Interesting Spot Play

[traynor]

As an example for those who firmly believe that anyone who manages to win without bleeding and suffering through agonizing years of losing is probably doing something illegal, immoral, or fattening, I give you a simplistic example of why you might consider learning a few basic coding skills. The output is not especially pretty, because I ran this search while doing other things.

The output:
0.24 DrvW%20Plus Won 0.24 of 1030 Pace races at GenH ROI 1.05

Translated:
Single factor--the ONLY thing considered is that the driver must have a win percentage >= 0.2

GenH means (US or Canadian) half-mile tracks.

1030 is the number of races in which one (or more--I don't agonize over trifles like whether I should bet on more than one horse in a given race) of the drivers had a win percentage >= 0.2.

Won 0.24 means (to no one's great surprise) that 24% of the wagers (including those in which multiple wagers were made in the same race--that part is irrelevant) won, for a net positive return, (ROI) of 1.05.

What is so hard about that?

(These are current results, from current models. Not something that happened way back when.)

[traynor]

Mile tracks:
1.00 DrvW%20Plus Won 0.21 of 972 Pace races at GenM ROI 1.00

Five-Eighths tracks:
0.95 DrvW%20Plus Won 0.28 of 807 Pace races at Gen5 ROI 0.95

(I switched the sort criteria to sort by ROI rather than win percentage.)

[traynor]

I wouldn't rob the piggy bank or bet the rent money, but you might want to spend a bit less time agonizing over "pace ratings" and "speed ratings" and a bit more time looking at the areas where profits actually exist. Like Woodbine.

1.26 DrvW%20Plus Won 0.23 of 418 Pace races at WDB ROI 1.26

[traynor]

Northfield. Northfield??? Yup, Northfield.

1.06 DrvW%20Plus Won 0.29 of 507 Trot races at Nfld ROI 1.06

[traynor]

At Monticello, look more at the trainers.

1.09 TrnW%20Plus Won 0.31 of 558 Pace races at MR ROI 1.09

[Stick]

How do you define a 20% driver? Year to date ? Previous 12 months? Stats for that track only or combined driving % at all tracks? Just trying to clear it up. Thanks for sharing the info.

[traynor]

Originally Posted by Stick

How do you define a 20% driver? Year to date ? Previous 12 months? Stats for that track only or combined driving % at all tracks? Just trying to clear it up. Thanks for sharing the info.

I keep fairly detailed records for drivers and trainers. Because they are by driver and trainer (rather than by current meet) they would be combined. YTD or previous 12 months would be a little harder to clarify. I don't make the same distinctions, so there are no time references in the data. When starts reach a given number, the ratios are preserved, and a new set of starts initialized, seeded with the ratios as starting values. That is the best process I have found for keeping the data current.

[traynor]

I should have mentioned that (for reasons that should be obvious to most, but perhaps not all) I don't include Woodbine in the generic mile track category. Specifically, the values for generic mile tracks ("GenM") do NOT include Woodbine.

To do so would create a two-fold deception: First, the higher ROI at WDB would distribute over the other tracks, making them appear more profitable than they actually are, and, second, the distribution would effectively conceal the rather generous return available at WDB.

[NorCalGreg]

Originally Posted by traynor

The output:
0.24 DrvW%20Plus Won 0.24 of 1030 Pace races at GenH ROI 1.05

Translated:
Single factor--the ONLY thing considered is that the driver must have a win percentage >= 0.2

GenH means (US or Canadian) half-mile tracks.

1030 is the number of races in which one (or more--I don't agonize over trifles like whether I should bet on more than one horse in a given race) of the drivers had a win percentage >= 0.2.

Won 0.24 means (to no one's great surprise) that 24% of the wagers (including those in which multiple wagers were made in the same race--that part is irrelevant) won, for a net positive return, (ROI) of 1.05.

What is so hard about that?

In your small sample...you've backfitted a subset that's showing a profit. Keep going...you'll dip into the red eventually.

Predicting tomorrow's outcome is where it gets tough.

[traynor]

One of the more useful aspects of studying single factor attributes is that it becomes apparent that when some believe they are "using pace and/or speed" the results may well be the (occasional) correlation of pace/speed with a confounding variable, the existence of which the pace/speed analyst may be overlooking completely.

Specifically, in the rank ordering of predictive attributes, pace/speed (in almost any configuration--WAY more than simplistic speed ratings, average pace, best final times, or other such definitions) rarely is anywhere near the top. Generically, out of the top 10 completely, and at specific tracks, often out of the top 15-20.

Correlations are deceptively misleading. What I suggest is that race analysts dig a bit more deeply, and at least attempt to discover other factors or attributes associated with winners that may be correlated with pace/speed often enough to make pace/speed seem more predictive than it actually is. Not rocket science. No PhD required.

Drivers (and trainers) are an often overlooked area that may be productively studied for the existence of those confounding variables that tend to keep pace/speed attributes on the low end of the profit scale.

