top 2 contender choices

[raybo]

Originally Posted by thaskalos

The game is not as predictable as some make it out to be; we stand to lose the vast majority of the time...regardless of what handicapping methods we use.

But we often complicate matters by compiling faulty logic on top of the chaotic nature of this game.

Handicapping should not just be an exercise in determining which horse(s) ran the best/fastest races in the recent past. It should be a method of determining what is likely to happen TODAY...and that isn't the same thing.

It has often been written -- and it's often repeated -- that the horse's last race is the best indication of its current form. Well...if that were true, I would have busted out long ago.

There are many reasons to ignore horses whose pps show that they are the "sharpest" in the race...and these reasons have nothing to do with the short prices that these horses are usually offered at.

For example...there are times when the short-priced favorite just jumps off the page due to its last race at the same class and distance...but I eliminate it from my list of win-contenders, and I am seldom wrong. A look at the horse's complete pps reveals to me that the horse's trainer likes to give this horse a "breather" in-between sharp efforts...and what good is a sharp effort last out...if the horse hasn't shown that it can run two sharp efforts in a row.

At other times...I may have spotted a tell-tale sign that one of the duller-looking horses may be in a position to improve markedly from what his recent record shows it is capable of.

Andy Beyer has argued, persuasively, that the "projection" method is better than the "mechanical" method in creating track variants.

And I have found that the projection method is a better way of handicapping...than the mechanical way of expecting horses to reproduce races that they have run in the recent past.

Contrary to popular belief...the horse's last race is NOT the best indicator of its current form...in my opinion at least.

I agree, you have to look at the field, today, and find the advantaged ones. While the last race can be an indication of a horse's form, it is not complete, other things must be considered. I gave up on trying to pick a paceline long ago, so that in itself is contrarian to much of the public.

[traynor]

Originally Posted by thaskalos

The game is not as predictable as some make it out to be; we stand to lose the vast majority of the time...regardless of what handicapping methods we use.

But we often complicate matters by compiling faulty logic on top of the chaotic nature of this game.

Handicapping should not just be an exercise in determining which horse(s) ran the best/fastest races in the recent past. It should be a method of determining what is likely to happen TODAY...and that isn't the same thing.

It has often been written -- and it's often repeated -- that the horse's last race is the best indication of its current form. Well...if that were true, I would have busted out long ago.

There are many reasons to ignore horses whose pps show that they are the "sharpest" in the race...and these reasons have nothing to do with the short prices that these horses are usually offered at.

For example...there are times when the short-priced favorite just jumps off the page due to its last race at the same class and distance...but I eliminate it from my list of win-contenders, and I am seldom wrong. A look at the horse's complete pps reveals to me that the horse's trainer likes to give this horse a "breather" in-between sharp efforts...and what good is a sharp effort last out...if the horse hasn't shown that it can run two sharp efforts in a row?

At other times...I may have spotted a tell-tale sign that one of the duller-looking horses may be in a position to improve markedly from what his recent record shows it is capable of.

Andy Beyer has argued, persuasively, that the "projection" method is better than the "mechanical" method in creating track variants.

And I have found that the projection method is a better way of handicapping...than the mechanical way of expecting horses to reproduce races that they have run in the recent past.

Contrary to popular belief...the horse's last race is NOT the best indicator of its current form...in my opinion at least.

No matter how the data is tweaked and massaged, ultimately the selections of a computer-based methodology are going to make comparisons based on history--what has already happened--rather than what is going to happen.

The classic example is two horses dueling to the wire to finish a head apart. Computer apps tend to "evaluate" those two as nearly identical, and "project" a similar scenario in today's race. In the real world, it is far more typical for one horse to have been maxxed out and require a cooler race or (relatively long) layoff to rest and recuperate, while the other attains peak conditioning from the intense exertion and is ready-to-run-back-in-a-best-ever effort within days. The skill of race analysis is in determining which is which.

As Quirin pointed out many years ago, "If a horse's best effort was in its last race, that is when you should have bet on it."

As you point out above, if a horse runs a Big Number in a race, what evidence is there that it is capable of repeating or exceeding that effort today? (I don't buy the "bounce" explanation for "swings in the form cycle." It is way too simplistically applied as an excuse for losing horses that should have been expected to lose by handicappers who were not daydreaming.) Form cycles are a fact of life in racing, but nowhere even remotely as neatly defined and easily predicted as some like to believe.

Ainslie may have believed that one should never expect a horse to do something it has not done before, but that seems to be where the money is. Good post. Good advice.

[HUSKER55]

hey Ray,....me too!

good advise!

[pandy]

Originally Posted by traynor

As Quirin pointed out many years ago, "If a horse's best effort was in its last race, that is when you should have bet on it."

.

I use this type of thinking. Today in the Belmont there are two horses that I bet and gave out as Best Bets last time, Oxbow and Freedom Child. But I'm not betting either today because last start it looked to me that they were both ready for big efforts, and they both ran career bests. Most of the time horses, and athletes in general, regress off this type of effort and the bottom line is, the time to bet them was last race.

