Bet Overlays Only

[Overlay]

On another thread, xfile posted the following:

Long Term Profit

--------------------------------------------------------------------------------

Why does long term profit elude so many players? It's really the simpler the better:

1. Bet overlays only
2. Place WIN BETS only
3. Skip races that do not offer an edge
4. Apply a good money management plan


In response to this, douglasw32 posted:

This is off topic for this post but maybe you (or any of you) could start a post explaining "1.", I get all the rest, but "1." seems to be very very subjective. Could use a good understanding of, what,when, etc an overlay "IS" an overlay.

Then "we" can all get back to talking handicapping.

In response to douglasw32's suggestion, number 1 in the above list for me means the following:

1. Assigning probabilities based on objective performance data rather than subjective analysis.

2. Using a properly weighted and distributed mix of primary factors (those that have the most powerful statistical correlation with winning chances) from the major performance categories (speed, class, form, pace, etc.), while avoiding redundancy and keeping the variables involved down to a workable number.

3. Rating all horses on a common scale, so that you can calculate fair odds based on the ratio of the rating for each individual horse to the total of the ratings for all the horses in the field. (This will also automatically produce a 100% line, so that you can compare each horse's final odds directly with its toteboard odds in order to find value, without having to do anything further to factor in the effects of take and breakage.) An overlay will then be any horse with toteboard odds which are higher than its assigned fair odds.

[douglasw32]

Thanks :)

[kenwoodallpromos]

How do you handle races involving horses without many races? Doesn't the importance of certain variables change? Assuming one is attempting to determine fair odds, I have thought determining which variables to give what importance to is critical and subject to change often. I find that very tricky.

[douglasw32]

oops hit something, wasn't done yet....

So do you think ranking factors that do not directly relate figure wise, say speed figures compared to morning line odds (I do not, it is just a good example of different types of numbers) you could not add the two, and they have a huge difference to them...so do you think ranking the catagories by percentage, say add all the speed figs and get the percent of the field for each horse, then the ML odds turned into percent so higher is better then ranked and get percent of field and do the same for each factor...then total all the percents???

Anyone think this would work for different "types" of factors.

Usually my posts do not even make sense to me I hope someone else gets what I am saying here.

[Overlay]

How do you handle races involving horses without many races? Doesn't the importance of certain variables change? Assuming one is attempting to determine fair odds, I have thought determining which variables to give what importance to is critical and subject to change often. I find that very tricky.

The only category of horses that I accord "special" treatment to are first-time starters with no race performance data that can be used for probability assignment. If a horse has even one race in its record, the variables and statistics that I use permit comparing it on an equal-footing basis to other horses in the field. The handicapping factors I employ were also chosen to have the widest possible applicability over the entire racing spectrum. I readily acknowledge that my approach is "bulk" in nature, and sacrifices some of the precision that is possible with detailed knowledge and treatment of individual factors. However, I've sought to compensate for that by a combination of the time savings that can be realized, and by the ability to develop workable probabilities for races that might otherwise not be considered "playable" according to "traditional" handicapping standards.

[Overlay]

oops hit something, wasn't done yet....

So do you think ranking factors that do not directly relate figure wise, say speed figures compared to morning line odds (I do not, it is just a good example of different types of numbers) you could not add the two, and they have a huge difference to them...so do you think ranking the catagories by percentage, say add all the speed figs and get the percent of the field for each horse, then the ML odds turned into percent so higher is better then ranked and get percent of field and do the same for each factor...then total all the percents???

Anyone think this would work for different "types" of factors.

Usually my posts do not even make sense to me I hope someone else gets what I am saying here.

I believe I do understand what you're saying. (Should I be worried about that? :) )

As you suggest with your percentage example, the key is finding a common scale to rate all the factors in question on, rather than comparing apples to oranges. Percentages would be workable for that purposes, except that they get cumbersome to deal with mathematically, especially if you're multiplying them. They also can be distorted by differing field sizes. For me, this is addressed by the use of impact values, and the assurance of independence among the variables I am examining.

[boxcar]

On another thread, xfile posted the following:

In response to douglasw32's suggestion, number 1 in the above list for me means the following:

1. Assigning probabilities based on objective performance data rather than subjective analysis.

Hi, Overlay, I visited your website and enjoyed the read and found it to be interesting in the abstract.

A couple of questions: First, regarding your remarks above do you mean for us to infer from them that unless we assign probabilities to a horse based on raw pp performance data that we can't be objective in our analysis? Or that there isn't any other basis for objectivity?

My next is this: From a practical standpoint how does your own odds line stack up against the public's? For example, we know historically that the crowd year in, year out is correct in its top betting choice about 1/3 of the time, in its top two choices about 50% of the time, and in its top three around 2/3 of the time, etc. Therefore, just from the perspective of hit rate, how has your line historically compared to the crowd's?

And finally, since post time odds reflect the collective knowledge and wisdom of the public -- the pubic's consensus -- would you consider the public's "odds line" to be equally as based on "objective performance data", as anyone else's personal line?

Boxcar

[kenwoodallpromos]

Yes, I can see where certain variables would change a lot less than some or could be viewed the same all the time. Like if you figure a way to use horse health or a speed or pace figure, or if certain lengths back or comments in a previous race counted for so much. Makes good sense with maybe a small amount of teaking but a good idea for a base method and if it produces a + ROI great!

[Overlay]

Hi, Overlay, I visited your website and enjoyed the read and found it to be interesting in the abstract.

A couple of questions: First, regarding your remarks above do you mean for us to infer from them that unless we assign probabilities to a horse based on raw pp performance data that we can't be objective in our analysis? Or that there isn't any other basis for objectivity?

My next is this: From a practical standpoint how does your own odds line stack up against the public's? For example, we know historically that the crowd year in, year out is correct in its top betting choice about 1/3 of the time, in its top two choices about 50% of the time, and in its top three around 2/3 of the time, etc. Therefore, just from the perspective of hit rate, how has your line historically compared to the crowd's?

And finally, since post time odds reflect the collective knowledge and wisdom of the public -- the pubic's consensus -- would you consider the public's "odds line" to be equally as based on "objective performance data", as anyone else's personal line?

Boxcar

No, I wouldn't say that there's only one way to be objective about odds assignment. I use the way that I've found has worked best for me. It might be entirely possible for someone with more experience or ability than I have in areas of handicapping that might be considered subjective (like body language characteristics, trip analysis, and so forth) to know (either from actual records maintained or as a result of intuition based on experience) what the winning chances of a horse that possesses a particular performance characteristic would be. The reason I tend toward the statistical is because I've found that the further away I get from it, the harder it is for me both to have sufficient confidence in the uniform application of personal judgment from race to race, and also to be able to go back and analyze which part of my handicapping model might need "tweaking" if results start turning out significantly different than I anticipate.

I've found that my line is generally more effective in approximating the true winning chances of horses than the toteboard odds, particularly at the lower end of the odds spectrum (from which the main contention in a race will come). This is due primarily to the fact that my odds are based on a 100% line, and are thus not distorted in terms of winning percentage by the effects of take and breakage, as the toteboard odds are.

As you say, the collective knowledge of the public is a formidable opponent to try to outwit, given its demonstrated ability in overall long-term odds assignment. However, the public's long-term proficiency doesn't guarantee its accuracy for every horse in every race. My handicapping is predicated on finding those instances where the public has misjudged the winning chances of one or more horses in a race, usually as a result of becoming oversold on a horse with a high figure, a hot jockey, or any number of similar factors which may very well make the horse the single entrant most likely to win today's race, but not at the low odds that the horse is being bet down to. Quite often, the result of my analysis will be to pass such a race altogether, which certainly has as much of a place in the handicapper's arsenal as betting an alternate horse other than the top overall selection.

[boxcar]

No, I wouldn't say that there's only one way to be objective about odds assignment. I use the way that I've found has worked best for me. It might be entirely possible for someone with more experience or ability than I have in areas of handicapping that might be considered subjective (like body language characteristics, trip analysis, and so forth) to know (either from actual records maintained or as a result of intuition based on experience) what the winning chances of a horse that possesses a particular performance characteristic would be.

Thanks for your reply, Overlay.

I apologize for, apparently, not making myself clearer in my post to you. What I was really asking in my first question was whether or not someone can't dispense with the odds assignment process and still be "objective". Is Price Making the only "objective" approach to the game?

The reason I tend toward the statistical is because I've found that the further away I get from it, the harder it is for me both to have sufficient confidence in the uniform application of personal judgment from race to race, and also to be able to go back and analyze which part of my handicapping model might need "tweaking" if results start turning out significantly different than I anticipate.

Ahh...personal judgment -- the most valuable asset a turf speculator can possess. I can relate to what you're saying. In my case, however, I developed sound personal judgment through a lot of study and a lot of practice, which kept me comfortable using a more "esoteric" approach.

I've found that my line is generally more effective in approximating the true winning chances of horses than the toteboard odds, particularly at the lower end of the odds spectrum (from which the main contention in a race will come). This is due primarily to the fact that my odds are based on a 100% line, and are thus not distorted in terms of winning percentage by the effects of take and breakage, as the toteboard odds are.

