Saratoga, Gulfstream and Kentucky Downs have all run races over wrong distances within the last six weeks – at least one half-furlong longer than the races were scheduled.

This 50th volume of #FreeDataFriday is not an explanation of some obscure method of timing races, it offers merely a sobering fact, easily exposed BECAUSE of the way in which America times horse races.

First, understand that nearly every distance of a race run in North America is not the actual distance traveled, but the distance which is timed. Horses run-up to the starting point and reach the spot which is the published race distance away from the finish, and then the clock starts. It might be 30 feet, 50 feet, 70 feet or more. It depends on many factors.

Yes, we think this is the wrong way to time races, but at least we know that run-up exists. 

But when the untimed portion of race is a half-furlong (1/16th of a mile) or more, and those wagering on that, riding in those races or preparing horses for such events are either unaware or not properly informed of this? Well, that's a problem - for the integrity of the sport and for the confidence of stakeholders.

Saratoga ran the Grade 2 Bowling Green on August 1, 2020 at a reported 1 3/8 miles on turf – the race was likely at least 1 7/16th miles, more than a half-furlong farther than reported to anyone, including owners of horses in the race, jockeys who rode it and the bettors who staked more than $1.7 million on this race.

Last Saturday, September 5, Gulfstream Park ran two listed, black-type awarding stakes (the Bear's Den and Miss Gracie) at a reported “about” 7.5 furlongs on the turf. The races were very likely about 540 feet, or roughly four-fifths of a furlong longer than that, much closer to 8.5 furlongs.

On Monday, September 7, Kentucky Downs ran four races at 6 ½ furlongs. The reported “run-up” for the race, acquired via the new Equibase-serviced Gmax timing and tracking system, was 330 feet, a distance that equates to a half-furlong. In other words, horses actually ran seven furlongs. The charts for these races (R2, R6, R7) are HERE - but a replay can be found via ADW replay providers. 

The circumstances of all of these races, and the impact of the extra ground covered, and the degree of harm done by presenting customers with these errors, assuredly, varies.

Here is what we know.

The times of all the races in question are not necessarily wrong – all of the races are timed from the point that is the published distance of the race from the finish. What that means is that the clock starts WELL after the race has actually commenced and makes it remarkably easy to “see” these errors.

In other words, once the horses get to the point that is 1 3/8 miles from the Saratoga finish, or 6.5 furlongs from the finish at Kentucky Downs – the timing system in place starts. Saratoga uses beam-based times, Gulfstream uses Trakus and Kentucky Downs is a new user of the Equibase-enabled Gmax. But an examination of the actual time the horses are racing differs substantially from the official times.

In all of the races noted above, horses raced for no less than seven seconds before the timer began and, again, the fractions recorded for each race are not disputed. At Gulfstream, horses ran for more than 11 seconds, a duration that is the equivalent of 12.5% of the actual time recorded for the race.

Below, review the chart which shows the observed times from video, either via YouTube or replays available from most ADWs, the actual time horses were racing according to those observations, a “hand-time” using a stopwatch from the break of the gate to the finish, the official time of the race as it was reported and the variance between the hand time and the official time.

FDF50 chart.PNG

The variances, in orange on the right, tell the tale. The actual time horses are racing is substantially longer than what is reported to the public. If there are approximately six horse lengths in one second (1 length = approximately 0.16 seconds), then a variance of 7.63 seconds is the equivalent of 45.6 lengths. A variance of 11.34 seconds is the equivalent of 68 lengths.

This is madness.

Over $6.8 million was wagered just within these individual races.

There seems to be a reckless disregard for the truth from track operators as it relates to running races at the distances they schedule.

Here are some questions:

1. Would a horse that won the Bowling Green have been demoted from first to fourth horse if the race was run over the published distance of 1 3/8 miles instead of the actual distance of 1 7/16 miles, given that the interference occurred in the “last half-furlong” of a race that was already a half-furlong too long? (Saratoga)

2. Would a filly have earned black-type for the first time, potentially increasing her future value, if the race was actually contested over 7.5 furlongs instead of nearly 8.5 furlongs? (Gulfstream)

3. Will bettors have won or lost because their analysis and bets were formulated and executed believing the distance published by the track, replicated and sold by Equibase, further sold and distributed by downstream providers like the Daily Racing Form, TimeformUS, Thoro-Graph, BRIS and Ragozin, was accurate?

