09/11/2013 9:39AM

Bob Pandolfo: Statistical predictors of winning


The word handicapping can have two different meanings. Some will say that we handicap to try and pick the winner. But you could also say that handicapping is about finding the best bet. But either way, you have to try to figure out what each horse's chances are.

[DRF HARNESS: Get Free Harness PPs when you wager with DRF Bets]

In my book, Trotpicks: Modern Harness Handicapping, I made the most important handicapper factor "final time." My two top angles were Key Driver Changes and a Drop In Class. I chose these as my prime handicapping factors based on experience, not statistics.

But there is statistical evidence to back up these theories. Ray Schell is an avid harness bettor who uses a computer program to pick horses. If you have access to the internet, you can find his postings under Ray2000 on the www.paceadvantage.com racing board. Ray recently displayed his top 12 factors for predicting the winner of a harness race. I contacted Ray and asked him if he would mind if I put this in my column. Here is his response, which explains the method used to compile the information:

Dear Pandy,

That's fine with me. I used over 100,000 past performance lines from 2012-2013 with 'logistic regression' as described in CX Wong's book, Precision: Statistical and Mathematical Methods in Horse Racing to get the relative impact of various handicapping factors for predicting Win outcome. I did not look at Return on Wagers, just Win probabilities - Ray

Ray Schell's Top 12 Win Factors
Projected Speed (Final Time)
Class Drop
Driver Change
Driver UDR
Horse WPS Percent
Days Off
Last Race Finish
Morning Line
Average Beaten Lengths
Last Race Odds
Trainer UTR
Post Bias

As you can see, the top-ranked factor is projected speed, which is final time. This should not surprise anyone. It is a "race" so it makes sense that the fastest horse has the best chance of winning. Class Drops and Key Driver Changes are great angles.

Some of you wonder why "trips" is not listed. Well, first of all, a computer has a tough time ranking trips because the past performances don't always tell the complete story.

Even if the computer knew the exact trips, it wouldn't make much of a difference. For instance, say you programmed in all of the trips by hand so you could see which trip is the best win indicator. And say that the horse on the lead is the best trip (which it is at some tracks). Well, you still have to figure out which horse is going to get the lead, and that is often more difficult than it appears.

Sometimes we tend to over complicate handicapping. Handicapping Thoroughbreds is complicated because of all of the different distances and surfaces. But harness racing is much more straightforward.

If a novice contacted me and said that they just started to follow harness racing and needed some handicapping advice, here's what I'd tell them (after I told them to buy my book, of course): 1) Mark each horse's best recent final time in the last 30 days. 2) Circle each horse that's dropping down in class. 3) Circle each horse that's receiving a positive driver change. 4) Eliminate any horse that starts from a post position that is winning at less than 8%.

The next step is to eliminate the slower horses, or horses that don't have a recent final time that appears to be competitive (within 3 to 4 fifths of a second) to the better final times in the race. Now that you've eliminated the bad posts and the slow horses, you have the contenders. Bet the one that appears to be the best value, or if this seems too hard at first, simply bet the contender that is the highest odds.

Then I would tell the novice to learn how to analyze trips and as he or she gains experience, they can use the trips to adjust the final times and fine tune the final selection process.

There are a lot of other handicapping factors, but these are the most important.

Pocono Downs Update: In my last column I wrote about the fast times at Pocono Downs.  Since then, as reported in Harness Racing Update, USTA Executive Vice President Mike Tanner wrote a letter saying that the track was measured prior to the 2010 Breeders Crown. According to Pocono's V.P. of Racing Operations, Dale Rapson, nothing has changed since then. In the same issue of HRU, driver Jim Morrill Jr. said that the bikes they're using now, such as the UFO and Tornado, are faster and "have really good wheels…they have made a difference." But, when asked about the Pocono track, Morrill said, "I think it's a little short." Morrill also said that even ten claimers go fast with ease at Pocono, and added, "I think it has to be a little shy."

In my last column I mentioned two year old filly trotter, Cooler Schooner, who set a world record of 1:51 3/5 at Pocono in her 6th lifetime start. Here's an update: In her next start, during the day time at Harrah's Philadelphia, she won all-out on the lead in 1:54 2/5, a difference of 14 lengths.

In my experience, I've found that when a physical being, human or equine, does something that seems too good to be true, it usually is. But you'll always find people that want to believe.

So far at Pocono this year, at least 16 world records have been set or equaled. Is it too good to be true? Obviously the Pocono track surface is very fast and faster this year than last year. But, horsemen seem to feel it’s safe, which is important. Just for the record, there was a change in track superintendents at Pocono last year (the former track super retired after a long career).

Analyzing this logically, the key could be the start. We have to take into consideration the fact that harness races are started by a moving car. I've seen harness races where the car did not start at the right spot. Does the mobile starting gate always open at the exact same spot? Does it start at the right spot? If it opens a few feet early, the horses would hit the starting point while already in stride, which would explain why the first quarters are so fast. A few feet can make a big difference. If they measure the track again they have to make sure that it's measured exactly from where the horses have been starting, based on video replays and actual live races, not where they're supposed to start. I would also check to see if the pylons are properly placed on the turns.

Personally, who you beat is more important than time. But this doesn't negate the fact that time is a part of the lore and history of the sport. If you research great horses on the internet, horses like Bret Hanover, Dan Patch, Greyhound, Varenne and others, it always mentions the world records they set.

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, 3386 Creek Road, Northampton, PA 18067.

[DRF HARNESS: Sign Up for the FREE DRF Harness Newsletter Today!]