I toyed around with price lines for a while. I used the lines generated by the A+ software, which a good online buddy incorporated into his own software in order to enhance A+'s performance. As I recall, the top A+ selection won only about 28% of the time. But was really disconcerting was the performance of that line diminshed significantly at its upper scales, as you also experience with your own line . Too often, for my liking, my methodology would come up with a false or vunerable longshot that wouldn't be bettable, according to the A+ line and the whole theory behind price making.

From what I've read about your approach and from my experience with A+, I think I now understand the limitations of price lines. It seems to me that in order to keep odds making "workable" (to borrow your term), it would be necessary to restrict the handicapping factors to a relatively small number.
What this would necessarily mean, then, is that programs like A+, or even manual methods like yours are often seeing a small portion of the overall "race picture" because too many pieces to the race puzzle (handicapping factors) are missing. To state this in other terms, it seems to me that this kind of approach, with its built-in limitations (whether by choice or necessity) can seriously violate what I consider to be the First Handicapping Principle, which essentially says that there aren't any isolated factors in handicapping -- that all factors are related to one another to some degree, and all are especially related to the Form Factor. To limit the comprehensiveness of an approach would necessarily violate this "First Principle".

For example, would anyone's price line be able to find and assess the factors that are subtly expressed in the pp charts? For sure, a method or a program would be able identify easily enough the obvious. But what about the obscure? After all, doesn't the "obscure" -- what's hidden from the public's view -- account largely for why the crowd will overlook a viable longshot play in the betting? Doesn't this fact account for why so many people lose at the game? Of course, the profound paradox here is that everyone is viewing the same data in horses' charts. But sometimes the crowd doesn't know what to make of data when it's expressed in certain ways. And other times the public doesn't even know what to look for, let alone interpret it properly.

As you say, the collective knowledge of the public is a formidable opponent to try to outwit, given its demonstrated ability in overall long-term odds assignment.

You bet the public is! A very "formidable opponent". In fact, when I can get around to it, I'm going to show everyone how to take the public's strength and claim it for your own use! How to use this strength to your own advantage -- with a couple of pretty obscure racing angles, of course. :)

However, the public's long-term proficiency doesn't guarantee its accuracy for every horse in every race. My handicapping is predicated on finding those instances where the public has misjudged the winning chances of one or more horses in a race, usually as a result of becoming oversold on a horse with a high figure, a hot jockey, or any number of similar factors which may very well make the horse the single entrant most likely to win today's race, but not at the low odds that the horse is being bet down to.

Quite frankly, this is something I've never understood. When a horse has lots of things going for him, as you've indicated above, and these positive factors might make him the "most likely" entrant to win the race, what do his odds have to do with him winning? His price is not a handicapping factor! Price is not a factor that exerts any influence upon the outcome of races. The horse doesn't know if he's an overlay or underlay; and besides, plenty of underlays win races!

Once again, thanks for your thoughtful reply.

Boxcar

[twindouble]

"Quite frankly, this is something I've never understood. When a horse has lots of things going for him, as you've indicated above, and these positive factors might make him the "most likely" entrant to win the race, what do his odds have to do with him winning? His price is not a handicapping factor! Price is not a factor that exerts any influence upon the outcome of races. The horse doesn't know if he's an overlay or underlay; and besides, plenty of underlays win races!" Quote;
Boxcar

Yes, I found it a little perplexing as well, you can't find winners just on overlays or underlays but at the same time you can't ignore consensus. M.L. the tote and final off odds, to me only come into play after handicapping the race, you know what horses you like then and only then can you put together a wager that will generate some profit. Theres nothing I like any better than having horses contrary to consensus. Maybe that's what your getting at when you say use the public's strength to your advantage. That's when I go for the score. I'm sure Overlay is handicaping the races as well, I assume he's just saying play value, can't argue with that.

T.D.

[timtam]

I find the money management plan is really tough to stick to. Alot of times

if you frequent the OTB in limited amounts you tend to 'force' bets. Maybe

you're looking for 3/1 on your horse but you still bet it at 5/2 because you're

not going to be back in a week.

[wes]

http://www.ozracetools.com/modules.php?name=Ratings_To_Price


Try you hand at your handicapping numbers. See which are overlays.

wes

[xfile]

Ok........I bet overlays, not which horse looks like the best. 90% of the players bet which horse they think is best. Hence, I guess, that is why 90% don't carry a profit long term. Somehow "magically" horses win who are offered at much higher odds than they should and they win often enough. Why this "magic" happens I don't have to understand. A great example was race 1 at Aqueduct yesterday (December 11, 2005). True Call was sitting on the board at 19-1 a few minutes before post. The horse looked like more of an 8-1 shot. Plus he had red hot Eibar Coa in the saddle. I jumped on it. True Call DID NOT look best, he offered a great overlay opportunity. I only cash about 15%-20% of my bets but I make profit long term (+15% for the year 2005 from January 1 through December 11) This is a +12% plus an average rebate of 3% from Tcbets. Some days I can hit 2 or 3 bombs then go zero for 15 for example the next 2 days. :cool:

[Overlay]

Price-making is certainly not the only objective approach to the game. You can use whatever set of criteria you wish to judge the chances of horses (including the same type of standards that I use to rank an entire field). But if you’re using those criteria to “pick winners” by narrowing every race field down to one horse that you plan to bet without regard for the relationship of its winning chances to its toteboard odds, you can be placing yourself on the wrong side of the percentages of the game, just as you would by playing craps or roulette in a casino, or accepting less than even-money odds on the flip of a coin. I guess the way a person looks at this issue boils down to how they would answer the question, “Is it possible to say before a race is run that any one horse in it has an absolute, guaranteed, 100% chance of winning?” If the answer to that question is “No”, then I don’t see how you can avoid concluding that more than one horse in the race has at least some chance of winning (however small you may judge that chance to be). That, after all, is the whole reason why we have betting and odds to begin with. The very concept of odds is an acknowledgement of the uncertainty involved in race outcomes. I definitely agree that if you choose to bet against the horse that your own handicapping puts on top, you would be making a bad bet if that were the only factor to consider. But if you agree that no horse has a guarantee of winning a race, it seems to me that you also then have to say to yourself, “Yes, I believe that this horse has the greatest chance of any of the entrants to win this race. However, since the horse does not have a 100% winning probability, I am taking some amount of risk by betting it. In light of that, will the odds established on the horse by the betting public compensate me for the amount of risk I’m taking?” You can bet horses that you think have as much as a 75% chance of winning in every race. But if the rest of the public reaches the same conclusion that you do, and bets those horses down to average odds of 1-5, the return that you get on the races where the horse wins won’t compensate you for the losses on the races where it doesn’t. You could say, “Well, I just wouldn’t play the horse if it were bet down that low.” But that would be making your wagering decision based on betting value, just the same as if you were betting on another horse in that same race with toteboard odds that were higher than its true chance of winning. The reason that it makes sense to me to bet that other horse is that, while I realize that the other horse does not have the single best chance of winning this one race, such horses will win often enough in the long run at a sufficiently high rate of return to produce a positive net result, whereas if you consistently bet on horses with odds which do not offer a rate of return great enough to compensate for the chance that they might not win, the long-term result will be an overall net loss.

To comment on some of the other points you brought up, I agree that there are no isolated factors in handicapping, and that, of course, the past performances can be mined almost endlessly for useful information. To me, however, the question is: in light of limited handicapping time, where can the player’s efforts most profitably be directed? While no factor works in isolation, and many factors may influence the outcomes of races, I don’t believe it can be disputed that some elements have a greater effect or correlation in that regard than others. In fact, statistics can show when the relationship of a particular factor to a horse’s probability of winning or losing is so strong that it can’t be accounted for by the fluctuations of random chance, meaning that the factor by itself is influencing the outcome of the race as an independent variable. This is the type of information I try to find and use in my approach. Yes, you could search the past performances for the “subtly expressed” factors that you refer to. But at what point does the law of diminishing returns kick in, where you’re spending more time on your race analysis than is justified by the return you’ll receive if you’re right? I prefer to invoke the Pareto principle instead, and concentrate on the relatively few most significant factors that account for or explain the great majority of race results, rather than getting bogged down in more peripheral elements which, although they might produce a lucrative spot-play angle from time to time, are less useful on a bulk, race-to-race basis.

I also agree that the horse has no consciousness of what its odds are. But its performance today is influenced by the effect of both its own attributes, and by how those attributes compare to every other horse in the field. And each of those attributes has a record of predictive power in indicating how the horse will run today. Some of these factors will have a stronger correlation in that regard than others. Because of this, these factors, in turn, can then be used to quantitatively estimate the horse’s probability of winning today. And it is the comparison of that probability with the odds established by the public that makes the determination of betting value possible. If you don’t capitalize on this pari-mutuel aspect of horse racing, you’re throwing away the advantage that the sport offers in comparison with games of pure chance, where odds are mathematically determined; nothing can be done to influence them; the player never gets back as much for a successful wager as the actual probabilities of the game call for; long-term loss is guaranteed; and any short-term success is strictly a matter of luck. It is undeniable that underlays win races every day. And, again, if the final measure of success were the percentage of winners that you picked, to bet against the one horse that you viewed as the best in the race would be foolish. However, the goal is not merely to pick the highest percentage of winners, but to realize the greatest net return on your wagers. And the key to that is to concentrate your betting on those occasions when you stand to receive a return that is greater than the degree of risk that the wager represents.