4. Will regulators – state racing commissions – step forward and hold operators accountable to ensure accuracy in the distances and times of races run in their jurisdictions? (Kentucky, New York. Florida does not have a racing commission)

5. What are the challenges keeping track operators from running races at the distances THEY set and how can they be overcome to ensure accuracy for all?

As it relates to the last question, some of these answers are clear.

Elements of tradition (“this is the way we’ve always done x”), course management, and safety concerns are the cause of these issues, while the product leaves us with duped customers or participants – bettors, horsemen, jockeys and fans. Accuracy matters, or at least, it should.


North American racing does not time races from the break of the gate. Almost every race is run over a distance LONGER than what is published. But run-up on dirt races is normally consistent, and as all seven examples above are turf races, it is clear that portable rails and turf management are partial causes, or exacerbating, the problem.  

But how much run-up is too much?

The answer to this should be, at the very least, when the run-up is a half-furlong or more considering the sport still measures distances in such ways.

Over the last year, the #FreeDataFriday series has covered a plethora of issues which impact racing. At its heart, this has been about the need to see access to racing data improved, preferably at reduced price points, and used to attract new customers to racing’s wagering markets. Remarkably, though, it seems the sport either fails to either check its own basic math on occasion, or worse, just ignores it.

Not only does accuracy matter, but inaccuracy is an affront to integrity.

Look at how long these “errors” are when extrapolated over a map of Gulfstream, in this example below. Using simple Google Map analysis, with the rail set at 108 feet off the inside, the red dot is located where the gate was approximately placed and the yellow dot roughly 540 feet beyond that (a total of 648 feet of measured distance on the map). This point is already on the first turn and is roughly the spot which is 7.5 furlongs from the finish at this rail setting where the timing would begin.

FDF50 - GP Turf.PNG

To put added perspective, consider this – 540 feet is more than half the distance from the top of the stretch to the finish in dirt races at Gulfstream. The image below provides more context. The yellow dot representing 540 feet back from the finish line.

FDF 50 Stretch Dirt.PNG


Those races at Gulfstream were “about” 7.5 furlongs like the Kentucky Derby is “about” 1 1/8 miles, which is to say, it isn’t. In 2020, racetracks in America should be running, timing and reporting the exact distance of a race as it is scheduled. Not 7.5 furlongs with 540 feet of run-up, not 6.5 furlongs with 330 feet of run-up. Schedule it, offer betting on it and run it at a precise distance. Precision, in both reporting and execution, are needlessly tricky. 

The simplest solution is for North American tracks to start accurately reporting existing distances, using existing measures, with completeness. Races at six furlongs on dirt at Churchill Downs have a run-up of 220 feet. The timed portion of the race may be six furlongs, but the race is contested over a distance that is really six furlongs and 73 yards. If you can be disqualified for an offense in an untimed portion of the race, why can't we report (and time) the actual distance horses run?

What might be more precise? Yes, the metric system. It's easier than it seems.

That same Churchill race reported as six furlongs and 73 yards would be 1,274 meters. 

Unarguably, the metric system presents one standard measure, boiled down to a precise, uniform number, and something we accept in human racing around the world, at both the highest levels of professional competition and the lowest levels of amateur, even children's sports. Oh, it’s used in racing too…just not here…yet. 

Regardless of the measurement used, accuracy is easily the most important need. 

Racing must start timing from the break of the gate, not some point which is the published race distance from the finish while the gate itself is positioned quasi-arbitrarily behind that point, 50 feet, 150 feet, or as was the case in the last week, maybe between 330 to 540 feet away from the actual point that is the distance everyone otherwise believes.

There are significant costs associated with this development too, but they must be borne.

This status quo is fraudulent to horse owners and bettors and misleading to jockeys. These frauds, perpetrated on the public, could leave tracks open to litigation from aggrieved customers.

Our sport, and its operators and regulators, are not taking it seriously. The actions needed to correct these errors are clear. It is up to those same entities to take those measures and ensure accuracy.​​​​​​


Popular Posts