America's Best Racing. The Jockey Club.

The 2012 Del Mar meet beckons and if the thermometer didn’t remind you that summer is well underway, the scratchy, but timeless recording of Bing Crosby crooning about turf meeting surf surely does. Trakus begins its sixth season tracking the races at Del Mar, and some interesting data from the 2011 meeting could be helpful for the 2012 edition.

The majority of statistics available regarding jockey performance are often limited to several basics – record at the meet, record for the year, winning percentages on specific surfaces, winning percentage when riding for a trainer, and some others. These statistics, however, do not identify how a jockey actually rides races on a regular basis. Some riders are anecdotally categorized by perceived skill, or potential bad habits, in certain racing conditions. Most horseplayers have made, or overheard, these sweeping statements in passing while discussing a race:

• “(Jockey X) is a great speed rider”
• “I love betting (Jockey Y) when he rides on grass”
• “(Jockey Z) never saves ground”

Quantified statistics, however, don’t exist for all of these chest-thumping claims. Potentially, Trakus could change some of that with an increased study of jockey performance. 

Getting the numbers

After the 2011 Del Mar meeting, we took the distance-traveled data from races run at three distances – 5 ½ furlongs (a furlong is an eighth of a mile) on Polytrack (the synthetic surface that makes up Del Mar’s main track), one mile on Polytrack, and one mile on turf. It is readily acknowledged that the sample size for a seasonal meeting, and at select distances, is small.  These distances were chosen due to the proximity of the first turn, and the added chance for more ground loss than in races of different distances – essentially, taking races where working out the best trip possible was of greater importance. Fifty-one races at 5 ½ furlongs, 30 races at one mile on Polytrack, and 33 races at one mile on turf were studied. 

To begin the analysis, we obtained the average number of feet each horse traveled (as tracked by Trakus), segmented by post position. For example, in races last season on Polytrack at 5 ½ furlongs, horses breaking from post one averaged 3,658 feet while horses starting from gate 10 traveled, on average, 3,680 feet.  The difference of 22 feet equates to approximately 2.5 lengths of extra ground. Races with two turns, which greatly improve opportunities to lose ground, often see increased average ground loss from outside posts. 

Taking the previous example above, if a horse broke from post one going 5 ½ furlongs, and actually covered 3,670 feet, that would be 12 feet more than average (and would earn the rider a +12 rating). Negative ratings indicate more ground saving, while positive ratings indicate the jockey’s mounts have covered more ground than average from their starting gate. Below are the top riders from the 2011 Del Mar meet, ranked in order from the lowest Trakus Jockey Efficiency Rating (JER), rating jockeys from most ground-saving to least.

Any definitive conclusions are difficult to reach from this admittedly cursory study. Interpreting the results for one of the jockeys above, the data suggests Martin Pedroza’s mounts over the three distances examined, covered about 2.5 feet less than the average expected from their post position. More specifically, digging into the individual distance data (which does not appear in chart form here), Pedroza’s mounts going 5.5 furlongs earned a JER of -5.25, mounts over one mile on turf earned a -3.33 JER, while mounts at one mile on Polytrack covered more ground, getting a +6.18.  A weighted-average based on the number of mounts is used to get the cumulative JER. Pedroza’s most ground-efficient mounts came when sprinting, and going two turns on the turf.

While Hall of Famer Mike Smith might check in at the bottom of the list, albeit from a total of only 31 mounts studied, his form at least suggests a style that perfectly fit the thrilling needs of future Hall of Fame mare Zenyatta, who closed from well off the pace and whose momentum could never afford to be stopped.

Criticisms abound

The pre-conditioned criticism of any type of study that reviews jockey performance based on ground-loss history is obvious. Some horses prefer racing in the clear, outside, and trainers are likely to suggest or demand this style for particular mounts. On occasion, a bias could exist, which leads jockeys and instruction-giving trainers to plot a wider course if the rail is perceived to be “bad,” regardless of the actual condition. But under almost no conceivable circumstance would a jockey’s mounts always be on horses that need to race outside, or always recorded when surface bias was believed to exist. 

One could even suggest (and I’m not, rather doing so for explanatory purposes) that the rail is always “bad.” If this was the case, and all jockeys knew it, the average ground coverage statistics would inflate across all posts. 

Advanced application of the Trakus data is underway. Between the Trakus-adjusted margins (TAM) referenced in our previous two blogs, or the JERs, the opportunities to go inside the race continue to grow.  We also welcome your comments, either here on the blog, or via email: press@trakus.com.