[RaceIsClosed]

On another thread, xfile posted the following:

Long Term Profit

--------------------------------------------------------------------------------

Why does long term profit elude so many players? It's really the simpler the better:

1. Bet overlays only
2. Place WIN BETS only
3. Skip races that do not offer an edge
4. Apply a good money management plan



Give me an example of how you find an overlay in a given race. What handicapping methods do you use?

I would think the more difficult part of this equation is finding the horses rather than betting them.

Also, I hear it's good to buy stocks that are going to double in the next six months. Please tell me which ones they are so I can get down.

[RaceIsClosed]

Price-making is certainly not the only objective approach to the game. You can use whatever set of criteria you wish to judge the chances of horses (including the same type of standards that I use to rank an entire field). But if you’re using those criteria to “pick winners” by narrowing every race field down to one horse that you plan to bet without regard for the relationship of its winning chances to its toteboard odds, you can be placing yourself on the wrong side of the percentages of the game, just as you would by playing craps or roulette in a casino, or accepting less than even-money odds on the flip of a coin. I guess the way a person looks at this issue boils down to how they would answer the question, “Is it possible to say before a race is run that any one horse in it has an absolute, guaranteed, 100% chance of winning?” If the answer to that question is “No”, then I don’t see how you can avoid concluding that more than one horse in the race has at least some chance of winning (however small you may judge that chance to be). That, after all, is the whole reason why we have betting and odds to begin with. The very concept of odds is an acknowledgement of the uncertainty involved in race outcomes. I definitely agree that if you choose to bet against the horse that your own handicapping puts on top, you would be making a bad bet if that were the only factor to consider. But if you agree that no horse has a guarantee of winning a race, it seems to me that you also then have to say to yourself, “Yes, I believe that this horse has the greatest chance of any of the entrants to win this race. However, since the horse does not have a 100% winning probability, I am taking some amount of risk by betting it. In light of that, will the odds established on the horse by the betting public compensate me for the amount of risk I’m taking?” You can bet horses that you think have as much as a 75% chance of winning in every race. But if the rest of the public reaches the same conclusion that you do, and bets those horses down to average odds of 1-5, the return that you get on the races where the horse wins won’t compensate you for the losses on the races where it doesn’t. You could say, “Well, I just wouldn’t play the horse if it were bet down that low.” But that would be making your wagering decision based on betting value, just the same as if you were betting on another horse in that same race with toteboard odds that were higher than its true chance of winning. The reason that it makes sense to me to bet that other horse is that, while I realize that the other horse does not have the single best chance of winning this one race, such horses will win often enough in the long run at a sufficiently high rate of return to produce a positive net result, whereas if you consistently bet on horses with odds which do not offer a rate of return great enough to compensate for the chance that they might not win, the long-term result will be an overall net loss.

To comment on some of the other points you brought up, I agree that there are no isolated factors in handicapping, and that, of course, the past performances can be mined almost endlessly for useful information. To me, however, the question is: in light of limited handicapping time, where can the player’s efforts most profitably be directed? While no factor works in isolation, and many factors may influence the outcomes of races, I don’t believe it can be disputed that some elements have a greater effect or correlation in that regard than others. In fact, statistics can show when the relationship of a particular factor to a horse’s probability of winning or losing is so strong that it can’t be accounted for by the fluctuations of random chance, meaning that the factor by itself is influencing the outcome of the race as an independent variable. This is the type of information I try to find and use in my approach. Yes, you could search the past performances for the “subtly expressed” factors that you refer to. But at what point does the law of diminishing returns kick in, where you’re spending more time on your race analysis than is justified by the return you’ll receive if you’re right? I prefer to invoke the Pareto principle instead, and concentrate on the relatively few most significant factors that account for or explain the great majority of race results, rather than getting bogged down in more peripheral elements which, although they might produce a lucrative spot-play angle from time to time, are less useful on a bulk, race-to-race basis.

I also agree that the horse has no consciousness of what its odds are. But its performance today is influenced by the effect of both its own attributes, and by how those attributes compare to every other horse in the field. And each of those attributes has a record of predictive power in indicating how the horse will run today. Some of these factors will have a stronger correlation in that regard than others. Because of this, these factors, in turn, can then be used to quantitatively estimate the horse’s probability of winning today. And it is the comparison of that probability with the odds established by the public that makes the determination of betting value possible. If you don’t capitalize on this pari-mutuel aspect of horse racing, you’re throwing away the advantage that the sport offers in comparison with games of pure chance, where odds are mathematically determined; nothing can be done to influence them; the player never gets back as much for a successful wager as the actual probabilities of the game call for; long-term loss is guaranteed; and any short-term success is strictly a matter of luck. It is undeniable that underlays win races every day. And, again, if the final measure of success were the percentage of winners that you picked, to bet against the one horse that you viewed as the best in the race would be foolish. However, the goal is not merely to pick the highest percentage of winners, but to realize the greatest net return on your wagers. And the key to that is to concentrate your betting on those occasions when you stand to receive a return that is greater than the degree of risk that the wager represents.

The "money management" and "overlay" gurus always do the same thing: they talk about the importance of value but do not give specifics on how to find it in an actual race.

As if I need their help converting my positive-ROI selections to profit. Finding the selections is the key.

Then of course there is my simple THROW OUT THE CHALK method.

[Overlay]

The "money management" and "overlay" gurus always do the same thing: they talk about the importance of value but do not give specifics on how to find it in an actual race.

I completely agree with you. That's why I place the importance that I do on the use of hard statistics in deriving a fair-odds line to locate value, rather than relying on "feel" or guess (no matter how well-educated). The line can then serve not only as a guide to each horse's winning chances, but also as an indicator of which horses to play and how much to bet.

[Overlay]

Give me an example of how you find an overlay in a given race. What handicapping methods do you use?

I use a statistical approach based on assigning proper weights to the specific published performance measures from the major handicapping factors/categories that are the most effective and durable in predicting the outcomes of races (while also taking into account the varying effects of considerations such as age, distance, and running surface). I then apply the rules of probability to them to produce a line that reflects the influence of each variable. For that reason, I really can't classify my procedure as a "type" of method in the usual sense (speed, class, pace, and so forth), since I look at a variety of elements.

[RaceIsClosed]

I use a statistical approach based on assigning proper weights to the specific published performance measures from the major handicapping factors/categories that are the most effective and durable in predicting the outcomes of races (while also taking into account the varying effects of considerations such as age, distance, and running surface). I then apply the rules of probability to them to produce a line that reflects the influence of each variable. For that reason, I really can't classify my procedure as a "type" of method in the usual sense (speed, class, pace, and so forth), since I look at a variety of elements.

Any profitable "value" system is going to have several bread-and-butter angles that the public repeatedly overlooks.

You're very generous in sharing insight into how to manage money, but you don't seem as liberal with actual methods that show actual profits in actual races. I tried to get that ball rolling with my "let's kill the goose" thread when I said you can safely throw out almost any horse who is claimed off a "supertrainer" (Dutrow, Lake, Klesaris, Pino, Pletcher, Capestro, Norman, etc.), unless it's going to another supertrainer, or to a barn that has previously won with such a tactic (such as Aristone claiming off of Lake).

See the difference? That's a real angle you can put to real use in real races. A "value" method would naturally yield this type of horse, but you don't have to do the actual calculations to know you have value by throwing out a horse like this.

So what's ONE thing that constantly tilts your value meter towards or against a horse? Anyone can manage their money once they have an edge.

[Handiman]

Hi Guys,


I'm writing a program that ranks horses and provides an odds line. Right now, the program offers the user the choice of Procapps file or TSN .50 file. Then it allows the user to select up to 8 racing factors to use to handicap the race. User can select route or sprint and then up to the 8 factors.
Then program offers user chance to enter scratches, then lists the horses by odds ranking. There is also two predefined handicapping setups...one for route and one for sprint.

Right now it will just do one track at a time. I Hope to have it able to run 5 tracks at once.

Now I am a hobbyist programmer...and not a professional, but the program is programmed to use in windows and not DOS. I don't want it to be just a replica of the other's on the market. I've also read where, and believe it to be true, that too much information just clouds the issue. Since it is based on finding value, I'm tying to keep the options down to a minimum.

I'm posting here cause this is a thread on Value. And I'm not offering it for sale...I'm not offering it up at all,(don't want to be accused of hustling software here) but would like to know, what else might fit in, without running into the 'too much info' neighborhood.

I've looked at many programs, and while they are beautifull, and have a gazillion options and screens, trying to put all the info into a usable form, is much to much for me.

One thing I have found, that most programs limit the odds of the horses, to around 20 to 1 or so. I don't believe that is right. Alot of people say each horse, say in a 10 horse race, have a basic 10 ino 1 chance of winning, by the mere fact there are just 10 horses entered. I again don't believe that to be true. My point is, I let the numbers speak for themselves. While testing my program, I had a horse listed at 73 to 1, and when checking the results, the horse went off at 73 to one. It finished 2nd. It was near the bottom of my list, to say the least. So for software to limit odds to 20 to 1 at most is unrealistic. It also throws the other end of the odds off. Unless the programmer just draws the line at 20.

As for the 10 in 1 chance of winning, that isn't true either. I was at Bay Meadows a few years back, and 6 horses were entered. Based on general odds belief, each horse had a 6 in 1 chance of winning. But if there had been just the number 1 horse in the gates...no other horse entered, he still had no chance of winning...cause he wasn't going to run...He left the gate, dumped the jockey and then jumped over the inside rail and ran around in the infield .

