Horse Exclusions Rules for Wagering?

[bigbrown]

What are the exclusion rules you consider useful in wagering ?

I.e., rules that make certain horses undesirable to bet and the selection is focused on the remaining (non-excluded) horses.

[shouldacoulda]

Any rider with less than a 5% win. I've just seen how that ends too many times. I don't care how good the horse is. If the horse is that good compared to the field and I am not getting that warm fuzzy feeling about the jockey then I usually pass the race.

[Overlay]

I tried checklist, elimination-type methods like that, but usually ended up with chalk/low-odds horses that still lost too often to produce an overall profit. I prefer to focus on positive performance aspects or full-field statistical rankings for factors, and look at all the horses in a race, in order to judge when a horse is going off at odds that are higher than they should be.

[Robert Goren]

5 yo mds and NWs of 2. There is no way this can be a good thing. I am not too fond of 4 yo in thoses races either.

[cnollfan]

My exclusion rules are largely based on the conditions of the race, e.g.

Race for winners -- horse off long layoff.
Maiden race -- $1,000,000+ horse.
N2X allowance -- horse that just won N1X allowance.
Claiming race for "non-winners of a race since such-and-such date" -- old horse.
Race for 3 year olds -- horses that have not equalled or surpassed their best 2yo Beyer as a 3yo.
Foreign horse -- horse that hasn't won in the U.S. after a couple tries, unless it was running in graded stakes.

[bigbrown]

I tried checklist, elimination-type methods like that, but usually ended up with chalk/low-odds horses that still lost too often to produce an overall profit. I prefer to focus on positive performance aspects or full-field statistical rankings for factors, and look at all the horses in a race, in order to judge when a horse is going off at odds that are higher than they should be.

Do you have a pointer to the "elimination-type" rules that you've tried? I agree that this methodology often results in getting low odds horses, however once in a while I am hoping to find value after exclusions.

[Trotman]

The horse and if it's in form and placed in the right race and conditions are to me more important than the jockey, after all the horse is doing the running. If a jockey can win 10% or better that's good enough for me.

[Overlay]

Do you have a pointer to the "elimination-type" rules that you've tried? I agree that this methodology often results in getting low odds horses, however once in a while I am hoping to find value after exclusions.

What caused my prices to be so short was the strictness of the elimination rules that I was using, mostly patterned after the standards suggested by Ainslie in his various writings. (I started out in the 1970's going by his early work, like The Compleat Horseplayer, that contained standards such as "two races or two workouts or one of each within the last seventeen days". He loosened up a bit over the years in regard to factors like distance and condition as the game changed, but the basic weakness was still eliminating a horse totally from consideration because of one flaw or performance area that didn't conform to what appeared to be a subjective benchmark, with no indication of how that criterion was derived. Even selection criteria based on statistical studies such as Quirin (for example, "no 'good race' within the last thirty days") were still subject to the same flaw, as long as they were used as go/no-go elimination tools. They would eventually be bet into unprofitability.)

On the other hand, there were admittedly elimination angles that were more effective in spotting likely losers, but that didn't apply to an appreciable number of horses, and/or that still caused the occasional longshot to be tossed. (For me, I guess the personal pain of missing a longshot is greater than the economies of having fewer horses to consider as contenders.)

I switched to a statistical/fair odds-based approach covering an entire field, in order to strike a balance between inclusiveness (maintaining visibility of higher-odds horses); accuracy (providing a consistent basis for deciding whether a horse was worth a wager (at whatever odds)); and rate of return.

[completebill]

I am absolutely convinced that negative handicapping is the wave of the future, and is the best road to profits (obviously the fuundamentals of handicapping can't be ignored). Ken Massa, a handicapping expert, a big winning tournament player, and the "author" of the HTR handicapping program, has volumes of testing and data to prove it.
I have no dog in the fight, but I'm a DELIGHTED user of HTR, a winning player, and I now rely heavily on exclusionary factors, all of which have recently become an integral part of the HTR program.
Go to the HTR website, read everything there, try the free demo, and CALL KEN. He'll be happy to educate you on this facet of the game.

[Overlay]

I am absolutely convinced that negative handicapping is the wave of the future, and is the best road to profits (obviously the fuundamentals of handicapping can't be ignored). Ken Massa, a handicapping expert, a big winning tournament player, and the "author" of the HTR handicapping program, has volumes of testing and data to prove it.
I have no dog in the fight, but I'm a DELIGHTED user of HTR, a winning player, and I now rely heavily on exclusionary factors, all of which have recently become an integral part of the HTR program.

I don't dispute the possible usefulness of negative factors in narrowing a field down to "contenders" (as long as the player can live with the (hopefully few) winners that will be missed in that fashion). But I still have difficulty picturing an elimination system that could narrow race after race down to a single selection (if that's what you're referring to), but that would never become obsolete through overbetting, or at the very least would not have to be revised at frequent intervals to avoid unprofitablity. And, from my personal standpoint, as long as I can develop a full-field betting line in the same time that it would take to apply such a system, I prefer the full-field approach.

Could you elaborate on how your qualifying statement about not ignoring the fundamentals of handicapping affects the negative handicapping process?

[markgoldie]

Some nice replies here on this subject.

ALL handicapping relies on some sort of elimination process. However, as pointed out, if you keep tightening the rules, you wind up with shorter and shorter-priced horses.

The problem in today's handicapping is that the prevalence of black-box programs, particularly those used by the whale groups, are running pp lines through a series of disqualifying filters. No one can possibly make money betting the "perfect" horse, because they are pounded into oblivion.

What the OP is looking for, then, is a way to eliminate (even if from the top position on a gimmick ticket), a well-backed selection. This is difficult, because it attempts to find flaws within the normal black-box filters that are out there.

But since this is what I do, more or less, I can comment (although I have said this before). For me, the most common flaw is the modern over-emphasis on early speed types. In particular, straight E-type horses.

Now, I have also mentioned that you need to properly assess the running style of the horse to be sure that the horse is (or more precisely, in current form is) a true E-type. This means that the horse MUST have the lead to win and WILL NOT pass horses, once outsprinted. Confusion here is immense, because Bris, for example, will routinely call a horse an E-type when in fact the horse is actually an E/P, who leans heavily on his E ability. This, in fact, is why you see so many horses listed as an E who will attend a pace and win late.

Since I don't want to write a book here, let me just say that elimination from top-ticket spots of true E-types who are not likely to establish a clear lead, has proven to me to be the best type of shorter-priced eliminations available.

[bigbrown]

Some nice replies here on this subject.

ALL handicapping relies on some sort of elimination process. However, as pointed out, if you keep tightening the rules, you wind up with shorter and shorter-priced horses.

The problem in today's handicapping is that the prevalence of black-box programs, particularly those used by the whale groups, are running pp lines through a series of disqualifying filters. No one can possibly make money betting the "perfect" horse, because they are pounded into oblivion.

What the OP is looking for, then, is a way to eliminate (even if from the top position on a gimmick ticket), a well-backed selection. This is difficult, because it attempts to find flaws within the normal black-box filters that are out there.

But since this is what I do, more or less, I can comment (although I have said this before). For me, the most common flaw is the modern over-emphasis on early speed types. In particular, straight E-type horses.

Now, I have also mentioned that you need to properly assess the running style of the horse to be sure that the horse is (or more precisely, in current form is) a true E-type. This means that the horse MUST have the lead to win and WILL NOT pass horses, once outsprinted. Confusion here is immense, because Bris, for example, will routinely call a horse an E-type when in fact the horse is actually an E/P, who leans heavily on his E ability. This, in fact, is why you see so many horses listed as an E who will attend a pace and win late.

Since I don't want to write a book here, let me just say that elimination from top-ticket spots of true E-types who are not likely to establish a clear lead, has proven to me to be the best type of shorter-priced eliminations available.

Very interesting rule; is there a documented (e.g., literature) way to classify running styles so that one can use raw data and craft a home-grown categorization instead of the BRIS default one? Or to simply "correct" the BRIS categorization?

[bigbrown]

What caused my prices to be so short was the strictness of the elimination rules that I was using, mostly patterned after the standards suggested by Ainslie in his various writings. (I started out in the 1970's going by his early work, like The Compleat Horseplayer, that contained standards such as "two races or two workouts or one of each within the last seventeen days". He loosened up a bit over the years in regard to factors like distance and condition as the game changed, but the basic weakness was still eliminating a horse totally from consideration because of one flaw or performance area that didn't conform to what appeared to be a subjective benchmark, with no indication of how that criterion was derived. Even selection criteria based on statistical studies such as Quirin (for example, "no 'good race' within the last thirty days") were still subject to the same flaw, as long as they were used as go/no-go elimination tools. They would eventually be bet into unprofitability.)

On the other hand, there were admittedly elimination angles that were more effective in spotting likely losers, but that didn't apply to an appreciable number of horses, and/or that still caused the occasional longshot to be tossed. (For me, I guess the personal pain of missing a longshot is greater than the economies of having fewer horses to consider as contenders.)

