Ask Beyer: Pars can only be published when there is confidence in the data
Q. Why do so many races in DRF have no Beyer Pars listed for them?
– Tom Brown
Q. Could you please explain the criteria for when you decide to publish a Beyer Par for a class designation in the paper or the Beyer Par Report?
– Wayne Crimi
A. We publish pars for any class where we have data in which we are confident – but there are many classes where we don’t. Most of these are races restricted to 2- and 3-year-olds. Thoroughbreds generally improve as they age, until they reach the age 4. In the month of June, the average 2-year-old maiden special weight race at Santa Anita has produced a winning figure of 72. In October, the average has been 79. (This is based on data from the last four years.) It’s not practical to try to generate month-by-month pars for 2- and 3-year-olds.
The other issue complicating pars involves the conditions for races. In a bygone era, race conditions used to be pretty straightforward: $10,000 claiming race. Allowance for non-winners of two races. But now, in their effort to attract large fields, racing secretaries regularly write races with multiple conditions of eligibility, like this one at Gulfstream: “For 3 year-olds and up which have never won 3 races, or which have not won a race on the turf since Oct. 31, or 3-year-olds.” Good luck making a par for that race.
– Andrew Beyer
Q. How many races for a particular class level does it take to determine an average winning par at an individual track?
– Brian Goodspeed
A. We try to list pars for as many classes as possible, and often we’ll use common sense to create a par for a class with inadequate data. At Keeneland the par for maiden $30,000 claiming races is 69; for maiden $50,000, it’s 75. There have been only two maiden $40,000 races run over the past three years – hardly a meaningful sample – but the par has to be between 69 and 75 and so we gave it a 73.
We use an internal report that shows the average winning figure for every class at a track, but we don’t automatically use that data for our pars. We review these average winning figures to make sure they don’t contain aberrations that may be produced by a small sample or a couple of extraordinarily fast or slow races that skew the data. The par figure for every class is supposed to be higher than the class below it. If a track’s N1X allowance races were faster than N2X allowance races, that would clearly be an aberration. We would adjust the pars so that the N2X par is higher. We want the our published pars to be logical.
– Andrew Beyer
Andrew Beyer and his associates regularly answer questions from readers about Beyer Speed Figures. To submit questions for publication, go to drf.com/beyer. Please include your full name and place of residence.