[Stick]

I believe I have written about it before and showed some stats about what I call 1st timers. This is a situation where a driver gets up on a horse that he has never been on before in the program lines. Certain drivers just wake up horses. It is generally a good driver but does not have to be the leading driver at that track. At certain tracks betting on these horses at 5-1 or better can be very profitable. Many of them will be bet down but others will not just because of the way people traditionally handicap. Sometimes people are shocked at the track when I bet one of these horses. How can you bet this horse? Do you see his lines? I see so and so is up tonight but he just looks so terrible on paper. It is most effective in cheap races or non winners where horses make great improvements from week to week. As a test mark down every time this happens with every regular driver at different tracks. Note the odds. You will be put on horses that you could never catch handicapping speed and class.

Now the bad news. Drivers change. Their connections and ability changes. You must always monitor this and look for another up and coming driver that may provide value in this situation.

[traynor]

Originally Posted by Stick

I believe I have written about it before and showed some stats about what I call 1st timers. This is a situation where a driver gets up on a horse that he has never been on before in the program lines. Certain drivers just wake up horses. It is generally a good driver but does not have to be the leading driver at that track. At certain tracks betting on these horses at 5-1 or better can be very profitable. Many of them will be bet down but others will not just because of the way people traditionally handicap. Sometimes people are shocked at the track when I bet one of these horses. How can you bet this horse? Do you see his lines? I see so and so is up tonight but he just looks so terrible on paper. It is most effective in cheap races or non winners where horses make great improvements from week to week. As a test mark down every time this happens with every regular driver at different tracks. Note the odds. You will be put on horses that you could never catch handicapping speed and class.

Now the bad news. Drivers change. Their connections and ability changes. You must always monitor this and look for another up and coming driver that may provide value in this situation.

An interesting observation. There is a phenomena in human behavior called "the Hawthorne Effect." Basically, people do better when studied. For awhile. It is not the study itself, or even the people, as much as it is "something out of the ordinary happening." You may have discovered the equivalent of the Hawthorne Effect in horses.

Right off the top of my head, I can think of some really weird results I have seen over the years that might be explained by what you suggest. In essence, a negative driver change is often predictive of an impending win. And at good prices. I thought it was strange, but assumed it was an anomaly, and never dug any deeper. Something to look at when I have a bit more free time. Thanks for the suggestion, and I will post the results when available.

[pandy]

It's possible that a driver could improve a horse, temporarily, for one race, just by driving the horse in a different manner. For instance, the horse could have had a top driver that is very aggressive for several races, then switches to a driver who lets the horse lag back and make one move and the horse wins from off the pace. Or it can work the other way.

One night a few weeks ago at the Meadowlands, on January 7, driver Bill Mann picked up two catch drives because the leading driver, Brett Miller, wasn't there. Mann, who is not regarded as a top catch driver, gunned both horses to the lead and let them roll, gave them their head, didn't try to back down the pace, and both horses went wire to wire impressively, Durant paid $13.20 and Rockin Robert paid $15.00. Mann drove both horses in a different way than Miller had been driving them, and they raced huge.

It's interesting, because on one mile tracks in particular, very few drivers go to the lead and just let the horse roll. This is the way great speed drivers like Luc Ouelette, Carmine Abbatiello, and Walter Case Jr. drove on the front end, they let the horse have its head. Abbatiello told me that speed horses tend to go faster and race better that way. Yet you still see most drivers try to slam on the breaks.

[traynor]

Originally Posted by pandy

It's possible that a driver could improve a horse, temporarily, for one race, just by driving the horse in a different manner. For instance, the horse could have had a top driver that is very aggressive for several races, then switches to a driver who lets the horse lag back and make one move and the horse wins from off the pace. Or it can work the other way.

One night a few weeks ago at the Meadowlands, on January 7, driver Bill Mann picked up two catch drives because the leading driver, Brett Miller, wasn't there. Mann, who is not regarded as a top catch driver, gunned both horses to the lead and let them roll, gave them their head, didn't try to back down the pace, and both horses went wire to wire impressively, Durant paid $13.20 and Rockin Robert paid $15.00. Mann drove both horses in a different way than Miller had been driving them, and they raced huge.

It's interesting, because on one mile tracks in particular, very few drivers go to the lead and just let the horse roll. This is the way great speed drivers like Luc Ouelette, Carmine Abbatiello, and Walter Case Jr. drove on the front end, they let the horse have its head. Abbatiello told me that speed horses tend to go faster and race better that way. Yet you still see most drivers try to slam on the breaks.

Absolutely. Good observation. It is just as evident with thoroughbreds as standardbreds. And just as completely missed by most bettors.

[LottaKash]

Originally Posted by pandy

One night a few weeks ago at the Meadowlands, on January 7, driver Bill Mann picked up two catch drives because the leading driver, Brett Miller, wasn't there. Mann, who is not regarded as a top catch driver, gunned both horses to the lead and let them roll, gave them their head, didn't try to back down the pace, and both horses went wire to wire impressively, Durant paid $13.20 and Rockin Robert paid $15.00. Mann drove both horses in a different way than Miller had been driving them, and they raced huge.

I observed those very two races as well, as they were real eye openers...I had no fat in the fire, so I was happy for the connections, having caught the M1 cadre with their panties down...haha...they should have been considered horses too..