My point is that some horses in any particular race just will not win and have no chance of winning. Obvious, by the fact that only one horse is declared the winner. So when my software lists a horse's odds at 85 to 1, I don't believe it is out of line. All that being said, again is there something I'm missing and, is there something more I should include....do you think?

Thanks in advance for your comments.

Handi

[RaceIsClosed]

Hi Guys,


I'm writing a program that ranks horses and provides an odds line. Right now, the program offers the user the choice of Procapps file or TSN .50 file. Then it allows the user to select up to 8 racing factors to use to handicap the race. User can select route or sprint and then up to the 8 factors.
Then program offers user chance to enter scratches, then lists the horses by odds ranking. There is also two predefined handicapping setups...one for route and one for sprint.

Right now it will just do one track at a time. I Hope to have it able to run 5 tracks at once.

Now I am a hobbyist programmer...and not a professional, but the program is programmed to use in windows and not DOS. I don't want it to be just a replica of the other's on the market. I've also read where, and believe it to be true, that too much information just clouds the issue. Since it is based on finding value, I'm tying to keep the options down to a minimum.

I'm posting here cause this is a thread on Value. And I'm not offering it for sale...I'm not offering it up at all,(don't want to be accused of hustling software here) but would like to know, what else might fit in, without running into the 'too much info' neighborhood.

I've looked at many programs, and while they are beautifull, and have a gazillion options and screens, trying to put all the info into a usable form, is much to much for me.

One thing I have found, that most programs limit the odds of the horses, to around 20 to 1 or so. I don't believe that is right. Alot of people say each horse, say in a 10 horse race, have a basic 10 ino 1 chance of winning, by the mere fact there are just 10 horses entered. I again don't believe that to be true. My point is, I let the numbers speak for themselves. While testing my program, I had a horse listed at 73 to 1, and when checking the results, the horse went off at 73 to one. It finished 2nd. It was near the bottom of my list, to say the least. So for software to limit odds to 20 to 1 at most is unrealistic. It also throws the other end of the odds off. Unless the programmer just draws the line at 20.

As for the 10 in 1 chance of winning, that isn't true either. I was at Bay Meadows a few years back, and 6 horses were entered. Based on general odds belief, each horse had a 6 in 1 chance of winning. But if there had been just the number 1 horse in the gates...no other horse entered, he still had no chance of winning...cause he wasn't going to run...He left the gate, dumped the jockey and then jumped over the inside rail and ran around in the infield .

My point is that some horses in any particular race just will not win and have no chance of winning. Obvious, by the fact that only one horse is declared the winner. So when my software lists a horse's odds at 85 to 1, I don't believe it is out of line. All that being said, again is there something I'm missing and, is there something more I should include....do you think?

Thanks in advance for your comments.

Handi

There are people who make money by using one factor exclusively, be it trainer, pedigree, speed figures, pace, class, shipping, or other angles.

The problem isn't that you would be using too much information if you tried to weigh all those factors, but simply that you might make a mistake with any one of them that winds up poisoning your numbers through the inflation of longshot ratings or deflation of favorite ratings.

You can try to find patterns that say if A occurs in the pps, then B will occur in the race, and it might hold up, but you'll have to do a lot of experimenting.

[xfile]

Any profitable "value" system is going to have several bread-and-butter angles that the public repeatedly overlooks.

You're very generous in sharing insight into how to manage money, but you don't seem as liberal with actual methods that show actual profits in actual races. I tried to get that ball rolling with my "let's kill the goose" thread when I said you can safely throw out almost any horse who is claimed off a "supertrainer" (Dutrow, Lake, Klesaris, Pino, Pletcher, Capestro, Norman, etc.), unless it's going to another supertrainer, or to a barn that has previously won with such a tactic (such as Aristone claiming off of Lake).

See the difference? That's a real angle you can put to real use in real races. A "value" method would naturally yield this type of horse, but you don't have to do the actual calculations to know you have value by throwing out a horse like this.

So what's ONE thing that constantly tilts your value meter towards or against a horse? Anyone can manage their money once they have an edge.

The one thing I have learned over the years that is CONSTANT is that NOTHING in this game is constant. It changes repeatedly and those who can make reasonably quick adjustments will do well. The problem I have seen is too many players looking for that constant "holy grail" that does not exist. Personally, I take every race on each day as a new chapter. Previous chapters MAY lead to what will happen in this current chapter OR it might be totally different. I guess that is why many consider this an art instead of a science. :cool:

[xfile]

Give me an example of how you find an overlay in a given race. What handicapping methods do you use?

I would think the more difficult part of this equation is finding the horses rather than betting them.

Also, I hear it's good to buy stocks that are going to double in the next six months. Please tell me which ones they are so I can get down.

With all due respect, when did I become your personal handicapper AND stock advisor?. Did I miss a meeting on this?.....lol......:D :cool: ......roll your sleeves up and get to work!......:cool: ...I spend 15 hours a day on my investing and horse wagering. My wife has to watch me do this from home 7 days a week 365 days a year. She is a very patient woman. At 40 years old I don't know everything but what I do know is learning from 20 years of experience and from mistakes and always be looking for trends....good luck and happy holidays :cool:

[sjk]

Handiman,

The most important question to answer is whether the program is being developed to sell and support or for personal profit and enjoyment. There could be a debate as to which is ultimately more remunative and satisfying.

If you are developing the program to use yourself I would dispute the premise that you want to limit the types and amount of information used. I have used everything that I can possibly think of in my program and I would think that any piece left out would devalue the whole and use the computer in a less than opimal way.

The advantage that a piece of software has over a person is that it can incorporate more types of information into the analysis and do it more comprehensively and far more quickly than a person could do in a reasonable amount of time. Of course your time is spent up front in data analysis and program design rather than later when others are handicapping individual races.


On the other hand if you are going to have to explain the program to others and more information translates into user confusion and complications in support I could see the downside.

As to placing artificial limits on a horse's chances to win I do not do so. I let the calculations dictate the probabilities although I do not play horses to win that I give less than a 5% chance of doing so or horses to run second that I give less than a 3% chance of doing so.

[Overlay]

Any profitable "value" system is going to have several bread-and-butter angles that the public repeatedly overlooks.

You're very generous in sharing insight into how to manage money, but you don't seem as liberal with actual methods that show actual profits in actual races. I tried to get that ball rolling with my "let's kill the goose" thread when I said you can safely throw out almost any horse who is claimed off a "supertrainer" (Dutrow, Lake, Klesaris, Pino, Pletcher, Capestro, Norman, etc.), unless it's going to another supertrainer, or to a barn that has previously won with such a tactic (such as Aristone claiming off of Lake).

See the difference? That's a real angle you can put to real use in real races. A "value" method would naturally yield this type of horse, but you don't have to do the actual calculations to know you have value by throwing out a horse like this.

So what's ONE thing that constantly tilts your value meter towards or against a horse? Anyone can manage their money once they have an edge.

As you indicate, one way to approach the question of value is through the use of a collection of angles that have known probabilities associated with them. The problem I’ve found there is that, sooner or later, the public will catch on to those angles, and, one by one, they’ll lose their profitability through overbetting (“killing the goose”). I prefer to assign an overall winning probability to each horse in a field through the use of multiple fundamental factors. None of the factors may be profitable individually if employed as a basis for selection to the exclusion of all other considerations. But the more important feature about them is that each one has demonstrated and retained its effectiveness in predicting race results over time. When combined through proper weighting, they provide an accurate estimate of a horse’s true overall winning chances. The only way that information of this type could become obsolete would be if the public suddenly started betting every horse in every race to the exact odds corresponding to the horse’s true chance of winning. And, despite the public’s demonstrated aggregate ability in odds-setting over the long term, I view accuracy of that degree to be a practical impossibility. That’s how I seek to obtain an edge, rather than through the use of isolated angles, no matter how profitable they may presently seem to be. Then I don’t have to always keep searching for new, undiscovered variables as current ones start losing their positive return through overplay.

I also noted your mention of throwing out horses claimed off a “supertrainer”. I agree that most races will have one or more horses that produce a reaction of, “Not in this lifetime” when you look at their past performances, or that have other strong negative factors in their records (like the one you cite). However, I prefer not to completely rule out the chances of any horse, since I’ve found that it’s from these horses with higher fair odds that the bulk of any race meeting’s highest paying longshots will come. But I also believe that my current approach will assign fair odds to those horses that are sufficiently high so that the horses will either not offer betting value at all, or not offer sufficient value to merit even a small wager. In my experience, horses that do qualify for wagers by my measures have enough redeeming qualities in their records that they truly do represent value in comparison to their actual winning chances.

[RaceIsClosed]

As you indicate, one way to approach the question of value is through the use of a collection of angles that have known probabilities associated with them. The problem I’ve found there is that, sooner or later, the public will catch on to those angles, and, one by one, they’ll lose their profitability through overbetting (“killing the goose”). I prefer to assign an overall winning probability to each horse in a field through the use of multiple fundamental factors.

Which factors, and how?

Is this a board for sharing handicapping information or a propaganda board for selling products?

