Who's on First, What's on Second,
and I Don't Care about Third

In Thoroughbred Handicapping: State of the Art, Dr. William Quirin offered one of my favorite definitions of class: "Class in the Thoroughbred relates to its ability to race competitively at a given level of competition." 

He goes on to suggest several broad ways to evaluate class by using the earnings box, wisely cautioning that all earnings fail to account for recent form--a potentially crippling weakness in this way of evaulating class. 

Quirin's computer studies over two seasons of racing showed that there was very little difference among the various rankings that can be performed even by beginning handicappers:

GROSS EARNINGS: Using simply the total earnings from the past year or two. "Favors durability at the expense of ability." 

AVERAGE PURSE VALUE: Using the purse distribution formula (.60 x wins, .20 x seconds, .10 x thirds and .5 x fourths OR local circuit purse payout percentages times earnings) 

AVERAGE EARNINGS (EARNINGS PER START or EPS): "Favors consistency at the expense of quality. 

I've always touted EPS as the broad class ranking of choice. Here's why, with a special twist added: 

Gross Earnings can mislead casual players who fail to readily see a horse who used to be a great runner but who is now desperately trying to find a race it can win. I find this to be especially true on the bush-league circuits, where $2500 and $3500 claimers who won a big purse last year and haven't won anything since appear to have a good record--based on gross earnings. 

Average Purse Value used to be a pretty good indicator, even with the purse distribution values varying in different parts of the country. But over the past ten years, we've seen purses cut drastically halfway through a meet, which ruins this approach quickly and obviously. Also, the ridiculously inflated purses of state-bred races and most maiden races can create some wild fluctuations, giving the occasional winner or up-close runner of one of these races a phony, inflated ranking. Worse yet, the math of APV will sometimes equate two horses, as in Quirin's example of a horse winning a race with a purse of $25,000 receiving the same APV as the horse who ran fourth for the same
purse! 

Earnings Per Start would seem, then, to be the best of the bunch. While Quirin says that a lower-class animal trying a higher class level might rank better than a higher-class animal dropping down after being unsuccessful against better with this approach, I prefer the consistent animal, the horse who tries hard on a regular basis over the overmatched allegedly higher-class horse struggling to keep up. But I use a variation that neither Quirin nor any other published handicapping researcher has ever reported on, to my knowledge. 

How does a horse run third ? 

I contend that horses who show fall into three categories: 

1. The horse who has a front-running style that enables it to be a serious contender for all or the last half of a race, who may be overmatched or outrun and winds up placing third by falling back. 

2. The horse who inherits third place by virtue of the horses in front of him burning out as the result of a sizzling pace duel up front, or as the result of the horses in front of him running out of gas due to a lack of conditioning or improper placement. The type of horse who merits the trip comment "passed tired horses." 

3. The horse, usually in the best races a track has to offer, the ones that offer serious purse money, who isn't good enough to win but who is kept running to grab a share of that large purse rather than just let run on its own. These races are more of the nature of the big feature races run once a week rather than an ordinary weekday feature race. 

So, case number 3 excepted, this says to me that in most races, any horse can run third. 

And even if you disagree with this, unless you observe the race or have excellent trip notes, it's usually a guess which of the first two scenarios took place to enable the show horse to show. 

So for a broad class evaluation, which has served me well, I submit that EPS be calculated this way: 

1) If less than 6 races this year, use both years' figures in the money box. 

2) Subtract the number of shows from the total number of races. 

3) Divide the earnings by this figure to get the EPS. 

Purists will object, saying that this is an adjusted and artificial figure. My response is that I am more interested in what I call front-end consistency, as well as a broad rating. Focusing too narrowly on numbers always results in overdependence on figures at the expense of  basic handicapping factors.

I'll admit that I concentrate more on win and exacta betting rather than trifecta wagering, which is why in most races I don't care who runs third.