How to determine a track bias

[Cratos]

Originally Posted by classhandicapper

Argue was a poor choice of words.

I was not disagreeing with you.

On the issue, I may have a different perspective than you because I've spent most of my years observing US dirt racing. US dirt tracks have different compositions, different bases, different depth, the moisture content is changing all day long as they water then and then it evaporates, the surfaces are not uniform, they seal and unseal the tracks, riders make pace adjustments etc..

I've seen so many patterns to the way races develop on dirt and there are so many possibilities, I think it's all too complex to pin anything down as to cause.

I'm moving further away from theory and more towards my ability to analyze the horses in a race and the results well enough to gain an edge (what actually happened).

Given the quality of each of these horses, how the race figured to develop, how the race actually did develop, do these results make sense or do they seem biased in some way due to some combination of the track/pace/path biases?

I'm taking the same approach to race analysis that many speed figure makers take with the track variant. There are several components to how fast the race will be (surface, wind, etc..), but rather then breaking it down into individual components, they/I capture it all in one adjustment.

As you move away from theory; what is your basis for quantification beyond conjecture?

[cj]

Originally Posted by bobphilo

I agree that that in slow paced turf races they come down to 2 furlong sprints to the finish so the value of a length is the same regardless of distance. However, in U.S. dirt races the opposite is usually the case. The early pace is much faster so that everybody is decelerating at the end and the effect of total distance run matters.

When an engineer is trying to figure out the area of a space he doesn't say, "I'm going to ignore math and base it on what my apparent observations tell me". Empirical observations are fallible, math is not.
Like Zeno said, "There is the way of truth and the way of opinion."

The numbers Beyer uses for values at different distances were chosen pretty arbitrarily, not by any scientific methods. Same goes for me and BRIS and TG too. But eventually mine will be.

[bobphilo]

Originally Posted by cj

The numbers Beyer uses for values at different distances were chosen pretty arbitrarily, not by any scientific methods. Same goes for me and BRIS and TG too. But eventually mine will be.

When I made my figures I based my different values for each length based on the different percentage they represented of the total distance run and these were constant and they proved accurate.
According to Beyer, he bases his on differences in times run at different distances which can vary widely.
I suspect your studies will conform to the more stable relationships that I use.
Of course, slow paced turf races are an exception, which we both understand.

[classhandicapper]

Originally Posted by Cratos

As you move away from theory; what is your basis for quantification beyond conjecture?

I have a classing scale that generates a "number" that tell me how strong a field was based on the class label and who was actually in the race. That scale is the starting point from which the analysis takes place. It's subsequently colored up/down subjectively by the horses' trips within their races and other factors.

The difference between what I am doing now and what I've been doing for a long time it that I'm not trying to quantify each component of the trip (pace figures, ground loss, any bias, how competitively the race was run, trouble etc..) individually. I'm focusing more on how well the horses ran in relation to each other within that race given their trips.

I still see that the fractions were fast/slow, that the track was tilted one way or the other, that the final time was fast or slow, that there was a battle etc.. but the conclusions are based on what happened instead of what theoretically "should have" happened.

It's a subjective process.

[Cratos]

Originally Posted by classhandicapper

I have a classing scale that generates a "number" that tell me how strong a field was based on who was in the race. That scale is the starting point from which the analysis takes place. It's subsequently colored up/down subjectively by the horses' trips within their races.

The difference between what I am doing now and what I've been doing for a long time it that I'm not trying to quantify each component of the trip (pace figures, ground loss, any bias, how competitive;y the race was run, trouble etc..) individually. I'm focusing more on how well the horses ran in relation to each other with that race given their trips.

I still see that the fractions were fast/slow, that the track was tilted one way or the other, that the final time was fast or slow, that there was a battle etc.. but the conclusions are based on what happened instead of what "should have" happened.

It's a subjective process.

Not speaking for you, but you are moving to become more of a "qualitative" handicapper as oppose to a "quantitative" handicapper; am I correct?

[cj]

Originally Posted by bobphilo

When I made my figures I based my different values for each length based on the different percentage they represented of the total distance run and these were constant and they proved accurate.
According to Beyer, he bases his on differences in times run at different distances which can vary widely.
I suspect your studies will conform to the more stable relationships that I use.
Of course, slow paced turf races are an exception, which we both understand.

Maybe turf races aren't really an exception, but the way all races should be rated?

How do you determine the value of a percentage of the race. Is one second 1/10th of a point? One point? 10? 100? That is the arbitrary part I was talking about.

[Dave Schwartz]

CJ,

The numbers Beyer uses for values at different distances were chosen pretty arbitrarily, not by any scientific methods. Same goes for me and BRIS and TG too. But eventually mine will be.

Would you be willing to talk about these values a little?

I certainly know about Beyer and BRIS, but nothing about yours.

Would love to hear more.


Regards,
Dave Schwartz

[classhandicapper]

Originally Posted by Cratos

Not speaking for you, but you are moving to become more of a "qualitative" handicapper as oppose to a "quantitative" handicapper; am I correct?

