Quote:
Originally Posted by Jeff P
And:
A few days ago, in post number 82 in one of the Handicapping Library threads, you posted a link to a PDF written by Bill Benter.
As a result I reread the PDF in its entirety.
I've probably read that paper at least 20 times over the years.
Every time I read it I manage to pick up a nugget - some little detail I somehow missed during all of my previous readings about the way Benter was doing things when he wrote that paper.
Why does it have to be one or the other?
Why is it ok to use Tote data but not ok to use past performance data?
It seems pretty obvious to me that Benter created a successful model.
Based on what he wrote - it seems pretty obvious that Benter used BOTH past performance data and tote data in his model.
Beginning on page 184 Benter penned a chapter on "HANDICAPPING MODEL DEVELOPMENT" and specifically mentioned a handful of factors based on past performance data that he used in his model including "normalized times of past races."
What are speed and pace figures? Are they not simply a way of expressing normalized times from past races?
-jp
. |
Regarding Benter’s paper I too have read it many times over. I even went so far as to post it on PA and high-lighted many items that I believed were for serious consideration. I’m not sure if many realize this, but Bill B. had no background in horses or handicapping them whatsoever. (Anyone can read about his background and eventual involvement with horseracing in Hong Kong by reviewing his recent published interview) He of course developed an extremely detailed program which according to what I read ultimately used over 130 race data variables. He certainly did not exclusively rely on Speed figures (which is what this thread is all about) in his model. Yet, he admittedly claims that without incorporating the tote data his overall profit margin would have been comparatively reduced. He also stated the following:
Quote:
It can be presumed that valid fundamental information exists which can not be systematically or practically incorporated into a statistical model. Therefore, any statistical model, however well developed, will always be incomplete.
An extremely important step in model development, and one that the author believes has been generally overlooked in the literature, is the estimation of the relation of the model's probability estimates to the public's estimates, and the adjustment of the model's estimates to incorporate whatever information can be gleaned from the public's estimates.
|
From my personal perspective “it is ok to use Tote data, but not ok to use past performance data” for the following reasons:
A) PP data is limited to providing dated factual data about a horse’s performance at specific race conditions while competing against a limited number of entrants.
B) Any detailed analysis of PP data (be it Speed figures or otherwise) is open for subjective interpretation which inevitably results in a variety of deductions.
C) No conclusions can be drawn from the PP’s about a horse’s current mental or physical condition.
D) The interpretations of the PP’s can very often lead to over bet and under-valued selections because critical assumptions are made about a horse’s competitive nature.
E) A detailed review and final interpretation of analyzed PP’s can become cumbersome, tedious and time consuming.
I won’t bore you with my feelings about the tote analysis because I’ve stated it many times before. I will say that I’ve tried both methodologies simultaneously and it became counter-productive.
Here are a few Excerpts (as related to tote information) from:
“
Computer Based Horse Race Handicapping and Wagering Systems:” Bill Benter.
A)
Quote:
|
Certain authors have convincingly demonstrated that profitable wagering systems do exist for the races. The most well documented of these have generally been of the technical variety, that is, they are concerned mainly with the public odds, and do not attempt to predict horse performance from fundamental factors.
|
B)
Quote:
|
There will always be a significant amount of 'inside information' in horse racing that cannot be readily included in a statistical model. Trainer's and jockey's intentions, secret workouts, whether the horse ate its breakfast, and the like, will be available to certain parties who will no doubt take advantage of it. Their betting will be reflected in the odds.
|