I switched to a statistical/fair odds-based approach covering an entire field, in order to strike a balance between inclusiveness (maintaining visibility of higher-odds horses); accuracy (providing a consistent basis for deciding whether a horse was worth a wager (at whatever odds)); and rate of return.

Thanks for the explanation and the literature pointers. Any specifics (e.g., book title) about the "statistical studies such as Quirin" you can point me to? Also any more recent Ainslie book or similar recent literature that treats this topic?

[JustRalph]

Very interesting rule; is there a documented (e.g., literature) way to classify running styles so that one can use raw data and craft a home-grown categorization instead of the BRIS default one? Or to simply "correct" the BRIS categorization?

I dealt with this issue a long time back. I decided that instead of trying to make the home grown analysis on my own, I opted for correcting the Bris data or building in a few rules for just that.

In a related way to the above; I recently went on a binge of trying to ID horses that hang or stop close to the wire...no matter their actual on paper advantages.....I went through a ton of data and several different methods and found that this phenom is often a short lived weakness related to fitness or even age. I found that you can ID these horses just as easy by looking back 2-3 races as you can looking back 10. The pattern is there, use what you see. The problem was I suspected the Bris data to be flawed.

But after a month or so of using it......it seems to be working out ok. The correction I built in, in this case was unnecessary. Therein lies the conundrum of using rules in a blanket fashion.

[Overlay]

Thanks for the explanation and the literature pointers. Any specifics (e.g., book title) about the "statistical studies such as Quirin" you can point me to? Also any more recent Ainslie book or similar recent literature that treats this topic?

Winning at the Races was the specific title by Quirin that I was referring to. Mike Nunamaker updated Quirin's work in the 1990's with titles such as Modern Impact Values and Longshots. Most or all of Nunamaker's studies are out of print or very difficult to find, however (although Nunamaker continues to be an active PA member). Prior to his death in 2007, Ainslie followed up The Compleat Horseplayer in the '70's and '80's with multiple editions of Ainslie's Complete Guide to Thoroughbred Racing, as well as titles such as Ainslie's Encyclopedia of Thoroughbred Handicapping, and Theory and Practice of Handicapping. But all his works were more qualitative than statistical in nature, and also largely (I would imagine) out of print.

Gambler's Book Club (www.gamblersbook.com) has various more specialized current titles such as sire guides, trainer guides, first-time starter guides, and statistical guides for specific tracks, sprinkled throughout their handicapping inventory.

[jamey1977]

Winning at the Races was the specific title by Quirin that I was referring to. Mike Nunamaker updated Quirin's work in the 1990's with titles such as Modern Impact Values and Longshots. Most or all of Nunamaker's studies are out of print or very difficult to find, however (although Nunamaker continues to be an active PA member). Prior to his death in 2007, Ainslie followed up The Compleat Horseplayer in the '70's and '80's with multiple editions of Ainslie's Complete Guide to Thoroughbred Racing, as well as titles such as Ainslie's Encyclopedia of Thoroughbred Handicapping, and Theory and Practice of Handicapping. But all his works were more qualitative than statistical in nature, and also largely (I would imagine) out of print.

Gambler's Book Club (www.gamblersbook.com) has various more specialized current titles such as sire guides, trainer guides, first-time starter guides, and statistical guides for specific tracks, sprinkled throughout their handicapping inventory. And I thought they were getting out the charts and compass and calculator and working for 5 hours. They're just disqualifying , just like me. Horses choke, these chokers are easy to spot. 1 for 17. 1 for 24. 2 for 32. An easy D.Q. I use to have an O for 10 - D.Q. rule, but too many kept winning and these cats Rosario and Bejarano can bring a mule in. Rosario brought in a horse who was 1 for 24. How can one explain this? My conclusion .It is Rosario, himself. Of course when I bet him, he blows it. The worst thing is when we D.Q. a horse and the damn horse wins, but we have to stick by our rules. Change the rules. Our original methods will come through and we would have had a winner. That's why we can't change. Unless, the same damn thing happens 6 times. The D.Q's are risky. I always D.Q. - 8 month layoff horses. An easy pass. But if they win. I Lose.

[Overlay]

And I thought they were getting out the charts and compass and calculator and working for 5 hours. They're just disqualifying , just like me. Horses choke, these chokers are easy to spot. 1 for 17. 1 for 24. 2 for 32. An easy D.Q. I use to have an O for 10 - D.Q. rule, but too many kept winning and these cats Rosario and Bejarano can bring a mule in. Rosario brought in a horse who was 1 for 24. How can one explain this? My conclusion .It is Rosario, himself. Of course when I bet him, he blows it. The worst thing is when we D.Q. a horse and the damn horse wins, but we have to stick by our rules. Change the rules. Our original methods will come through and we would have had a winner. That's why we can't change. Unless, the same damn thing happens 6 times. The D.Q's are risky. I always D.Q. - 8 month layoff horses. An easy pass. But if they win. I Lose.

