New York Racing Association challenges methodology used in N.Y. Times article

The New York Racing Association is challenging the methodology used by the New York Times in a March 25 article that reported a high incident rate of injury at Saratoga Race Course over a three-year period.

According to the Times article, Saratoga had an incident rate of 5.5 horses per 1,000 starters from 2009 through 2011. The Times purchased data from Equibase and used the terminology used by chart callers to calculate their figures. Among the terms the Times used to calculate their incident rate were “vanned off,” “lame,” and “went wrong.”

In a press release issued Thursday, NYRA “asserts that it is unreliable and potentially deceptive for the Times to rely on chart callers’ descriptions of the running of a race to estimate how often horses get injured.” The release noted that chart callers do not follow up with veterinarians or trainers to find out if a horse was injured.

Further, NYRA asserts that horses may be vanned off for many reasons “that have nothing to do with an injury.”

As an example, NYRA noted that on Aug. 25, 2010 the footnote for the horse Santo Gato read “pulled up, vanned off” but the horse was not injured. Santa Gato raced 16 days later at Presque Isle Downs and has won three races from 19 starts since his Saratoga race.

According to NYRA, of the horses that were vanned off – and NYRA did not disclose how many there were – at Saratoga from 2009 through 2011, 19 came back to race, making a total of 149 starts through the end of last month.

In its release, “NYRA concludes, therefore, that there is plausible cause to regard the New York Times’ incident rate metric as faulty and to consider that its purported goal of assessing how often horses break down or get injured leads to misleading and incorrect results.”

In an e-mailed response from Matt Purdy, investigations editor for the Times, the paper stood by its reporting and said no one has proven any inaccuracies with its data.

“NYRA is choosing to use a different – and questionable – definition than the one we used in analyzing racing results,” Purdy wrote. “As we stated in our article, we focused on incidents where horses broke down or “showed signs of injury” based on the unbiased, contemporaneous reporting of chart callers.

“NYRA says that some horses that were vanned off later returned to racing. But that does not prove they were not injured or did not “show signs of injury.” As our story showed, for example, it is not unusual for injured horses to be given pain medication and returned to racing. To make an assessment of a horse’s health based on whether it ran again is speculation. We made a point of not relying on speculation but rather on the official record of races.”