Are Speed Figures Still Important?

In 2010, I did a column for Youbet.com regarding the importance — or lack thereof — of speed figures. Back then, I only had Trackmaster figs to work with, so I thought it might be kind of fun to revisit my study using the Brisnet numbers. But first, some background…

It’s become a common racetrack lament, best summed up by a guy I ran into on my last trip to the OTB… well, actually, he ran into me — I still have the beer stain on my shirt to prove it.

“See that?” My booze-loving buddy asked me, pointing vaguely to one of the dozens of big screens scattered about the establishment. I had no idea what he was talking about, of course, and looking into his glazed and unfocused eyes I doubted that he did either, but I played along and waited for him to continue.

“The six-horse that just won at Aqueduct had a best Beyer of 66. And he beats the favorite who ran a 78 last time. Speed figures are [equine excrement],” the “Least Interesting Man in the World” informed me, shaking his head in disgust and staggering away (presumably to go bet on a horse with lousy Beyers).

Not since the famous “tastes great/less filling” debates of the mid-1970s, has a product or service created the kind of controversy that speed figures do among horseplayers today. Andrew Beyer, a speed handicapping pioneer and the man whose figures have appeared in the past performances of the Daily Racing Form since 1992, called the discovery of his ratings “one of the most momentous moments of my life.” Yet, even Beyer concedes that their worth — at least as a betting tool — has greatly deteriorated over the years.

“In 1993, as I was writing ‘Beyer on Speed,’ I asked the Daily Racing Form to analyze thousands of races and assess the performances of horses with the top speed figures,” Beyer wrote in a 2007 preface to “The Winning Horseplayer,” one of his many classic tomes on Thoroughbred handicapping. “A $2 bet on each of them would have returned an average of $1.85 — a 7 ½ percent loss. Speed figures remained a vital tool for understanding the game, but my old fears had proved correct. The figures were no longer sufficient to beat the game.”

But were speed figures ever “sufficient” to master the Sport of Kings? After all, in “Picking Winners,” his very first book, Beyer didn’t even delve into speed handicapping until Chapter 7. Instead, he chose to first discuss “Larceny and Betting Coups” (Chapter 3), “Track Biases” (Chapter 4), “Trainers” (Chapter 5) and “The Horse’s Appearance” (Chapter 6). If speed is, or was, the predominant factor in determining the winner of a horse race, shouldn’t it rate as the lead topic?

The fact is a speed figure — whether it’s produced by Beyer, Len Ragozin (The Sheets), Jerry Brown (Thoro-Graph), TrackMaster, Brisnet or the Racing Post — provides handicappers about as much information as the current price of a stock provides investors. It is merely a snapshot in time, a reflection of a particular race run at a particular track on a particular day. And while it may certainly serve as a guide to future performances, it does not in any way determine those performances.

However, to truly test the effectiveness of speed figures as a prognostic tool in the new millennium, I decided to consult my database of over 6,200 races contested from 2004 to present. Unfortunately, I only have access to the TrackMaster numbers, which are a bit more mechanical in nature than some of the other figs on the market. Nonetheless, they should be more than adequate for our purpose here.

Let’s start by looking at horses possessing the best last-race TrackMaster speed rating (by at least one point, to eliminate ties):

Number: 4,519
Wins: 1,316
Rate: 29.1%
Return: $7,853.80
ROI: -13.10%

Apparently, things have gotten much worse for speed handicappers since Beyer’s initial query in 1993. Sadly, the situation didn’t improve much when I eliminated those horses above that failed to meet par in their latest start either:

Number: 3,076
Wins: 925
Rate: 30.1%
Return: $5,354.70
ROI: -12.96%

The stats got a little less gloomy if the qualifying steed above was also the post-time favorite, but the digits were still decidedly dour:

Number: 1,536
Wins: 660
Rate: 43.0%
Return: $2,777.50
ROI: -9.59%

On a positive note, when the aforementioned public picks returned to the races in two weeks or less, the ink finally turned green:

Number: 345
Wins: 168
Rate: 48.7%
Return: $698.50
ROI: +1.23%

Now, let’s conduct the same surveys using the Brisnet speed figures and a more contemporary database, consisting of over 14,300 races.

Test #1: Horses possessing the best last-race Brisnet speed rating (by at least one point, to eliminate ties).

Number: 11,679
Wins: 3,375
Rate: 28.9%
Return: $19,916.30
ROI: -14.73%

Test #2: Horses possessing the best last-race Brisnet speed rating (by at least one point, to eliminate ties) that met today’s par.

Number: 5,001
Wins: 1,589
Rate: 31.8%
Return: $8,830.20
ROI: -11.72%

Test #3: Horses possessing the best last-race Brisnet speed rating (by at least one point, to eliminate ties) that met today’s par and are favored in today’s race.

Number: 2,575
Wins: 1,127
Rate: 43.8%
Return: $4,507.60
ROI: -12.47%

Test #4: Horses possessing the best last-race Brisnet speed rating (by at least one point, to eliminate ties) that met today’s par, are returning to the races in two weeks or less, and are favored in today’s race.

Number: 521
Wins: 262
Rate: 50.3%
Return: $1,043.40
ROI: +0.13%

Note that the stats are very similar, despite the passing of five years and the use of different speed figures. Really, the biggest difference was the higher rate of winning favorites and the lower associated prices, which has been documented by myself and others in the past.

Still, it is interesting that favorites possessing the best last-race speed figure and returning to the races quickly, i.e. “unrested,” continue to churn out a profit, albeit a very small one.