Track level EID data examined

[FenceBored]

When the Jockey Club originally announced the Equine Injury Database project a few years ago, some tracks were reluctant to come onboard without assurances that track level data would not be released. Presumably, this was to prevent people from "misusing" the data. Knowledge is not inherently misuse. Besides, how can we correct misinformation (such as that contained in the recent NYT article) if we don't have accurate information on which to base the correction?

Flash forward to last month, some tracks have agreed to their individual data being released, and more tracks seem to be coming on line with that. As of this afternoon, I've catalogued 51 different track/surface/breed combinations and ranked them by their 3-year cumulative total fatality rate.*

Total numbers
51 track/surface/breed combinations
23 Dirt courses
20 Turf courses
8 Synth courses

Top 10
4 Dirt
4 Turf
2 Synth

Top half
7 Dirt
12 Turf
7 Synth

Bottom half
16 Dirt
8 Turf
1 Synth

Bottom Ten
7 Dirt
3 Turf
0 Synth


*Yes, I know, the switch from synth back to dirt at SA makes their data quirky, but we'll just note the change and move on, ok?

[FenceBored]

Alrighty then, how do the stats for the publishing tracks compare to non-published tracks?

Well, the 27 tracks give us 51 track/surface/breed combinations which account for 34.91% of the starts recorded over the past three years in the EID (http://www.jockeyclub.com/pdfs/supplementaltables_eid.pdf) and 30.10% of the estimated deaths.*

This is not to say that all the safer tracks have published and the less safe tracks haven't, but overall those who haven't published certainly account for a slightly higher percentage of the deaths than their percentage of starts.

*The 2012 table of EID data gives # of starts and fatality rate, but no deaths figures. But, given the # of starts and fatality rate we can estimate the # of deaths recorded in the EID.

[FenceBored]

Presque Isle, which has a synthetic Tapeta surface, has a fatality rate of 0.45 per 1,000 starts, the lowest of the tracks that made their data public. Penn National Gaming Inc. doesn’t make the data public for its tracks, but one of them, Penn National, has averaged 1.65 fatalities per 1,000 starts the past two years, below the national average and an improvement over previous years’ numbers, sources said.
Read more: http://www.bloodhorse.com/horse-racing/articles/68515/pa-horsemen-call-for-state-level-meetings#ixzz1qhPm3mFY

Hmm, Presque Isle did indeed have a 0.45 rate for 2011, but it's 3-year is 0.84 (still pretty good). But, no, it's not the lowest for last year. You have to look to the California Fairs for that. Fresno, Stockton, and Sacramento all had no reported TB fatalities last year.

Sacramento is our overall winner so far with no reported fatalities in the last 3 years.

[Tom]

What is the rate of breakdowns for the World Cup Races - the best in the world?

2 so far from 25 horses......what's that, about 80 per 1,000?

[FenceBored]

Track summaries of EID data (http://www.jockeyclub.com/initiatives.asp?section=2) look to be staying current, not just a yearly release. Fresh summaries up for tracks which were publishing their data and have had a meet that has ended this year.

Santa Anita's main track didn't seem to get the same press it did last year, but that's not because it was safer: 9 fatalities from 2997 starts for a 3.00
fatality rate. Well, its defenders might say, at least that's better than their (non-downhill) turf course's 3 fatalities from 878 starts (3.42 per 1000). :faint: