Why use speed figures at all?

[Capper Al]

First of all, let me say that I may joining this forum as a regular member. I haven't figured out your vcash yet, but that will come in time.

Now about speed. Beyer, Free, and many more authors have documented that when a horse wins they usually improve by 9 or more speed points. This means the handicapper has to guess which horse is going to jump in speed making past speed unreliable because the horse with the highest speed figure might not necessary be the same one that improves in today's race. Just think about the numbers and forget that they are speed figures for a while. If a field of 6 horses has the following speed:

A: 89 B:90 C:84 D:92 E:86 F: 87

The lowest one C at 84 could win if the other horse ran back to their regular number and C improved by the statistical improvement of 9 points or more. C would rate a 93, 84 + 9. The variance is too much to get a handle on what's about to happen in the race. The contenders, generally, are close in their figures anyway making it even harder to use speed.

The best that speed might be able to do is act as an elimination. If a horse isn't within 10 or 12 points of the top horse, eliminate him.

[Cangamble]

You are assuming that they all have an equal chance to win. The horse with the highest number most likely has the highest chance of winning on average.
Probably the second or third highest horse in your example provides the best ROI.
Horses sometimes improve by 25 points, but that doesn't mean we should throw away our Forms and just use a Ouija Board.

[PaceAdvantage]

Originally Posted by Aspiteri

I haven't figured out your vcash yet, but that will come in time.

vCash is the Monopoly money we use to bet with in our free contests at P.A. Downs

[kenwoodallpromos]

With Beyer's method, points are also given depending on the level (class) of the race; so if a horse drops for a win, it may not need to improve its speed figure at all. For non-claiming races I think your statement should be true for most races, especially for a 2-4 year old.

[Overlay]

Originally Posted by Aspiteri

This means the handicapper has to guess which horse is going to jump in speed making past speed unreliable because the horse with the highest speed figure might not necessary be the same one that improves in today's race.

I think the keys to addressing this are:

1) considering an average of speed figures from multiple recent races;
2) dealing with the relative rankings of the horses with regard to their competitors, rather than concentrating on a comparison of the speed figures themselves;
3) viewing the situation in terms of the probabilities associated with the various rankings (as part of a general approach of betting based on value), rather than as a guarantee that any one horse will win today; and
4) not treating speed as a stand-alone element, but giving it its proper weight in conjunction with consideration of other major handicapping factors such as condition and class.

I've found that, when treated in this manner, speed figures continue to have predictive reliability and significance.

[rastajenk]

I'm always amused by the anti-figurists and the way they create situations that prove their point in advance, ignoring all the other possibilities that could co-exist with their situations, or could exist in real life. In other, off-topic realms, these are known as straw man arguments, because they are as flimsy as a scarecrow.

Using the numbers in the original post, I would start off by assuming that B and D could be fit and ready enough to win, but I couldn't say D is clearly superior to B, or for that matter that B is superior to A. Nor could I eliminate any of them just off a single, last-race number. There is so much more to it than that.

[pandy]

I forget the actual stat, but I think that 80% of all races are won by a horse with one of the top 4 Beyers. Your point spread of 10 to 12 might be about right also, so for terms of eliminating horses, you could do worse. The bottom line is, if you bet on a horse that's a longshot (say 8-1 or higher) that recently ran one of the top 4 Beyers ( or Bris) or within 12 points, you have a much better chance of cashing than if the horse doesn't fall into those parameters.

[classhandicapper]

Originally Posted by Overlay

I think the keys to addressing this are:

2) dealing with the relative rankings of the horses with regard to their competitors, rather than concentrating on a comparison of the speed figures themselves;
3) viewing the situation in terms of the probabilities associated with the various rankings (as part of a general approach of betting based on value), rather than as a guarantee that any one horse will win today; and

Good insight. I think rank is probably often not appreciated enough.

If a horse is only 7 Beyer points slower than fastest horse, but there are 6 horses faster than him, he probably doesn't have as good a chance of winning as a horse that is 10 points slower, but who is the 2nd fastest horse. In the latter case, all that horse may need is a sub par effort or very bad trip from one horse to become the most logical winner.

[OTM Al]

Plus it is all about the context in which the fig was earned. I always discount big fig career tops that won by open lengths for example as horses that get loose on the lead get brave. Even Andy B's books will tell you that you need to consider the trip and the way the figure was earned. It seems that most critics of figs think they are some black box guide to picking a winner. Sorry to break the news to you, but......

[Overlay]

PA:

I see you contributed to this thread already, but might it not get exposure to a potentially bigger target audience over on the Handicapping board (among those who might not check the General Racing Discussion board as regularly or as thoroughly), since it does deal with an aspect of handicapping?

[Capper Al]

Thanks for all your comments. And believe it or not, in your own way most of you seem to agree. The fact that we need other attributes to figure a race only speaks to the weakness of speed. I especially agree with handling speed different for claimers and non-claimers.

The point still stands that when we read the PPs, and now know one of the horses will jump 9 or more points, makes the speed figures we are looking at of less value. And yes, I agree top speed last out has to be one of the most impressive stats out there, about 30%. Yet, it is seen by everyone and bet down. And yes, one of the top four speed figures will win about 80% of the time seems right. But also one of the trackman's top four will win about 80% of the time also. So of what value is speed? We could substitute one of the trackman's top picks for the top speed picks with the same results.

[eddie10]

you also have to consider early speed and best closing speed as a factor .etc

[Mike at A+]

What might also be interesting is to analyze the percent of horses that run increased speed figures from their last race in a typical race. Intuition would tell me that it is probably somewhere near 50% because there are only two possibilities (excluding a tie). A horse can run higher or lower. And if he runs lower, his chances of running higher in his next race are that much better.

That being said, in the original example, I'd need pretty attractive odds to put any money on "C" because even if he does improve, one would expect 2 or 3 of the 5 above him to also improve. But then you'd also have to consider what those 5 horses did in their last race compared to the next to last. If all 5 improved, the chances of a lower rating are that much greater.

Perhaps a moving average with some consideration given to standard deviation should be used.

[garyoz]

Originally Posted by Aspiteri

Thanks for all your comments. And believe it or not, in your own way most of you seem to agree. The fact that we need other attributes to figure a race only speaks to the weakness of speed. I especially agree with handling speed different for claimers and non-claimers.

It is all about form cycle not a specific speed figure. This whole topic has been dealt with in many forms on the PA Board. For example see this recent thread on Len Ragozin's book. http://www.paceadvantage.com/forum/s...ad.php?t=38163. There are many other posts on this topic. It isn't that simple.

[PaceAdvantage]

Originally Posted by Overlay

PA:

I see you contributed to this thread already, but might it not get exposure to a potentially bigger target audience over on the Handicapping board (among those who might not check the General Racing Discussion board as regularly or as thoroughly), since it does deal with an aspect of handicapping?

Excellent catch. You are 100% correct.