Trainers and Form Points

[andicap]

Lets try and get back to good old fashioned handicapping talk around here.

This idea occured to me today where most of crazy thoughts do -- in the shower.
Most of you might remember "Form Points," the book Tom Hambleton published a while ago that sought to quantify a horse's form. Not really a successful effort but an interesting try. He gave points for horses in winning at today's class, distance, recent workouts, form, etc. It was kind of a pain to calcuate every horse on the card and never caught on. However, I believe it did give a decent gauge on the "form" horses in the race.

What if you calcuated the Form points for winning trainers and tried to see which ones "overachieve" on a regular basis. Which ones don't "figure" to win by conventional handicapping methods but seem to anyway? I suppose you could also determine this by their average odds, but if you looked at HOW the FP were calcuated you might be able to see patterns in each trainer. E.G., Brings in longshots with horses with poor wins at the distance or class or don't get the workout points. It might be a shortcut to linking form to each trainer.

Since you're only calcuating FP for 9 or so horses each day it's not that much added work.

[BillW]

Andi,

Scary, I had the same thought awhile back and have it on my todo list to implement (along with other trainer stuff). Haven't got there yet, but plan to in the next few months. It seems to be a close enough measure of form to dig out those trainers that are achieving "form" by other than classic (or documented as in hidden works) means.

Bill

[JimL]

Andi, I really think you could be on to something. It would seem to follow that if a trainer keeps scoring with low form point total horses, then you know he or she has a ready horse regardless of the pp. I know you are saving the pp of winning horses: wondering if you have detected any patterns? Has it been usefull? JimL

[Dave Schwartz]

You ARE on to something.

In our system we call the "performance factors."

Try building some simple trainer stats based upon things like rank for final time in the last race. The most powerful stuff is how they perform with horses that rank in the back half of the field. It can become an almost absolute elimination.

There are others that can be powerful... such as earnings per start rank, APV rank, etc.


Regards,
Dave Schwartz

[gino]

andicap-

i applaud your creative efforts. the more esoteric your creation the more likely it is to make money.
i've never been real big on trainer/jockey stats, but several years ago while in the thrall of a famous program that only produced 3 numbers, i plugged in the 3 numbers against the world famous Bay Meadows jock colony...things weren't so bad then, Ron Hansen was still alive and the fields had more than 5 horses. anyhow i found all kinds of goofy patterns...all of Baze's winners were like 1-2-2 or 1-1-1 or maybe 2-2-2....while some guy like V. Belvoir was 4-3-3 0r 5-3-3, etc....i didn't wait for 10,000 races, good thing cuz i would have had to throw out all the hansen stats, i just started betting with reckless abandon...
my point is your idea is yours, and it will give you numbers nobody else has, and you just might make some $$$..go for it!!!

The creation of stats nobody has/looking at things from a wacky perspective is the crux of A's G.M. Billy Beane's approach to MLB...profiled brilliantly in Michael Lewis' new book Moneyball...
it's a must read for baseball fans, and i think iconoclastic handicappers can derive some serious inspiration from the methods the A's organization employs to kick ass on a low budget.

gino

[BillW]

Andi,

In looking back in my notes another thought occurred. While you're recording form points, it may also be instructive to record the gap between the winner and the leader in points.

Bill

[VetScratch]

From what I've been told, GP used to be a very fertile ground for trainer angles. Some barns perennially had the inside track while others had no chance even though they might be very good barns in NY, KY, FL, and Canada during the rest of the year. This was back in the 70's and very early 80's (before the PC), and Beyer numbers were golden, but so were trainer angles because they were so laborious to compile.

At GP, in the same era that Beyer celebrates, another group supposedly made much more money by compiling a list of trainers who were successful at GP with either first-timers or layoff horses. They used the DRF result charts on microfiche to build their list and sort it by both trainer name and average payout. One of the members of the original group told me that FTS and layoffs were winning between 18% and 22% of all GP races for almost a decade while paying like they won 8% and 12% of the races. He said their list identified every FTS and layoff winner in their best year, when 20% of the races were won by first-timers or layoffs, including every single payoff over $30.

I suppose the years may have added a little gloss to their memories of a golden era, but my mon's oldest brother trained for two of them and vouches that they had a "very good thing" going for them back in those days. To whatever degree such legends are true... today, too many PCs and databases make it hard to imagine that such a simple angle would hold value for very long.