Putting TimeformUS Speed Figures to the Test


As part of our relentless quest to present the best handicapping data possible while being as transparent as possible, today we will look at the results of a recent study we did on our speed figures and Spotlight Speed Figures. Although we are generally pleased with the results, we’ll use them to identify areas where we can improve. Moreover, we believe that our customers will profit from knowing that certain situations call for extra enthusiasm on the part of the horseplayer, while other situations require an even more nuanced view.

The speed figure aspect of the study is simple enough, since just about everybody is familiar with speed figures. For our purposes here, we focus on the horse who has the fastest last-out speed figure, and we examine his win percentage and ROI in various situations.

Our Spotlight Speed Figures are the result of a complex process that aims to find in each horse’s past performances the recent race that is most “representative,” meaning most indicative of the quality of the effort that he will put in today.

A note on the ROI stats below:

The amount of money that racetracks take out of the betting pools (the track takeout and breakage) varies from state to state and track to track. The size of this “rake” ranges from confiscatory to positively obscene. As a practical matter, while avoiding hairsplitting, this means that the ROI stats presented below should be considered “neutral” if they are in the 1.50-1.60 range. Anything significantly above that means we are seeing overperformance. Anything significantly below it means we are seeing underperformance.

Here are the results:



What we find most surprising in these results is the lack of surprises. Ordinarily, such studies produce some results that defy logic. But no such results seem to be present here.

The overall results for top last-out speed figure show a win percentage of 29 and an ROI of $1.75. Keeping in mind that these results are based on “blindly” betting the top last-out speed figure into a diabolical track takeout, we find these results pleasing–because our speed figures are causing a substantial amount of the track takeout to disappear. And the same effect is visible in the overall results if one is blindly betting our top Spotlight Speed Figures (27.5% and $1.72). Indeed, the results are almost identical.

But of course one is not obliged to bet the top last-out speed figure or Spotlight Speed Figure blindly. As one can see on the chart above, requiring that the race being used (whether it is last-out speed figure or Spotlight Speed Figure) be on the same type of surface as today’s race improves the results. Now we see win percentages of 30% and 28.5% and ROIs of $1.81 and $1.76.

Requiring the race to be on the same type of surface as today’s race AND run within the last 120 days improves the results even further. Now the ROIs are up to $1.84 and $1.80.




Profitability is within sight–and with only this minimal amount of handicapping. It’s enough to make one wonder what would happen if one used these tools in combination with full-fledged handicapping.

But the results of our study are not all so rosy. Betting the top last-out speed figure or Spotlight Speed Figure places one at the mercy of how representative these numbers truly are, given the circumstances of today’s race. And the results from the study plainly show that certain situations are treacherous indeed. The ROIs plummet when the races being used by either the top-last-out-speed-figure method or the Spotlight Speed Figure methodology are either on a type of surface that is different from today’s surface or come after a layoff of more than 120 days.



This is hardly a surprise. After all, betting the top last-out speed figure puts one at the mercy of changing circumstances, and programming the Spotlight Speed Figure to find the most representative race is a tall task when layoffs and surface-changes are involved. However, we do plan to learn from this study and adjust our formulas for selecting the most representative Spotlight Speed Figures.


While we are largely pleased with the results of this study, they are a reminder (as if we needed one) that there is no totally satisfactory substitute for handicapping horse races in the most thorough manner possible. Much as we love our Spotlight Speed Figures, and much as we are pleased with their performance in much of this study, we always find it important to remember that what they are, more than anything else, is a convenient way to get INTO races. They are a quick and easy way to spot races that have potential for the handicapper. But getting into the handicapping of a race is not to be confused with being done handicapping a race.

Related: Putting the Pace Projector to the test

This entry was posted in Data Studies, Product features. Bookmark the permalink.

11 Responses to Putting TimeformUS Speed Figures to the Test

  1. john cosides says:

    my grandmother can pick 30% n knowing nothing as the saying go’s. I look for 60 to 70% l winning as my goal. not losing 7 0ut of ten. . thanks for the info it will help in playing in the future. while i’m at it. do you have jockey stats as to what distance there wins are at. I use to keep a record lots of work. but I did find some gems one that sticks out was jose santos he had say 11 wins but 9 were at 1 1/16. that to me made me look twice when he rode at that dist. keep up you good work


  2. Barry Morrison says:

    Did you include the same type of race (sprint or route) as well as surface in the results for the data?


  3. Cal says:

    This is indeed encouraging, as it is clearly beating the take. What would be even more interesting, since you are operatiing in a competitive environment, is how does this compare to last out Beyer figures, Brisnet, Colt’s Neck, Rags, or other systems. This answering of course, not the question of whether Timeform figures add value, they certainly do, but also whether they outperform the competition.


    • marcatrtr says:

      We encourage you to ask other providers for similar studies. We provide ours as a service to our customers and as a way of helping us improve.


  4. Tim says:

    Also, in view of these results, it would be a big time saver if the Preview page could show both the Spotlight and the Last Out speed figures, rather than just the Spotlight number.


  5. Tim says:

    Presuming the ROI numbers could be further increased with some fairly simple handicapping (e.g., eliminating jockeys / trainers with low percentages), and eliminating plays on horses that were odds-on or similar.

    Looking forward to your further refinements here.


  6. Hi David,

    All tracks, yep.

    Good suggestion on breaking it out by class. We’d like to use our race ratings as a proxy if possible, but those are effectively generated on the spot when entries are drawn. Let’s see what we can work out. Stay tuned.


  7. Zippy Chippy says:

    I think the goal of the spotlight figure should be to outperform a very simple “last race”, otherwise why not just give last race.


    • There’s a number of different ways of looking at races. While last out figs are currently outperforming Spotlights, the thing we like about spotlights is that it’s more likely to show the customer a live longshot, as people are simply more likely to lean on last out figs. Our goal is to refine both figure sets, and we’re not entirely opposed to your suggestion. Consider it a work in progress.


  8. David G. says:


    Was this study done at all tracks from Assiniboia to Wyoming Downs or just the major circuits in North America?

    Also, did you guys perform a segmented study for different race class structures? Curious to see if you came up with the same overall results regardless of whether it’s a Breeders Cup race at Santa Anita or a nickel claimer at Mountaineer.




Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s