break even or a loss - in which case they were not Real Overlays, but just misperceptions. Or is the sample too small, or the way of evaluation to complex to truly know, at least statistically.

One might know better by having wagered on them enough, and turning enough of a profit to say, yes, I am indeed usually betting on overlays.

One could be trading dollars, or we're not winning, and doubting that our Perceived Overlays are Real Overlays after all.

Or one could be getting their wallets cleaned out, desperately wondering what is going wrongs, because, after all, "I'm betting Overlays".

So, hopefully you're at least betting succesfully to know you're finding enough 'Real Overlays', since you don't currently have 'statistical proof' that betting on your line's 'Perceived Overlays' are indeed Real Overlays and turning a profit - outperforming the actual odds.

I was hoping you actually could prove it, and that they were Real Overlays, outperforming the odds. And hopefully for you, you will someday, or are at least succesfully betting.

For your own method, you probably had to create this first piece, and others have their own ways of determining what price they will or will not take on a horse, but for all of us, the 'Real Work', the rare, but real and important accomplishment, comes in actually producing the profit when those horses going off at over the required odds actually do win often enough to earn a profit, and doing it with real dollars, so that it includes the intangibles like the 'choke factor', etc.

I do wish you good luck in your efforts to get there someday. That's what most of us on the board strive to do, and chat with each other so we can get there and/or stay there.

[aaron]

I have a friend who monitored his odds line for thousands of races.He basically came up with what you thought was a great break through and wrote a book about.
He told me his 4-1 shots won at 20%,but when they went off at 10-1 they won at 10%,so basically even though his lines were accurate it was almost impossiable to make money just using his line without making adjustments based on the actual odds.An example he gave me is that,he has a horse 2-1 on his line,but it is going off at 10-1.There is another horse in the race he has at 6-1 and the horse is going off at even money thus creating overlays on other horses.I hope this makes sense,because there are many instances where one horse could be overrated and another underrated.

[GameTheory]

I have a friend who monitored his odds line for thousands of races.He basically came up with what you thought was a great break through and wrote a book about.
He told me his 4-1 shots won at 20%,but when they went off at 10-1 they won at 10%,so basically even though his lines were accurate it was almost impossiable to make money just using his line without making adjustments based on the actual odds.An example he gave me is that,he has a horse 2-1 on his line,but it is going off at 10-1.There is another horse in the race he has at 6-1 and the horse is going off at even money thus creating overlays on other horses.I hope this makes sense,because there are many instances where one horse could be overrated and another underrated.But if his line is accurate, then the overlays just gotta make money! They just gotta! Isn't it obvious? :lol:

[DJofSD]

He told me his 4-1 shots won at 20%,but when they went off at 10-1 they won at 10%,so basically even though his lines were accurate it was almost impossiable to make money just using his line without making adjustments based on the actual odds.An example he gave me is that,he has a horse 2-1 on his line,but it is going off at 10-1.There is another horse in the race he has at 6-1 and the horse is going off at even money thus creating overlays on other horses.

Isn't that when using the Kelly criteria is usable and, in theory, allowing you to optimize your profit?

[ryesteve]

Isn't that when using the Kelly criteria is usable and, in theory, allowing you to optimize your profit?
Kelly can't make a profit out of bets that aren't profitable in the first place (which I believe was the gist of that example)

[aaron]

His line is accurate to the point that all is 4-1 shots combined will win at 20%.That doesn't mean that his 4-1 shots going off at 10-1 will win at 20 %.His 4-1 shots going off at even money win at 50%,so even with the overall 20% hit rate he cannot show a profit.

[DJofSD]

Kelly can't make a profit out of bets that aren't profitable in the first place (which I believe was the gist of that example)

I guess I didn't understand that important point.

No, not even used car salesmen can make money in the long run when they lose it in the short run.