06/02/2009 11:00PM

Tweaking formula for synthetic figures


Since synthetic tracks were introduced in North America, one fact about them has become indisputable: They are different. They are less speed-favoring than traditional dirt. As a result, the style of racing on the new surfaces is different. Typically, the early pace is slower and the horses are more tightly bunched at the finish.

Speed figures are different on synthetic surfaces, too. The distinctive nature of these races has prompted a change in the mathematical underpinnings of the Beyer Speed Figures.

The new calculations go into effect on Wednesday and will be applied retroactively to Jan. 1. They will make synthetic figures closer to those for dirt. Top-class stakes horses will earn figures as much as 4 points higher. Many bottom-level races will be 3 or 4 points lower. Figures for horses in the middle of the class spectrum will be changed to a smaller extent, if at all.

Users of the Beyer Speed Figures know that individual numbers are often changed for a variety of reasons. The calculation of turf figures was amended several years ago. Now there is enough data about synthetic tracks to suggest that a revision of these figures is warranted.

Readers of Daily Racing Form have often observed that figures on synthetic tracks appear too low - a judgment based mostly on high-profile stakes races. In Southern California, the average winning figure in all stakes for older males is 102 on synthetics, compared with 106 on dirt. Many fans therefore assumed that the figures were low overall. This was not the case.

While stakes and allowance horses were running lower figures, bottom-level races produced an unexpected result. At Hollywood Park, the average winning figure for maiden $25,000 filly claiming races on dirt was 65. But after Hollywood installed its Cushion Track, these same races produced an average winning figure of 68. At Woodbine, the figures for maiden $10,000 males jumped 3 points when they started running on Polytrack. At every synthetic track in North America we observed the same phenomenon.

While there were reasons to explain the lower figures in stakes races across the country (inferior horses, slow early pace, etc.), there was no good reason why maidens should improve their figures. When my partner, Mark Hopkins, and I finally found the explanation, we felt like physicists who had discovered a corner of the universe where E doesn't equal MC2.

Underlying all of the Beyer Speed Figures is a chart that assigns a numerical value - the speed rating - to any final time at every distance. (The rating is added to the track variant - which gauges the speed of the racing surface - to produce the Beyer Speed Figure.) The speed chart was first published in my book "Picking Winners" in 1975, and it has stood the test of time. Its accuracy has been almost magical - for dirt races. But synthetic tracks have proved to be a new game.

Accordingly, we have revised the chart for synthetic tracks, expanding the range of the numbers so they are higher at the top and lower at the bottom. At 1 1/8 miles, for example, every tenth of a second in our old calculations equals 0.9 points on the Beyer Speed Figure scale. In our new formula for synthetics, one-tenth of a second equals 1.04 points. While these numbers may appear abstract, they reflect the different nature of racing on the two surfaces.

Over synthetic tracks, winning margins are narrower and runaway victories rarer. In the last 10 runnings of the Blue Grass Stakes on dirt, the average winner scored by 3 1/2 lengths. In its three editions on Polytrack, the winning margins have been a nose, a neck, and 1 1/2 lengths. The horse who wins a race on Polytrack by 1 1/2 lengths (or about three-tenths of a second) is scoring by a more significant margin than if he won by the same amount on dirt. The numbers in our new chart quantify the fact that a tenth of a second on a synthetic surface deserves more weight.

Because of the same principle, we have made a slight change to our calculation of beaten lengths in synthetic races, although these changes will be imperceptible to most readers.

The net effect of this new methodology will be to make Beyer Speed Figures on synthetic tracks closer to their counterparts on dirt. Maiden $25,000 filly claimers at Hollywood will earn approximately the same figures that they did on dirt. So will $8,000 claimers at Woodbine and maiden special weight runners at Turfway Park and allowance runners at Golden Gate Fields.

However, the best stakes horses and even some top-quality allowance horses will not necessarily earn figures on synthetic tracks that are as high as the best horses run on dirt. The reason has to do with the horses themselves, not the calculation of the figures.

The U.S. Thoroughbred industry produces horses bred for speed and bred (for the most part) to run on dirt. It would not be logical to expect that America's very best horses would be just as productive on non-dirt surfaces that usually don't favor speed. Moreover, the early pace in stakes races on synthetic surfaces is often unrealistically slow - so slow that horses can't make up enough time in the stretch to produce their best possible final time.

Although the revised synthetic numbers will generally be comparable to dirt figures, they will not magically predict how a horse will perform when switched from one surface to another. Horses usually prefer one surface to another - just as most prefer either turf or dirt. A horse who earns 90 on a synthetic track and switches to dirt is just as apt to run an 80 or 100 as he is another 90.

Nevertheless, we believe it is important to make these changes.

Since the Beyer Speed Figures were introduced into the Daily Racing Form in 1992, readers have come to understand what the figures signify about horses' abilities. If a 2-year-old wins a maiden race with a figure of 95, he is probably a future stakes winner. If a 3-year-old colt wins a major prep race with a figure of 97, he is a pretender, not a legitimate Kentucky Derby contender. We want readers to have the same confidence in our speed figures on synthetic surfaces as they do for the figures on dirt, and with our new calculations we believe that they will.


The changes in the calculation of Beyer Speed Figures for synthetic tracks will generally raise the figures for higher-class horses (who run in the 90s or 100s) and decrease those for lower-level horses with numbers in the 60s or below. For example:

We have recalculated every day of synthetic-track racing in 2009, producing new track variants. As part of this project, we have made other tweaks to the synthetic figures. In some cases, the figures at a track were raised or lowered by a point overall so that the numbers at all tracks are perfectly aligned with each other. As a result of these changes, there is no strict correlation between the old figures and the new ones - e.g., a figure of 98 is not automatically translated into a 101, as in the example of Shadowless, above. Another horse's 98, in different circumstances, might remain a 98. But the general rule is: The fastest races produce higher figures, the slowest races produce lower figures.