[Overlay]

I don't believe your last post is justified. Yes, I have my own website where I market handicapping titles that I've developed through many years of experience, research, and trial and error. But I think that I have a record of posting to this board that is long enough and diverse enough that any number of "regulars" would attest that I am not here just to persuade those who read my posts to buy my items. My primary objectives (to the extent that I have any type of agenda) are to emphasize the personal importance that I place on basing handicapping and wagering decisions on the assignment of fair odds, and the presence of wagering value, as opposed to the more "traditional" handicapping mindset of concentrating solely on narrowing a race field down to the one horse that is most likely to win; and to give my fellow board members an appreciation for that point of view. I keep coming back to this point because I believe strongly in it, and because it seems apparent from the questions that I am asked on the board that there is still a great deal of misunderstanding (even among the knowledgeable players who post here) as to why it might make sense to bet a horse other than the one that your own handicapping puts on top in a particular field. It is only because my titles are so closely tied to this philosophy that I find it difficult to discuss this subject without leaving myself open to the interpretation that I am posting merely for the sake of exposure or self-promotion. I think that reading my posts in the aggregate will reveal enough about my methodology (including the general handicapping categories I look at) that the only other thing about it left to discuss would be the exact factors, values, and calculation procedures that I use. I don't withhold this information out of the normal concern about driving down mutuel prices if the word gets out, since my whole approach is designed to eliminate that type of concern. But I believe that I'm entitled to compensation for that information in view of the time and effort spent in developing it, the same as with any of the other respected board members who independently market their own items without revealing all of the mechanics of them on the board.

[BIG RED]

http://www.ozracetools.com/modules.php?name=Ratings_To_Price


Try you hand at your handicapping numbers. See which are overlays.

wes

Nice little site wes, I'll look around a bit when I have some time,thx.

[RaceIsClosed]

I don't believe your last post is justified. Yes, I have my own website where I market handicapping titles that I've developed through many years of experience, research, and trial and error. But I think that I have a record of posting to this board that is long enough and diverse enough that any number of "regulars" would attest that I am not here just to persuade those who read my posts to buy my items. My primary objectives (to the extent that I have any type of agenda) are to emphasize the personal importance that I place on basing handicapping and wagering decisions on the assignment of fair odds, and the presence of wagering value, as opposed to the more "traditional" handicapping mindset of concentrating solely on narrowing a race field down to the one horse that is most likely to win; and to give my fellow board members an appreciation for that point of view. .

That approach is useless without the means to assign a better oddsline than the track handicapper, which you are implying you have. So what is the track handicapper missing? If you had a profitable method you would have no problem whatsoever doing this. Want another example? Kirian McGlaughlin's second-time starters in New York often run second first time out with a low figure, yet always seem to contend for the win even when apparently "outclassed." The public tends to underestimate these horses, and they often pay good prices.

What you're looking to do here is sell stuff (otherwise we wouldn't even have to know you have a website), get credibility for yourself as someone who can help others make a profit, and to do so not through any handicapping insight whatsoever, but by the very obvious advice to "make an oddsline" and "bet overlays." If the answer to "how do I FIND these overlays" is "do it yourself" or "buy my stuff," then that's beyond the purpose of a message board and becomes little more than propaganda.

You may want everyone to roll over and say what a great product you have, but you can't come into a place banking on the discusssions favoring you and then turn tail and run when a question you don't want to hear is asked.

[traveler]

Obviously some members are big advocates of only betting value, while others are selection oriented. If the selection oriented players are making money, their overall picks are value by default it would seem to me, or they would be losing money. My question to the "value only" players is - you have to make a line to decide if you have a value play, do your horses win at the percentage you have made your line to be? In other words do your horses you make 2-1 win 33% of the time, 4-1 win 20% etc. etc.? If so great, if not what does that tell you about your line making? I am not trying to be a wise guy, just trying to understand the line-making, value concept better than I do now. Thank you.

[twindouble]

Obviously some members are big advocates of only betting value, while others are selection oriented. If the selection oriented players are making money, their overall picks are value by default it would seem to me, or they would be losing money. My question to the "value only" players is - you have to make a line to decide if you have a value play, do your horses win at the percentage you have made your line to be? In other words do your horses you make 2-1 win 33% of the time, 4-1 win 20% etc. etc.? If so great, if not what does that tell you about your line making? I am not trying to be a wise guy, just trying to understand the line-making, value concept better than I do now. Thank you.

Your not alone, from what I've read here there's a new breed of handicappers that are attempting formulate new scientific tools to measure all aspects of handicapping and wagering. They have the magnifing glass out looking at every possable angel and nuance, then extrapulating that into numbers. To me it's like looking for a hidden code that will produce winning strategies.


When some tracks went to measuring times in a 100th of a second, I laughed Because who would say to themselves this horse could win by 2/100's of a second so I'll unload. Just knowing all the unforeseen variables that can happen in any race makes it a joke. I want open lengths when I make a serious wager.

I'm not suggesting that breaking down things in a scientific way won't produce something of value, I just won't take the time and effort to understand it all when I can pick the the DRF and do just fine without it.

All I can say is, some here swear by what they do and profess to make a lot a money using these new approaches to handicapping and wagering. It's up to you decide.


Good luck,

T.D.

[boxcar]

Essentially, Overlay, we’re on the same page in terms of our betting philosophy. I’m in complete agreement with you that it would be foolhardy to bet any horse, no matter how well he “figures”, apart from any consideration to his post time odds.

And again, serious players should search for, identify and focus their attention on good wagering opportunities wherein the return would be greater than the degree of risk.

Where we differ significantly, however, is how we approach our wagering, how we assess horses’ “chances”, and the critically important role a comprehensive handicapping methodology plays in providing the handicapper with ample opportunities throughout the month, meet and year to identify viable long shot plays and cash bets on nice-sized mutuels.

I’ve always looked at handicapping methodologies as tools in the hands of workmen. As a workman, my goal was to realize the greatest return possible on my investments. And I knew that the only way I could do this was to capture my fair share of long shot winners. And I also knew that the primary reason the public allows viable long shot plays to slip through its collective fingers is because they focus nearly all their attention on the Obvious, and because they too often accepted the Obvious at face value. These two indisputable facts account for why they public is so consistently good at picking the right horse in a race about 1/3 of the time, since very often the Obvious can be accepted at face value . Frequently, it is nothing more than a “WYSIWYG” situation in terms of a horse’s pp chart. But by the same token, there are times when the Obscure is overlooked and a horse’s poor appearing pp chart is blindly accepted as being the Obviously Bad. The Public only unwittingly, unknowingly, unintentionally overlooks viable long shot plays. And while these “times” may be relatively rare compared to how often low-priced “obvious logical selections” score, nonetheless nothing, in my opinion, stings worse than letting a big one slip away. I can easily endure a few chalkish winners getting away for one reason or another, but missing a vulnerable or, worse yet, false long shot hurts badly. Why? Because these kinds of slip-ups are very costly, and detrimental to one’s financial well being! Conversely, recognizing and acting on viable long shot plays will put you miles ahead in the game. Will drown you in the Black. (What a way to go!) :)

I never forgot something a former “track rat” buddy of mine (who is now deceased) from many years ago told me when I started out learning the game. He told me that “a few long shot winners during a meet will cover a multitude of sins -- and then some”. I never forgot this. Truer words were never spoken. The “sins” to which he was alluding were such things as poor personal judgment, poor racing luck, less-than-physically fit selections, unintentional handicapping oversights, etc.

These are the reasons why I devoted so much time "mining gold" out of pp charts. I wanted to avoid the specific mistakes the public often makes. And they made these mistakes because they either didn’t have the right tools for the job, lack of know-how, or as you have said due to time restraints.

Another reason I avoided going the personal price line route was because I felt it was “splitting hairs” – in a game that is as much “art” as it is “science”. And it would seem to me that odds making is a very inexact science because unlike craps, roulette, coin tosses, etc., the variables in racing aren’t static. The probabilities change from race-to-race because every race is a unique problem unto itself. No two races are exactly the same. Therefore, I have to wonder about the “accuracy” of such lines.

Permit to give you an illustration of what I mean. Let’s take the Drop in Class Last Start (DCLS) angle that I discussed in my long shot thread. There are two ways to determine the value of this maneuver in today’s race. One way is to compute the value on the basis of long-term win percentages of horses that, historically, have had this maneuver in their charts. In this situation, every horse in today’s race that has this maneuver would get equal credit for it.

But does or even should this trainer maneuver angle have the same value in every race? Absolutely not! Too many other variables, for one thing, would affect the value of the angle in a horse’s particular chart in today’s particular race. What if we were to assess the value of this angle on a race-specific basis – as it pertains to this particular race? Let’s say we were to evaluate the worth of the maneuver off the race that forms the basis to this angle (i.e. 2RB for this angle) and calculate the difference between the claiming value of that race and the horse’s LR, then we could well come up with very different values in situations wherein more than one horse in the race has this angle in his chart!

