News - Videos - Blogs

Polytrack & Horses for Courses - Part III

Arlington Park Communications | 07/09/2008 #

To lead into this discussion, consider two fans’ posts to my blog. THG 4559 stated, “Polytrack is a unique surface. There doesn’t appear to be any dependable correlation between ability on dirt or turf to that on an artificial surface. Therefore, you need to consider the horse’s previous efforts on Polytrack when handicapping a Polytrack race.” John Drake added, “When it comes to AP Poly, I only use AP Poly races to rate the horse. It is AP’s Poly and there is no other like it.”

THG 4559 later added, “If a horse has no current Polytrack figures, I will use a figure from last year’s AP meet. This is much more effective than using a current dirt figure.”

That raises a very legitimate question: Can we reach back that far? After all, horses can, and do, change over time. Their ability, or talent level, may not be the same from one year to the next.

True. For example, considering physical maturation only, a two-year-old or a three-year-old Thoroughbred should get better (faster) when he becomes a three-year-old or four-year-old, respectively. Some late-developing four-year-olds will even be faster as five-year-olds. Conversely, at some point in his career, an older horse will start down the other side of the slope and run slower from year to year as he ages.

Obviously, an injury can occur at any time to these athletes. Depending on the serious of the injury, the impact on the horse’s ability will vary.

There are many other things that can impact a horse’s demonstrated ability. Changing trainers, moving from weak hands to strong, can help a horse perform better, and the reverse is also true. Medication and equipment changes, some of which we are notified and some of which we are not, can impact performance. Toss in changes in the training regimen, changes in feed and supplements, and more, and we can understand how a horse’s demonstrated talent can be affected.

Does that mean we cannot assume a horse will perform the same at this year’s Arlington meet as he did at last year’s meet? Yes. It is not safe to make that assumption. Does it mean all is lost? No.

First, we can make a qualitative analysis. Look back to last year’s AP races and note where the horse was placed and how he was performing. For example, if the trainer had him running in $25,000 claiming races and he was having some success (winning and/or finishing in the money), he was a legitimate $25K claimer. Can the same be said for this year? If the horse now must run at the $10K level to have some success, he obviously is not the same horse.

I prefer to be quantitative. The more we can accurately (and I stress the word accurately) convert performances to a number, the more exacting we can be in our analysis and ultimately our handicapping. Here are two variations of extrapolation I use. If you attended my workshop at Arlington on June 28, you received some insight into how this works.

This first method is for horses who have made one or two starts at Arlington this meet. Look back to see if a horse progressed in successive starts over the AP Poly in last year’s meet. For example, say the horse ran a 60, then a 62, then a 64 in his third race over the track. Let’s say that this year he is now a seven-year-old and in his first start over the track this year he ran a 50 with a clean trip (no trouble during the race). Will I assume he is going to a run a 62 next time out just because he ran a 62 in his second race over the track in 2007? No. He is not the same horse, talent-wise; he is 10 points (or lengths, by my speed figures) slower. But I will project an improvement of two points since that is how much he improved last year, and predict that he will run a 52 in his upcoming race. I can now compare that prediction to the predictions I have made for the other horse’s in his upcoming race.

Here’s another approach, useful for horses just starting out at the current Arlington meet. Look back to see how fast the horse ran at his 2007 meet prior to coming to Arlington and compare that to how fast he ran at his pre-Arlington 2008 meet. Is he faster or slower, and if so, by how much? Now you can extrapolate how fast to expect the horse to run this year by taking his typical AP 2007 figure and adjusting accordingly. For example, let’s say our horse typically ran 54-56 in dry-track sprints during the 2007 spring Hawthorne meet. This year at HawthorneHawthornHaHaHawthonreHawHaw Hawthorne he typically ran 59-61, or five lengths faster. Now I can take his typical figure from the 2007 Arlington meet and add five points to it in order to predict how fast I think he should run today.

Too many fans simply take a horse’s last race and predict that he will run the same today. Quick, simple, expedient? Yes. Sufficiently efficient and effective? No. We must allow for change. One of those changes can be improvement in successive starts over the track. Now you know how to better quantify that expected improvement.

Join the discussion.