[DeltaLover]

I think that as far as winning probabilities alone are going the last race of the horse is close to neutral as a prediction factor. But considering the way the crowd perceives it, I agree that it turns to a negative expectation.

As far as Oxbow, I have to disagree with the bounce theory followers.
In the begining I was tossing him well. Studing the race closer, I admit I have changed my mind.

I am expecting a very strong race from him today. His problem remains the distance as he seems to prefer shorter despite his breeding who is suggesting quite the oposite.

[pandy]

Career Tops are more likely to create a bounce effect if the horse was used hard or ran dramatically faster than any of his prior starts and Oxbow had a pretty easy trip, so I do think he could race well today. But, at the shorter odds, longer distance, different trip, he is probably not worth a bet. Plus there is the possibility that he will regress wheeling back two weeks off a career best. And on top of that, over the long run this is a losing scenario, i.e.--betting on horses that shake loose on an easy lead and win at long odds in their next start at shorter odds.

[pandy]

A horse that comes off a career best that seems unlikely to bounce today is WABBABACK in the 6th at Belmont. I bet him last time at 2-1 in a 5 horse field and he won but he had what Voegele referred to as a "smooth" trip. This horse ran in that huge key race at Gulfstream two back and he finished IN FRONT of Freedom Child, who came back and won the Peter Pan.

[traynor]

Originally Posted by pandy

I use this type of thinking. Today in the Belmont there are two horses that I bet and gave out as Best Bets last time, Oxbow and Freedom Child. But I'm not betting either today because last start it looked to me that they were both ready for big efforts, and they both ran career bests. Most of the time horses, and athletes in general, regress off this type of effort and the bottom line is, the time to bet them was last race.

I agree completely. I also think that determinations of when a horse is likely to improve, when it is likely to replicate, and when it is likely to decline from a particular race are among the most profitable areas of horse racing to research and analyze.

[erikeepper]

Originally Posted by traynor

I also think that determinations of when a horse is likely to improve, when it is likely to replicate, and when it is likely to decline from a particular race are among the most profitable areas of horse racing to research and analyze.

This is the final step of my handicapping process and the most important! I use negative angles to eliminate "bad bets", then I have my main contenders, then I analyze those contenders and determine which are the most likely to improve/decline/stay the same and why. Then I check the odds and make my decisions.

Regression to the mean, or "to the horse's notional mean", is a wonderful concept in betting

[PIC6SIX]

Simply, bet the same amount on both provided the odds are usually split like you say at 5/2 and 5/1.

[Overlay]

Originally Posted by PIC6SIX

Simply, bet the same amount on both provided the odds are usually split like you say at 5/2 and 5/1.

Would it be preferable to dutch them in a situation like that (so that you would get approximately the same total return in case either of them won), or (for a fair odds-line bettor) to size bets for each according to Edge/Odds (or a uniform percentage thereof), rather than betting the same amount on each?

[traynor]

Originally Posted by Overlay

Would it be preferable to dutch them in a situation like that (or, for a fair odds-line bettor, to size bets for each according to Edge/Odds (or a uniform percentage thereof)), rather than betting the same amount on each?

That can only be determined by accurate, detailed records of one's selection and wagering processes. That is, accurate, detailed records will indicate the (relatively) exact probability of of each odds range winning, and that information can be used to determine when (and if) dutching is more or less profitable than other wagering strategies.

[Robert Fischer]

If you have 2 contenders you pass, unless they are high enough odds that you can bet them both*.

Same goes for "x" number contender(s).

*The exception to this rule, is when one or more of your contenders is an underlay or overlay. In these cases you pass the underlay, and bet the overlay.

All of this is moot if your value estimate is poor.

If you are getting a true .92ROI, then you have to be .08 more conservative overall on your value estimate in order to break even.

This game is extremely difficult. however, principles like formulating a value estimate can be tackled by serious players and weekend warriors alike.

[Robert Fischer]

Originally Posted by Overlay

Would it be preferable to dutch them in a situation like that (so that you would get approximately the same total return in case either of them won), or (for a fair odds-line bettor) to size bets for each according to Edge/Odds (or a uniform percentage thereof), rather than betting the same amount on each?

If you have 2 contenders that you plan to bet you will dutch them.

The only reason to have 2 betable contenders is that both horses are showing a positive return on your value estimate.

Dutching is enough in this situation. It's unlikely that your value estimate will be accurate enough to do "adjusted dutching" sized precisely to the value estimate. More than dutching is going to be overkill.

[Robert Fischer]

Originally Posted by Robert Fischer

If you have 2 contenders you pass, unless they are high enough odds that you can bet them both*.

Same goes for "x" number contender(s).

*The exception to this rule, is when one or more of your contenders is an underlay or overlay. In these cases you pass the underlay, and bet the overlay.

in which case you don't really have 2 contenders

you have 1 contender