Let’s say one horse was dropped to the $5,000. level in his LR off his 2RB that was a $10,000. claimer. The dif is $5,000. or 50 points, if for ratings convenience we drop all but the first two numbers of the difference. And another horse in the same race dropped to the 5K level last start off his 2RB where he was entered for $6,500., making his maneuver worth only 15 points, etc, etc. Even here with this rating system, I wouldn’t dare take these numbers at face value. There would be several other closely related factors ([perhaps what you’d consider to be “peripherals”) that I’d have to first examine. For example, the horse with the 35 point lead might really be sporting a big Back Class deficit compared to the cheaper appearing horse. But we could only find out by examining the speed and pace ratings of each horse’s “back class” race. Or maybe the form cycle of the horse with the raw 35 point angle advantage indicates that the horse is on the downward side of his form cycle curve, whereas the cheaper horse is on the upward side of his curve, etc. And so it would go with each closely related factor in turn – each factor weighing in on the worth of the one under consideration. Each factor either contributing to or detracting from the value of the angle.

The fact that horse races are unique problems unto themselves is recognized by all seasoned players, is it not? This is why savvy handicappers dismiss the idea of “hard ‘n’ fast rules”, isn’t it? And rightfully so! These kinds of players know the dynamics of every race are different! This is why guideliens are preferred to rules by such players.

This is precisely why I asked you earlier how your personal line – your odds making methodology stacked up against the public’s. The public’s line is very race-specific, as it brings its collective expertise to bear upon a particular race, which largely accounts for why the crowd is such a “formidable opponent” at the windows. However, price lines can’t really be all that race-specific since the lines (i.e. percentages) find their ground in thousands of races that were run prior to today’s unique event. These long-term percentages, then, are static or fixed, whereas they really should be dynamic or variable in nature– according to the unique mix of factors or attributes in each horse’s chart, and how all these in turn uniquely compare to every other horse’s unique attributes in their charts.

We also differ in how we should best budget our time in our quest to “realize the greatest net return” on our wagers. To me, the factors that actually exert an influence upon the outcome of races should be given the highest priority. Why? Because when it comes down to it, you still want to be on the horse that is a) most likely to demonstrate the most improvement in today’s race; and b) who is well meant. Now, I’m not saying here that you want to bet on these kinds of horses at any price. Of course not! I, for one, have always desired the twin inducement of “logical selections” or “solid contenders” + “good prices”. And, naturally, some races will have more than one “logical selection”. Other races might have none because they shape up as “chaotic”, or even too contentious.

Furthermore, genuine handicapping factors would be my “first line of defense”, as it were, in sifting the wheat from the chaff. I wouldn’t care if a horse under consideration is being sent off at 4-1 or 40-1 when I toss him on the basis of handicapping factors. If my handicapping tells me that an entrant is in too tough of a spot today , then I care not a whit about his price. So, why waste valuable time figuring such a horse’s winning chances in today’s race?

(Permit to define a “tough spot” in the context of my methodology. A horse would be taking the worst of it whenever he’s angle-poor in terms of positive angles, and/or too angle-rich in terms of negative ones. An angle-poor horse might be one with simply too few angles going for him, or little or nothing in terms of “power angles” which are powerful angle combos.)

So then, how did I approach the important Price Factor with my contenders? Well, being a big believer in the K.I.S.S. principle, once I isolated my bona fide contenders after making all my eliminations, the toss-outs, for all practical intent and purposes, became non-existent. This usually left me with 2 to 4 horses with which to concern myself for wagering purposes. (Life is good when we can keep it simple!) :)

At this point, I turn my undivided attention to the board to see what the public thinks about my contenders. Remember: I’m pitting my handicapping expertise (first and foremost) against the collective expertise of the public. Let’s say, I have three contenders in a race. And a lot could be said for all of them. In my opinion, any one of these could take the race. They all stand about an equal chance of winning (not wanting to split hairs here). Now I look at the board and see that the crowd has bet down one them to 8/5, the other is at 7/2, and the other is at 6-1. It doesn’t take a mathematician to see where the overlays are in this race – if my handicapping assessment is more or less on the money – if my evaluation of the specific mix of factors and attributes in this particular race is more or less accurate. In this situation, I’d dump the lowest price horse, leaving me with the two “overlays” in the race, and the opportunity to implement my money management plan in order to realize an acceptable profit if one of my selections make good in the win pool, or both of my selections make good in the exacta pool.

My “edge” in the game, therefore, was two-sided: One edge being my knowledge of racing angles, and my ability to properly evaluate them – being able to often see the “little” (but nonetheless important) things the public overlooked. This was my handicapping edge –one largely but not exclusively of my own making. As for my wagering “edge”, the public provided me with that whenever they blundered badly in the handicapping department! Therefore, my “overlays” came in two forms – in the form of handicapping, since one or more good angle horses would have advantages over the rest of the field, and in the form of betting, since the public would frequently misjudge or not even acknowledge in the wagering any advantages that my contenders enjoyed.

In summary, therefore, we’re in basic agreement over the necessity of overlays, if one hopes to survive in the game. And we’re both in agreement as to needing a price line – an investment guide, if you will to help us recognize overlay situations and take advantage of decent investment opportunities. However, where we differ significantly is how we go about finding a reliable “investment guide” – a decent “odds line”.

Boxcar

[Overlay]

"That approach is useless without the means to assign a better oddsline than the track handicapper, which you are implying you have. So what is the track handicapper missing? If you had a profitable method you would have no problem whatsoever doing this. Want another example? Kirian McGlaughlin's second-time starters in New York often run second first time out with a low figure, yet always seem to contend for the win even when apparently "outclassed." The public tends to underestimate these horses, and they often pay good prices.

What you're looking to do here is sell stuff (otherwise we wouldn't even have to know you have a website), get credibility for yourself as someone who can help others make a profit, and to do so not through any handicapping insight whatsoever, but by the very obvious advice to "make an oddsline" and "bet overlays." If the answer to "how do I FIND these overlays" is "do it yourself" or "buy my stuff," then that's beyond the purpose of a message board and becomes little more than propaganda.

You may want everyone to roll over and say what a great product you have, but you can't come into a place banking on the discusssions favoring you and then turn tail and run when a question you don't want to hear is asked."



The track handicapper tries with the morning line to anticipate how the public will behave. The morning line is also artificially constrained to a certain degree as to how low the line can be for favorites or how high it can be for longshots, in order to stimulate betting interest on all entrants. The public (which is the bettor's real competition, although it is influenced by the morning line) takes over from there in refining the line. And as proficient as the public is in setting odds on a long-term overall basis, there are and will always be individual cases where the public allows itself to be overly influenced by factors which may indeed have a bearing on a horse's chances of winning the race in question, but not to the extent that the public believes (as substantiated by performance statistics).

You point out that the public tends to underestimate the winning probability of McGlaughlin horses in their second races, as reflected by their high odds, based on your experience with these horses, or on hard statistics of which you have knowledge. That's exactly what I'm doing with the performance factors I apply. For now your approach is profitable. But there may come a time when enough people make the same observations that you have about McGlaughlin's horses, so that the second-timers start getting bet down to odds below what their actual winning probability is. At that point, you will have to start looking elsewhere for another angle that the public has not caught on to yet. I try to address this by developing a composite picture of an entire field's chances based on proababilities associated with fundamental, recurring factors (such as class, condition, speed, pace, distance, and running surface), so that the need to make such basic or frequent changes in my handicapping model is minimized.

Aside from the fact that I am already an established presence on this board (rather than having just "come into [this] place"), I have no need to "get credibility". My items have been featured by reputable handicapping outlets for years. My main "message" on this board is to communicate the value of the concept of betting overlays to those who may have never looked beyond "picking winners" as being the object of the game, or who have not adequately considered the importance of getting the better of the odds on a wager rather than just cashing a ticket. That is the handicapping insight I am offering, and any references I may make to the specific elements of my approach are subordinate to that. As I said, it appears from questions I have seen that the concept of betting for value is still not too widely understood or appreciated. And if that basic premise has not yet become apparent to someone, of what use will more detailed discussion of my methodology be? Also, I am far from the only poster on the board to include their website address in the signature line on their posts, and I do not use my posts as a blatant invitation to frequent my website. (I do not believe that merely discussing my general handicapping philosophy or approach constitutes a "plug" for my specific site or products, any more than it would for any of the other individuals on this board who market their own items. Nor have I noticed or, to my knowledge, has anyone else commented unfavorably on, my being less forthcoming than they are in the comments that we post on the board. Apparently our views on this differ.)

It seems that we approach this subject from divergent (but I would hope not irreconcilable) perspectives. One of the great things about thoroughbred handicapping is how many different roads there can be to success, and the sheer number of varying emphases or points of view that can be of equal utility in beating the game.

[Handiman]

I'd like to weigh in here. Value is a must. win % in relation to odds is like a teeter totter. One goes up the other goes down and vice versa. If not, then you lose moeny....Unless your absolutely amazing and your win % goes up and so do your winning odds. Not likely these days anyway. So for every level of winning % you can achieve you must have a corresponding odds level to produce a profit. I see value betting as the only real option.


As I stated somewhere before, I am a hobbyist programmer. I'm working on a program right now. Infact testing it today at Tampa. Anyway, it's not just finding horses with value, that's important, it's also a matter of deciding if there is value in the race that is detectable and predictable...

I've set my software up, so that when I detect value, my top contenders show good odds. But when race is way too contentious for me, then the top contenders have worse oddds and I look to skip. It's really not important if your odds reflect the same hit percentage as the crowd. It's important that when you bet your horse or two that they pay more than they should. Thus, a value play....and not play when you can't detect value.
Hope that all makes sense.

