Pace and Cap - Sartin Methodology & The Match Up - View Single Post

alydar_ David · 05-09-2009, 10:41 AM

(posted here, by me on June 29, 2007)

Chris,

Thanks for your erudite post. You have some creative ideas and have generously shared them.

Which paceline to use? Aye, there's the rub.

I think I first asked that question in 1969 and have never found a satisfactory answer. I think I've had this converstation thousands of times and have written about it more times than I care to remember. However, your post has inspired me. Here goes.

I've used almost every program on the market (well, not as many as Jim G ;>) and have employed automatic paceline tools, creating several of them myself with purchased software as well as code I've personally written. All of them have left me wanting.

I recall Michael Nunamaker doing a study. It may have been in Modern Impact Values, I'm not entirely sure. I don't have those stats handy, but I remember the gist of what he discerened. He discovered that using the horse's second best speed figure was more predictive than anything else. Unfortunately, it produced a negative return on investment.

After reading that I scampered around like a busy little beaver and created a program using Phase III type readouts with variant energy. That program selected the second best paceline from the horse's last ten pacelines (assuming all ten races were run on today's track, surface and at today's distance, otherwise there were exceptions and rules made which I won't bother going into here). The results? Disheartening, to say the least.

I've also worked with artifical intelligence programs which select pacelines. They are, as Dick Mitchell said, the wrong technology for horse racing. "A broken clock is better. At least that's right twice a day."

This is not to say one can't successfully implement rules and/or guidelines. Look at Binder. He is like an unwavering pit bull with this "best of last three comparable" paceline guidelines. Granted, Bill deviates from this at times, but he can tell you exactly why and under what conditions he'll go back to a fourth paceline. For example, high class & APV, the last three were sprints and today is a route, etcetera. Bill has had undeniable success with his approach.

Do you remeber Dick Quigley (sp)? He had a set of guidelines where he used different colors which allowed you to go to different paceline zones based on certain criteria. For example, a horse was in the green zone if such and such, and you could only pick a paceline from the red zone under specific conditions. I used this for a while with success. It worked well in the old days when horses didn't run all year long. Frequent layoffs and "freshenings" have rendered his method obsolete.

I've been involved with a couple of statisticians and an actuary on a paceline selection project. We came up with very different paceline selection methods based on the race at hand. For example, at one particular track, for one particular sex and age group, we found that for maiden claimers the best paceline in the entire past performances of the horse was the best to use. In other cases it would behoove (pun intended :>) the bettor to use the best of the last two pacelines.

We spent a lot of time and money on this research and the result was...well, there was no result to speak of; all our findings were statistically insignificant. Time and money down the drain, along with all our high hopes and expectations. Sigh. Those guys threw up their hands and gave up. They left horse racing.

That's not to say it can't be done. Maybe it can. We were not able to do it though.

Personally, I don't know if pacelines "loom off the page" at me, but I sort of do it half by zen now. I don't give it a great deal of thought and allow it to happen.

I'd prefer to know exactly which paceline to use at all times, but that may be something best left to Walter Mitty.

05-09-2009, 10:41 AM	#49
alydar_ David Grade 1 Join Date: Jun 2005 Posts: 2,654	(posted here, by me on June 29, 2007) Chris, Thanks for your erudite post. You have some creative ideas and have generously shared them. Which paceline to use? Aye, there's the rub. I think I first asked that question in 1969 and have never found a satisfactory answer. I think I've had this converstation thousands of times and have written about it more times than I care to remember. However, your post has inspired me. Here goes. I've used almost every program on the market (well, not as many as Jim G ;>) and have employed automatic paceline tools, creating several of them myself with purchased software as well as code I've personally written. All of them have left me wanting. I recall Michael Nunamaker doing a study. It may have been in Modern Impact Values, I'm not entirely sure. I don't have those stats handy, but I remember the gist of what he discerened. He discovered that using the horse's second best speed figure was more predictive than anything else. Unfortunately, it produced a negative return on investment. After reading that I scampered around like a busy little beaver and created a program using Phase III type readouts with variant energy. That program selected the second best paceline from the horse's last ten pacelines (assuming all ten races were run on today's track, surface and at today's distance, otherwise there were exceptions and rules made which I won't bother going into here). The results? Disheartening, to say the least. I've also worked with artifical intelligence programs which select pacelines. They are, as Dick Mitchell said, the wrong technology for horse racing. "A broken clock is better. At least that's right twice a day." This is not to say one can't successfully implement rules and/or guidelines. Look at Binder. He is like an unwavering pit bull with this "best of last three comparable" paceline guidelines. Granted, Bill deviates from this at times, but he can tell you exactly why and under what conditions he'll go back to a fourth paceline. For example, high class & APV, the last three were sprints and today is a route, etcetera. Bill has had undeniable success with his approach. Do you remeber Dick Quigley (sp)? He had a set of guidelines where he used different colors which allowed you to go to different paceline zones based on certain criteria. For example, a horse was in the green zone if such and such, and you could only pick a paceline from the red zone under specific conditions. I used this for a while with success. It worked well in the old days when horses didn't run all year long. Frequent layoffs and "freshenings" have rendered his method obsolete. I've been involved with a couple of statisticians and an actuary on a paceline selection project. We came up with very different paceline selection methods based on the race at hand. For example, at one particular track, for one particular sex and age group, we found that for maiden claimers the best paceline in the entire past performances of the horse was the best to use. In other cases it would behoove (pun intended :>) the bettor to use the best of the last two pacelines. We spent a lot of time and money on this research and the result was...well, there was no result to speak of; all our findings were statistically insignificant. Time and money down the drain, along with all our high hopes and expectations. Sigh. Those guys threw up their hands and gave up. They left horse racing. That's not to say it can't be done. Maybe it can. We were not able to do it though. Personally, I don't know if pacelines "loom off the page" at me, but I sort of do it half by zen now. I don't give it a great deal of thought and allow it to happen. I'd prefer to know exactly which paceline to use at all times, but that may be something best left to Walter Mitty.