02/18/2016 4:16PM

Pandolfo: Using statistical analysis for handicapping


Computer harness handicapper Ray Schell recently published some stats for the year ending 2015 on the paceadvantage.com online racing forum.

Ray is generous with his information. He has a large database and a computer handicapping program he developed. The stats were based on a statistical analysis of over 37,000 races at 33 different harness tracks. Here are some of the stats Ray's Robot came up with for 2015:

Horses dropping in class by at least 3 class points (based on numerical class ratings): 21% wins, 0.821 ROI.

Horses moving up in class by at least 3 class points, 8% wins, 0.606 ROI.

These stats don't surprise me because I know from my own betting that I hit a lot more bets with horses dropping in class as opposed to going up in class. This doesn't necessarily mean that you should never bet a horse that's moving up in class. But, I personally feel that an average handicapper is at a disadvantage when betting horses that are moving up in class.

I'll explain what I mean. Deciding on the class of competition that a horse can win in is tricky. Many bettors look at the class in the past performances and make a snap decision. That isn't good enough. Classification is complex. A horse could be dropping in class but actually facing a tougher field. How could that happen? Well, what if there are four horses in the race that are all dropping in class?

My point is, it takes a skilled and experienced handicapper to evaluate class. And even if you are a very good handicapper, if you're handicapping a race at a track that you don't follow closely, chances are you won't have a good feel for class drops. I handicap all the races at the Meadowlands. Sometimes I'll see a horse dropping out of a particular race that I remember handicapping the week before. As soon as I see the field, it clicks, "Oh, he's dropping out of that tough NW5000 field last Saturday that had three big class droppers in it including two horses that have won over a million dollars." If you don't follow a particular circuit it's hard to make those distinctions.

I've met thousands of horseplayers over the years and sometimes the words that come out of their mouths astound me. Quite a few horseplayers have told me that it's smarter to bet horses moving up in class "because the horse is in form," rather than dropping in class "because horses dropping in class are not in good shape." You often hear simplistic nonsense like this at the racetrack.

Many horses that drop in class are in good enough shape to win, if they find the right field. And just because a horse is in good form doesn't mean that it can step up and beat faster animals. One of the biggest problems with horses moving up in class is that they're usually coming off a win. The public likes to bet winners, so many of these horses are overbet.

Let's look at a few more stats that I found interesting.

Horses coming off a layoff of fewer than 8 days won at 14% with a 0.766 ROI. Horses coming off layoffs between 8 and 15 days or 15-22 days won at 12% and produced an ROI which was 10 points lower, 0.667 on average, a significant difference. Whether a horse was off two weeks or three weeks didn't seem to matter. The key was that horses that were only off one week fared better.

This stat shows that if you bet a horse that's been off a few weeks, you'd better be sure that you're getting excellent value, because the numbers are against you. There are certain situations where the effect of this is mitigated. An example would be if bad weather forced racing to be canceled a few times, and all of the horses missed a start. What you want to avoid is betting a horse that missed a few weeks when it's racing against horses that didn't.

Here are some of the best stats:

The highest win percentage category was 73%, which was for horses that were less than even money on the morning line. The ROI was 0.884. All horses that went off at less than even money won at 57% with a 0.873 ROI.

This trend continued through the first four odds ranges that Ray's Robot tested. That is, horses that went off in the same odds range of their morning line won at a higher percentage than horses that went off at lower odds than their morning line.

For instance, horses that went off at odds between the range of 1-1 to 9-5 won at 36% of the time with a 0.862 ROI. But horses that were between the range of 1-1 to 9-5 on the morning line won at 47% with an ROI of 0.856. Horses that went off at odds of 2 to 3-1 won at 25% but horses that were between 2 and 3-1 on the morning line won at 33%. Horses that went off at odds between 3 and 4-1 won at 19%, but horses that were between those odds on the morning line won at 22% with similar ROI's.

This doesn't surprise me. From my own betting and testing that I've done, both with thoroughbred and harness races, I've found that you have to be careful with horses that get bet down significantly from their morning line. This is even more important at tracks that have a good morning line. Some people think that it's "smart money" coming on when a horse that's 4-1 on the morning line goes off at 8-5. Sometimes it is. But, more often, it's an underlay.

This is another little tidbit I've heard players say at the track, over and over again: "You should bet the horses that are going off below their morning line, that's the smart money coming in"…and, "Don't bet those horses going off higher than their morning line, they're dead on the board and probably not trying today."

Oh, such wisdom. Of course, the exact opposite is true. Yes, sometimes it is smart money and maybe sometimes a horse is cold on the board for good reason. But, over the long run, horses that go off higher than expected are better bets than horses that go off lower than expected.

Another thing I hear people say is "there are no more overlays because everyone has all the info these days." That's absurd. I must have missed something. Since when did picking winners become so easy? If there are no overlays, that would mean everyone has the winner in almost every race. If everyone has "all the info" and it’s so easy, why do I have such a hard time hitting the late Pick 4 at the Meadowlands?

No more overlays? Here's an example: I provide an odds line for my top four contenders on my free Meadowlands picks that are published at www.ustrotting.com. Last Friday (February 12) at the Meadowlands four horses that won went off higher than my odds-line: Pink Pistol 4-1, won at 9-1 odds, Miss Defiance 6-1, won at 10-1 odds, The Summer Wind 7-2, won at 5-1 odds, and Let Her Rock 7-1, won at 13-1 odds.

The most profitable statistic Ray's Robot spit out was horses that were claimed in their last start won at 22% and produced an ROI of 0.906, which was the highest ROI of all of the categories tested. Food for thought.

To find out more about Pandy’s handicapping theories check out his www.trotpicks.com or www.handicappingwinners.com websites, his free picks at handicapping.ustrotting.com/pandycapping.cfm or write to Bob Pandolfo, 112 Michael Ct., Northampton, PA 18067.

[DRF BETS: Sign up for DRF Bets & wager on your favorite harness tracks - Up to a $200 Signup Bonus + Free Harness Eye PPS]