Handi

[kenwoodallpromos]

For me how far above public odds the "fair odds" are = value = overlay.
I determine the probability (fair odds) of my bet succeeding based on my handicapping which can include looking at the other horses if a straight win bet. But I do not actually make up a set of odds covering all horses because I have a good idea based on my systems what horses I am considering.

[boxcar]

Which factors, and how?

Is this a board for sharing handicapping information or a propaganda board for selling products?

That was a little harsh and below the belt, wasn't it? Neither Overlay or anyone else is obligated to share handicapping info. We can share opinions, but that's not the same as revealing hard information of any real worth.

And so what Overlay is selling products? I have to think he's selling things in which he has a good deal of confidence. Just because he's an information vendor doesn't mean that he can't write about (even passionately!) the very things about which he has strong convictions. About what else is he going to write? Methodologies or approaches he doesn't believe in or use!?

Moreover, the man appears to be a real gentleman. Not once, in his passionate, eloquent and well-articulated defense of his approach has he even hinted that "his way is the only way".

Shame on you, sir. :confused:

Boxcar

[RaceIsClosed]

"That approach is useless without the means to assign a better oddsline than the track handicapper, which you are implying you have. So what is the track handicapper missing? If you had a profitable method you would have no problem whatsoever doing this. Want another example? Kirian McGlaughlin's second-time starters in New York often run second first time out with a low figure, yet always seem to contend for the win even when apparently "outclassed." The public tends to underestimate these horses, and they often pay good prices.

What you're looking to do here is sell stuff (otherwise we wouldn't even have to know you have a website), get credibility for yourself as someone who can help others make a profit, and to do so not through any handicapping insight whatsoever, but by the very obvious advice to "make an oddsline" and "bet overlays." If the answer to "how do I FIND these overlays" is "do it yourself" or "buy my stuff," then that's beyond the purpose of a message board and becomes little more than propaganda.

You may want everyone to roll over and say what a great product you have, but you can't come into a place banking on the discusssions favoring you and then turn tail and run when a question you don't want to hear is asked."


You point out that the public tends to underestimate the winning probability of McGlaughlin horses in their second races, as reflected by their high odds, based on your experience with these horses, or on hard statistics of which you have knowledge. That's exactly what I'm doing with the performance factors I apply. Aside from the fact that I am already an established presence on this board (rather than having just "come into [this] place"), I have no need to "get credibility". My items have been featured by reputable handicapping outlets for years.

Which performance factors do you apply, and how? I gave a very simple angle that is not even in my top 10, but which is something people can verify on their own by looking into it.

If you have "credibility" in "other threads" then surely you've already parted with some type of concrete handicapping wisdom. Being an "established presence" means nothing, as I could quote many touts who don't win who are also established.

If you put a link to your site in your sig, you have chosen to market your website through your poss. You also have an agenda that says you want us convinced that you know the way to profitability, a road that obviously would go through the checkout at your website. It's perfectly fair game to call this marketing, because that's what it is.

I said before, if the answer to the question "how do I find overlays" is "do it yourself" or "buy my stuff" then you are not participating on this board except for using it as a marketing vehicle, or telling me to do it myself, in which case I would, according to your own words, have absolutely no use for you.

I'll share ANOTHER example of an angle that has been profitable for years, might still be profitable, but which doesn't come up enough for me to care if anyone else uses: if a horse wins its debut in a maiden-claimer by three or more lengths, at odds of 5-1 or greater, and returns within two weeks in allowance company, the ROI is generally very positive. Horses who come close to this criteria also do well. How'd I find the angle? I was filling out a pick-3 ticket at Garden State in 1990, where I took a solid even-money shot in a four-horse field (he won), singled a lukewarm 3-1 shot in a 12-horse field (he won), and took four of the six horses in an allowance race, including the 30-1 winner who met the above criteria. I included him when I asked "what is a maiden-claimer doing in an allowance race," and then thought "no, what WAS an allowance horse doing in a maiden-claimer?" (!!). He won by six, and the pick-three came back $939.00. Notice, however, that I don't talk about value and overlays, but instead how to actually FIND them.

That's all I'm asking of you: how do you FIND your overlays? If the answer is "buy my stuff and find out" then you are marketing, not sharing, and that is held to a very skeptical standard, as it should be.

[RaceIsClosed]

Win bets only? You've GOT to be kidding.

While it is *possible* to turn a profit this way, it's extremely difficult, and I've yet to meet the professional horseplayer who does it.

If you have an edge in the win pool, that edge is going to increase exponentially in the exotics, and that's basic math.

[Overlay]

I started with available statistical data on the performance of various thoroughbred handicapping factors (in my case, in the form of impact values, since they accurately take field size into account as opposed to the use of simple winning percentages; they can be used to compare widely differing factors on a common scale; and they're mathematically easier to work with). I grouped the factors by major categories (as I've said, things like speed, form, class, and pace). When the data had been grouped, I looked for the factor in each category that had shown the closest correlation or the greatest power in predicting the outcomes of races, as reflected by the range of its impact values (the difference between the highest and lowest value for the factor), and by the smoothness of the flow of the values from top to bottom. As a further means of avoiding possible redundancy (even though I had already grouped the data once), I also preferred factors with either a top or bottom value (or preferably both) that passed statistical tests designed to show that the value was so far above or below the expected value (based on the odds of the horses involved) that the factor constituted an independent variable that would keep replicating its performance in the future, rather than a less powerful trend that might currently appear strong, but then might change in effectiveness as time passed. Once I had the most powerful factors identified in each category, I tested them on past-performance data by multiplying the impact values for different combinations of one factor at a time from each category for each horse in a field, to arrive at a composite product for each horse that reflected the influence of all of the factors in the combination. I assigned fair odds to each horse in the field based on the ratio of the individual horse's final product to the total of the products for the entire field. I then checked the odds that were produced against the results of the races to see which combination produced odds that most closely approximated the actual race outcomes. I also performed this procedure on races grouped according to different distance categories and different racing surfaces, so that I could assure that the values were properly weighted to reflect variations in outcome relating to those considerations. After arriving at my final value matrix, I devised a variety of handicapping options based on them, from development of a full-dress odds line, to spot plays, to fair odds for exotic-wager combinations, to a shorthand full-field method that made maximum use of independent variables, so that the handicapper would have a choice of strategies based on available time or personal preference. Since my methodology is designed to produce fair odds that are as accurate as possible, this also allows the use of the disparity between fair odds and tote board odds as a means of optimum sizing of wagers, which I incorporate in my money-management strategy.

Once again, I by no means claim that this is the only way (or even the best way) to handicap. But I know that it has worked for me, and I also know how much confidence I personally gained from having a reliable basis for determining when I was getting the better of the odds. And it is that same assurance, and the improvement in overall performance that it makes possible, that I want to give other players the opportunity to experience. I hope that I have met the intent of your request with this response.

I realize that I have not gone so far in this response as to give the specific factors or impact values that my approaches employ (which may be what you're seeking). However, I think that to do so (given the degree of detail I have already supplied about my methodology) would cross the line into rightfully proprietary information, the further revelation of which I believe I would be (and am) entitled to be compensated for, given the amount of research and effort that has gone into my titles.

[joeyspicks]

RACEISCLOSED:

Win bets only? You've GOT to be kidding.

While it is *possible* to turn a profit this way, it's extremely difficult, and I've yet to meet the professional horseplayer who does it.
.................................................. ......................................

Another foolish, inaccurate statement...........stated as absolute fact ! ? ?:ThmbDown:

[xfile]

Win bets only? You've GOT to be kidding.

While it is *possible* to turn a profit this way, it's extremely difficult, and I've yet to meet the professional horseplayer who does it.

If you have an edge in the win pool, that edge is going to increase exponentially in the exotics, and that's basic math.

Ignorance Is Bliss :D

[RaceIsClosed]

I realize that I have not gone so far in this response as to give the specific factors or impact values that my approaches employ (which may be what you're seeking). However, I think that to do so (given the degree of detail I have already supplied about my methodology) would cross the line into rightfully proprietary information, the further revelation of which I believe I would be (and am) entitled to be compensated for, given the amount of research and effort that has gone into my titles.

Here's the flaw: if you're selling it for $60, or giving it away, that matters little, because if it's profitable you'd be killing your prices and obviously that's not a concern. The net "compensation" would be negative as the prices dropped. Never mind that anyone winning so consistently wouldn't need to sell a thing, or would just sell picks based on the selections, while giving away enough free selections to justify the cost to the consumer.

When you go from "I'm here to share" to "I deserve to be compensated" you become just another guy trolling for customers and not really sharing what helps you win with others.

[Overlay]

Here's the flaw: if you're selling it for $60, or giving it away, that matters little, because if it's profitable you'd be killing your prices and obviously that's not a concern. The net "compensation" would be negative as the prices dropped. Never mind that anyone winning so consistently wouldn't need to sell a thing, or would just sell picks based on the selections, while giving away enough free selections to justify the cost to the consumer.

When you go from "I'm here to share" to "I deserve to be compensated" you become just another guy trolling for customers and not really sharing what helps you win with others.

One of the main points of my previous posts was to explain that the use of fair odds allows the handicapper not to have to worry about prices being driven down on any particular horse, since, if you're betting based on the presence of wagering value, the only way that there would never be any overlays in a race would be if the public assigned each horse in each and every race the odds that corresponded exactly to its actual chance of winning. And, as I've said, although the public is remarkably accurate in assigning odds on a long-term aggregate basis, I regard that degree of precision on its part as a practical impossibility.