I was always very comparative class oriented because I started before good figures were even available. I'm pretty good at that. I always tried to incorporate pace and final time numbers into my thinking to go along with making track biases. Doing that, I would see some races where the numbers were saying one thing, but that wasn't what I thought happened based on the rest of my analysis. That makes things difficult. I think I'm getting better at doing a subjective post race analysis from just looking at field makeups, watching races, and analyzing charts. I think that also takes me away from consensus thinking.

[bobphilo]

Originally Posted by cj

Maybe turf races aren't really an exception, but the way all races should be rated?

How do you determine the value of a percentage of the race. Is one second 1/10th of a point? One point? 10? 100? That is the arbitrary part I was talking about.

Generally turf races are run very differently than dirt races, though in the rare cases where there are, I understand the exception.
In my figure making I used the standard that 1 point differences would be of some significance and 2 points or more to be very significant. This is what Dr Quirin found in his excellent study (I strongly urge everyone to look at his books based on his studies).
In other rating systems, including Beyer and TFUS (no offense) I find these differences are exaggerated leading to a false sense of security of the difference between horses.

[Cratos]

Originally Posted by bobphilo

Generally turf races are run very differently than dirt races, though there are exceptions.
In my figure making I used the standard that 1 point differences would be of some significance and 2 points or more to be very significant. This is what Dr Quirin found in his excellent study (I strongly urge everyone to look at his books based on his studies).
In other rating systems, including Beyer and TFUS (no offense) I find these differences are exaggerated leading to a false sense of security of the difference between horses.

Are you sure?

[thaskalos]

Originally Posted by cj

The numbers Beyer uses for values at different distances were chosen pretty arbitrarily, not by any scientific methods. Same goes for me and BRIS and TG too. But eventually mine will be.

Beyer's method of assigning a greater deduction to a length in a sprint than to a length in a route is readily understandable...and it is easily realized when related to human track events. A high school sprinter is capable of running within a second of Usain Bolt's 100-yard world record...but only a world-class miler could run within a second of El Guerrouj's mile world record time.

The fault in Beyer's thinking, IMO, was in his determination that a horse's sprint figures should come into mathematical alignment with a horse's ROUTE figures...as if horses are equally competent in running at both these distances. Beyer sought to "equalize" these differences so that an 80 in a sprint is equal to an 80 in a route...or an 80 at Saratoga equals an 80 at Charles Town. THAT was his mistake...IMO.

[cj]

Originally Posted by bobphilo

Generally turf races are run very differently than dirt races, though in the rare cases where there are, I understand the exception.
In my figure making I used the standard that 1 point differences would be of some significance and 2 points or more to be very significant. This is what Dr Quirin found in his excellent study (I strongly urge everyone to look at his books based on his studies).
In other rating systems, including Beyer and TFUS (no offense) I find these differences are exaggerated leading to a false sense of security of the difference between horses.

I've read all the Quirin books, learned a ton. I'm just not as sure the values he came up with and that you used are as good as they could be.

[cj]

Originally Posted by thaskalos

Beyer's method of assigning a greater deduction to a length in a sprint than to a length in a route is readily understandable...and it is easily realized when related to human track events. A high school sprinter is capable of running within a second of Usain Bolt's 100-yard world record...but only a world-class miler could run within a second of El Guerrouj's mile world record time.

The fault in Beyer's thinking, IMO, was in his determination that a horse's sprint figures should come into mathematical alignment with a horse's ROUTE figures...as if horses are equally competent in running at both these distances. Beyer sought to "equalize" these differences so that an 80 in a sprint is equal to an 80 in a route...or an 80 at Saratoga equals an 80 at Charles Town. THAT was his mistake...IMO.

Again...I know this. I've said it many times. It is the amount I'm questioning.

[thaskalos]

Originally Posted by bobphilo

Generally turf races are run very differently than dirt races, though in the rare cases where there are, I understand the exception.
In my figure making I used the standard that 1 point differences would be of some significance and 2 points or more to be very significant. This is what Dr Quirin found in his excellent study (I strongly urge everyone to look at his books based on his studies).
In other rating systems, including Beyer and TFUS (no offense) I find these differences are exaggerated leading to a false sense of security of the difference between horses.

The difference, of course, being that Quirin was talking about FIFTHS of a second when he was trying to determine "significance"...and, in your arguments, you are talking about ENTIRE seconds. You made the comment that you felt a one-second difference at the half-mile of a sprint would have a NEGLIGIBLE effect on the horse's final time...which is a comment that would give William Quirin a HEART attack.

[thaskalos]

Originally Posted by cj

Again...I know this. I've said it many times. It is the amount I'm questioning.

There can BE no "right amount". The entire "figure equalization" idea is flawed. Some horses like to sprint, and some horses like to run a longer distance of ground. The competent handicapper KNOWS that a sprinter's figures can't be trusted in a route race...and vice versa. And we all know that an 80 speed figure at Monmouth isn't the same as an 80 at Belmont.

This tinkering with the figure point values is an effort to smooth-out the rough edges of a very complicated sport...and that is a futile undertaking...IMO.