# COMMENTARY : New Speed Index Doesn’t Figure

THE WASHINGTON POST

The Daily Racing Form carries the authority of holy writ; every symbol and number in its past performances is invested with significance.

The paper’s publishers rarely tamper with their product. But the Form now has introduced a whole new component to the record of every horse in North America, the Speed Index, and has ballyhooed the innovation daily in its pages.

At the bottom of each horse’s record there now appears a line such as this:

“Speed Index: Last Race: (plus)2.0. 3-Race Avg.: (plus)3.3. 9-Race Avg.: (plus)0.9. Overall Avg.: -1.7.”

This is the Form’s belated response to decades of criticism that its speed ratings are worthless. It was surely prompted by the arrival of a competing paper, the defunct Figs Form, which made its sophisticated speed figures the centerpiece of its product.

The Speed Index is derived from the two numbers that always have appeared in every past-performance line, the speed rating and the track variant.

When a horse’s record shows a pair of numbers like 72-23, the 72 is the rating, which shows how fast he ran; the 23 is the variant, which supposedly reflects the condition of the racing strip. Horseplayers traditionally have added the two figures together, but now the Form saves them the trouble.

It compares the sum with 100, and that’s the Speed Index. In the case of the 72-23 horse, those numbers total 95, so the Speed Index is -5.0.

This calculation evokes the old aphorism of computer programmers, “garbage in, garbage out,” because the speed ratings and track variants on which the Speed Index is based are riddled with flaws.

In the Racing Form’s speed-rating system, the best time run at a particular track and distance during the past three years has a value of 100. One point is subtracted from 100 for each fifth of a second slower than this standard.

If the best six-furlong time at Laurel is 1:09, a horse who runs in 1:12 has run 15 fifths of a second slower than the record and thus gets a rating of 85. The defects of this method could be seen clearly in a race this week at Laurel.

A horse named Daisy Patch has an overwhelmingly superior Speed Index for his last race -- which he earned by running 1 1/4 miles in 2:07 3-5 at Penn National.

That is ridiculously slow time, but Daisy Patch gets a high rating because Penn National cards few races at this distance and the track record for the distance is unrealistically slow.

Any horse who runs 1 1/4 miles at Penn National will get a good speed rating. Track records are a flimsy basis for evaluating horses’ times.

If the Racing Form’s speed ratings are filled with such logical inconsistencies, the track variants are even more misleading.

They purport to show how fast the racing surface is on any given day -- again, by making comparisons with the track record. Last Thursday, a filly named Find Rachele had the superior Speed Index in a field, based on a speed rating of 82 and a track variant of 22 -- numbers that suggested she had run reasonably fast on a track that was fairly slow. But these figures proved to be misleading, because of the way the variant was calculated:

On Oct. 24, Find Rachele ran 1 1-16 miles in 1:46 1-5, earning a speed rating of 82, because she was 18 fifths of a second slower than the record for that distance.

One other route race was run the same day; it happened to be an event for cheap maidens, and the time was 1:47 4-5 -- 26 fifths slower than the record.

The Form averaged 18 and 26 to come up with the variant of 22. Find Rachele came up with a good figure because the other route race happened to be a cheap slow one.

If Find Rachele had run on a day when a fast stakes race was being run, the variant would have been lower and she would have wound up with a much worse Speed Index for the same performance.

Serious handicappers recognize these flaws and ignore the speed ratings in the Form but, even so, the introduction of the Speed Index is galling.

If the Daily Racing Form wants to improve its product, it could make plenty of changes.

It ought to show when horses have been treated with Lasix in previous starts.

It might show (as Figs Form did) an indication that horses have worn bandages in their previous starts.