I really don't know what else I can say to address your other comments beyond what I have already stated in this thread to explain my motivations and objectives. I am no different in regard to the type and amount of information about my personal handicapping approach that I discuss on this board (especially with the degree of detail that I have gone into on this thread) than any number of other individuals who post here, and who also sell their own products. However, I do want to sincerely thank you for providing me with the opportunity (at your prompting, rather than on my own initiative) to discuss my approach in the manner which this thread has allowed me to do, and to continue to educate others on my main point concerning the advantage of betting based on the presence of wagering value.

[sjk]

One of the main points of my previous posts was to explain that the use of fair odds allows the handicapper not to have to worry about prices being driven down on any particular horse, since, if you're betting based on the presence of wagering value, the only way that there would never be any overlays in a race would be if the public assigned each horse in each and every race the odds that corresponded exactly to its actual chance of winning. And, as I've said, although the public is remarkably accurate in assigning odds on a long-term aggregate basis, I regard that degree of precision on its part as a practical impossibility.



You are saying that you find a playable horse in every race in the win pool. Doesn't leave much room for error.

What about a 2 horse race where you give both an absolutely equal chance to win. Both horses are less than even money. Do you bet on the longer priced one?

[Overlay]

You are saying that you find a playable horse in every race in the win pool. Doesn't leave much room for error.

What about a 2 horse race where you give both an absolutely equal chance to win. Both horses are less than even money. Do you bet on the longer priced one?

I didn't intend my comment to imply that. (Sorry if I was not as clear as I could have been.) One of the advantages of betting for value is that it also allows you to pass races where value is not present, and that is just as important as betting when value is present. While I think it's safe to say that at least some horses in every race will be overlaid in comparison with my fair odds line, the actual winning chances of those overlaid horses may be so small that it makes no sense to bet them as a matter of regular practice. I base wagering decisions on the degree of disparity between my fair odds and the odds established by the public. And even though horses at the high end of the odds spectrum may technically be overlays, they will often not be overlaid enough to justify a wager of any size.

However, I would like to also note that one of the advantages of betting with established probabilities rather than against them is that the more frequently you bet, the greater the chance that long-term probabilities will have to assert themselves to your advantage, just as if you were betting on each flip of a fair coin and were getting better than even-money odds on every flip. Sure, you'd still lose roughly half the time in the long run. But on every one of the other 50% of flips that you won, you'd be realizing a net profit. So, as long as your bankroll and bet size were sufficient to withstand the longest likely runout under those conditions, the more often or the longer you played, the greater the chance that you would come out ahead, and the greater the net profit you would realize, which is the exact reverse of the situation with most wagering games.

In the example of the race you cite, it would certainly be possible for both horses to be underlaid. In that event, I would pass the race, or look for value in the exotic pools, since (according to the way I figure fair odds with a 100% line), neither of the two could have fair win odds below even money if they both had an equal chance to win (and even that would be assuming that all of the other horses in the race had an absolute, literal, zero chance of winning).

[sjk]

I fully agree with the overlay approach but I require that the odds offered be a significant multiple of the fair odds (to account for take-out, imperfections in my odds, changes after the bet is made).

If others were using the same methods as I do and they bet on my overlays I would not be nearly so sanguine that other alternative overlay bets would result. In many cases it would turn a playable race into one that could not be played.

My answer to the question "Why would you be willing to sell a program rather than just use it to win money?" is that I would not be willing to do so.

[Overlay]

I certainly understand the reasoning of any individual bettor who requires an odds premium to justify a wager, especially in light of last-flash downward changes in the odds that have been known to occur. But if you're using a sound quantitative model for your fair-odds assignment, I think you have a firmer basis for pursuing that course of action than if you were assigning odds using intuitive judgment alone. Also, I find that the effect of any inaccuracies that may exist (for example, in betting a horse that is only a slight overlay) is mitigated by the limit which that small odds disparity will impose on the indicated wager size for the horse.

I think its safe to say that, as the odds on one or more horses in a race go down, the odds on one or more others will go up (and vice-versa). That may result in a race becoming unplayable from a win-bet perspective (or it may not). It all depends on the ongoing relationship between the fair odds and the public's odds for each of the horses involved.

[cj]

I certainly understand the reasoning of any individual bettor who requires an odds premium to justify a wager, especially in light of last-flash downward changes in the odds that have been known to occur.

All you said is true, but the above cannot be overstated. At all but the biggest tracks, and even those to some extent, you really don't know what the odds are going to be when the pools close.

At places like Lrl, Tdn, TP, and FL, it is a crapshoot. For the most part if you are betting horses at 5-1 and higher, you will get more, but there are on rules. I've seen 5-1 horses pay $5, and 2-1 turn into 3-5 on the turn. I've also bet horses at 9-1 and gotten back $35.

Betting overlays is not the exact science some would like it to be, even if you have a perfect betting line, which noone does. I think that is why a premium is not only good to have, it is necessary.

[RaceIsClosed]

However, I would like to also note that one of the advantages of betting with established probabilities rather than against them is that the more frequently you bet, the greater the chance that long-term probabilities will have to assert themselves to your advantage, just as if you were betting on each flip of a fair coin and were getting better than even-money odds on every flip. Sure, you'd still lose roughly half the time in the long run.

This is basic math, not anything that will help a player win. He still has to have handicapping methods that will give him the horses with "value."

Too many "experts" in this game will pontificate endlessly about money-management and value as a way of sounding astute while not really saying anything. Of course these things are important, but they are nothing earth-shattering, and certainly nothing worth paying for.

Compare this with someone like CJ, who puts actual ratings on actual horses in actual races. You can debate whether or not his figures are worth the price (I'm sure they are to many), but you can't debate that he is actually producing something that people can use at the track to pick those horses in search of that "value." If you ask CJ how he arrives at his selections, he'll tell you point blank, and he'll even tell you what he likes in some races, before they go off.

[Overlay]

This is basic math, not anything that will help a player win. He still has to have handicapping methods that will give him the horses with "value."

Too many "experts" in this game will pontificate endlessly about money-management and value as a way of sounding astute while not really saying anything. Of course these things are important, but they are nothing earth-shattering, and certainly nothing worth paying for.

Compare this with someone like CJ, who puts actual ratings on actual horses in actual races. You can debate whether or not his figures are worth the price (I'm sure they are to many), but you can't debate that he is actually producing something that people can use at the track to pick those horses in search of that "value." If you ask CJ how he arrives at his selections, he'll tell you point blank, and he'll even tell you what he likes in some races, before they go off.

The whole concept of value is derived from comparison of your idea of a horse's winning chances with the public's estimation, as reflected by the horse's toteboard odds. If the odds that the public assigns to a horse are higher than your estimate of the horse's chances, you have betting value, and if they are lower, you don't. So the questions then become, how do you develop your personal estimate of a horse's winning chances; how do you do so in a manner that will be more accurate on an aggregate basis than the public; and how do you maximize the probability of that estimate remaining valid and consistent from race to race over the long term? For me, the answers to those questions are: by the use of hard statistical data that has demonstrated its durability and accuracy over time, rather than by subjective opinion; and by the use of a blend of multiple properly-weighted fundamental factors, rather than by excessive reliance on any single performance measure. Yes, the combining of those factors is done through the use of basic math. But it is the blend of factors and the weights assigned to them that are the basis for the degree of accuracy of the probabilities and fair odds that are produced.

I have great respect for cj and his work, just as I did (and still do) for the pioneering research done by Beyer in earlier years. In cj's case, he may very well share insights on how he derives his figures. If he decides to do so without cost, that is his prerogative, and I would not second-guess him on that, nor would I presume to expect or demand that he do so. Beyer shared similar insights about his methods in Picking Winners, which told how he generated his figures, and then allowed the handicapper to replicate that procedure on his own. That is exactly what I do with my approach. Beyer rightly put a cost on his information through the price of his book, and so do I. It's all about value-for-value, rather than something-for-nothing. In the final analysis, I'm doing a greater service to bettors by showing them how to derive their own fair odds (just as Beyer discussed the derivation of speed figures), than I would be if I were selling or even giving away fair-odds information piecemeal for specific races. (As the old proverb says, "Give a man a fish, and he eats for a day. Teach a man to fish, and he eats for a lifetime.") And I'm even providing that along with a feature that single-factor handicapping, which normally produces the type of "actual ratings" that you mention, cannot provide, regardless of the ratings' usefulness in highlighting potential winners: the ability to keep applying the same methodology indefinitely into the future while never having to worry about losing profitability. (As powerful as Beyer figures were (and still are) in forecasting a horse's winning chances, is it still possible today to make a profit by betting the top-figure horse blindly to the exclusion of all other considerations (as could undoubtedly be done at one time in the past)? I don't believe so.)

And what are fair odds themselves if not a rating tool, and a means of ranking a field of horses, just as speed figures, pace figures, or any other similar quantitative measure, would be? In fact, fair odds are even more so, because they provide a concrete means of judging when a horse is worth a bet and when it is not, whereas if you bet a top-figure horse without knowing what its actual winning probability is, you wouldn't be able to tell whether you were getting the better of the odds, or not. (And you certainly would not be, after the information you were using became too widely known to the betting public).