Your post highlights the point that I made earlier about the reason I switched from elimination/D.Q.-oriented handicapping with its rules, to a probability/value-based approach.

[bigbrown]

I dealt with this issue a long time back. I decided that instead of trying to make the home grown analysis on my own, I opted for correcting the Bris data or building in a few rules for just that.


How did you correct the BRIS categorization?

[jamey1977]

Your post highlights the point that I made earlier about the reason I switched from elimination/D.Q.-oriented handicapping with its rules, to a probability/value-based approach.
Your methods are good. But I hate losing. I can not take it. I want my hit rate at 20 percent winners 16 dollar average. Could be in any sequence. A 9 dollar winner and a 7 dollar winner, as long as I make the total units. 20 percent at 16. I can't take losing 14 in a row, That's why I have to use the elimination. It works on a day with nothing but favorites, I only bet maybe 1 official play because of my D.Q. rules. If I had played overlay, I would have been 0 for 7. Maidens with 14 through 17 starts do come in. The public is dumb, they'll let them go up to 8 to 1. I do bet those now. Now 24 maiden starts, I will pass those. But 2 months ago. A maiden with 24 starts did win. Paid 10 dollars. I would have had to bet 10 to get that 24 start maiden winner.

[markgoldie]

Very interesting rule; is there a documented (e.g., literature) way to classify running styles so that one can use raw data and craft a home-grown categorization instead of the BRIS default one? Or to simply "correct" the BRIS categorization?
The Bris historical designation of a horse is generally correct. That is, at some point in the animal's career, the designation was correct. Problem is, they don't update it regularly in keeping with current form. When an E horse is in very good form, he may be an E/P. That is, he can attend a pace, often falling to third or fourth position in the early going and still pass these horses to win. Conversely, when an E/P is out of form, he may degenerate into a straight E. That is, unless he is able to establish a clear lead, he has no chance of winning. Sharp P types may run more like E/P animals in that they they may actually take an early lead if the pace is rather ordinary. Off-form P's may actually be more like S types in that there is no pace slow enough for them to press. And so on.

So the trick is to convert the horse to its proper level based on recent efforts. When I spoke earlier about weak E types, you want to be certain that NO recent races show the passing of horses after the E2 call. As far as eliminating them- yes, there is a chance that some of these animals will shake clear and win on the front. But if the pace profile seems to be a reasonably contested one, I'll take my chances that he won't shake free and beat me. The odds' value in such eliminations is generally favorable.

[completebill]

I don't dispute the possible usefulness of negative factors in narrowing a field down to "contenders" (as long as the player can live with the (hopefully few) winners that will be missed in that fashion). But I still have difficulty picturing an elimination system that could narrow race after race down to a single selection (if that's what you're referring to), but that would never become obsolete through overbetting, or at the very least would not have to be revised at frequent intervals to avoid unprofitablity. And, from my personal standpoint, as long as I can develop a full-field betting line in the same time that it would take to apply such a system, I prefer the full-field approach.

Could you elaborate on how your qualifying statement about not ignoring the fundamentals of handicapping affects the negative handicapping process?

I'll elaborate. One of the tools in HTR is the Robot, which is used to research the viability/ profitability of various handicapping factors, either alone or in various combinations. Not only can you do the research, but you can create Spot Plays, of various combinations of factors you find profitable, store them, and then automatically print them out each day.

The various factors that can be used are myriad, some common handicapping features such as 1st fraction, 2d fraction, turn time, class, jockey & trainer stats, and some are proprietary to the program.

In building these spot plays, you can research various angles to see if they have a NEGATIVE effect on the play, and you can then incorporate these exclusions into your spot plays. There are dozens and dozens of them.--------Bad pedigree, bad trainer, bad class, questionable layoff, vets list,-----MANY, MANY more--again, some proprietary.

The exclusions can make a losing play a winner, and can GREATLY improve the win % and/or r.o.i. of a play that is already profitable.

[tbwinner]

I began using a point system at the beginning of the year, "assigning" points based on performance of a number of factors (such as last 3 races, speed fig, jockey, trainer, surface, class change, and on). It's really just getting to the point where I have to fine-tune the point values to get an accurate representation of the race in front of me (for example, weight of points of last 3 races versus jockey, currently at a 2:1 weight).