11/16/2006 12:00AM

What we've learned about Cushion Track


INGLEWOOD, Calif. - By the time deep closer Harvard Avenue clunked up for third on Nov. 1, opening day of the Hollywood Park fall meet, it was clear the new artificial racing surface had altered the dynamics of the game.

Speed has always been an asset in California, but it was no help opening day on Cushion Track. In fact, speed was a liability as all six main-track winners rallied from behind and every pacesetter weakened to third or worse.

Harvard Avenue merely rode the bias for a suck-up third. It looked good on paper. But his performance, and every other opening-day stretch run, was summarily downgraded. The closers benefited from the anti-speed bias, and did not run as well as it appeared.

Performance analysis is not that simple however, and bettors faced a dilemma this week when Harvard Avenue wheeled back just 14 days later. The anti-speed bias on Cushion Track proved temporary. Since opening week the track has stabilized and is fair. To win the seventh race, Harvard Avenue would need to outrun his rivals without assistance from the racing surface.

The 5-year-old gelding did just that. Harvard Avenue powered to a sharp come-from-behind victory in a race that was run over a track that neither complemented nor compromised his closing style. The track played fair. The best horse won. Imagine that.

It was no surprise (Harvard Avenue returned $7), nor was the race particularly fast. Harvard Avenue earned a 92 Beyer Speed Figure, only a point higher than his previous effort.

The victory did, however, illustrate the danger of over-analysis. Although he benefited from the bias on opening day, wagering against Harvard Avenue on Wednesday required one to suspend handicapping fundamentals. Not a good idea.

Harvard Avenue had recent form. But 2-1 favorite Gotaghostofachance was racing for the first time in a year. Truly a Judge was a distance horse in a sprint and running for the first time in seven months. Harvard Avenue is a dirt horse trained by frequent winner Doug O'Neill. Third choice Buckland Manor is a turf horse trained by Paco Gonzalez, who is 1-for-40 this year. Romeo Plus did not belong.

The point is, while Harvard Avenue was flattered by bias opening day, he could not have been a bet-against on Wednesday in such a suspect field. Perhaps horseplayers are worrying too much about track surface and too little about the basics of handicapping.

Nevertheless, after just two weeks, there has been a rush to judgment on the nuances of Cushion Track.

Another example was the apparent domination during opening week (Nov. 1-5) by horses that trained over the surface. The first week, 34 races were run on Cushion Track. Horses whose last work was on Cushion Track won a gaudy 31 of those races; horses who last workout was on a dirt track won only 3.

The instantaneous reaction was predictable. Knee-jerk handicappers would prefer horses that had worked on Cushion Track.

The data is relevant only in the proper context - percentage of starters. It turns out that the impulsive conclusion during opening week was correct. Cushion Track workers made up 72 percent of the starters and won 91 percent of the races. Non-Cushion Track workers made up 28 percent of the starters and won just 9 percent.

Put another way, there was one winner per 6.25 Cushion Track work starters, and one winner per 25.6 non-Cushion Track work starters. (There were 194 Cushion Track workers, and 77 others.)

The data suggested that horses that had trained on Cushion Track had an advantage over horses that trained on dirt.

However, there was reason to reconsider after the second week of racing (Nov. 8-12). Perhaps Cushion Track-trained runners dominated opening week merely because their trainers were stabled at Hollywood and pointed to the start of the meet.

The second week of racing revealed only minimal advantage to Cushion Track workers. As other horses filtered onto the programs, winners spread out. During the second week, there were 28 races on Cushion Track. Horses that last worked on the surface won only 18 of them. Horses whose last workout was on a track other than Hollywood won 10 of the 28.

Again, the relevance depends on number of starters, and the percentage of Cushion Track workers decreased. Locally trained workers made up 57 percent of the runners and won 65 percent of the races; non-Cushion Track workers made up 43 percent of the starters and won just 35 percent.

If there is any edge to Cushion Track works, it decreased radically. In the second week, there was one winner per seven Cushion Track work starters, compared to one winner per 9.7 non-Cushion Track work starters. (There were 127 Cushion Track workers, and 97 others.)

While deliberating the surface on which horses work might require re-evaluation, another pre-meet idea has proved valid. That is, the notion that turf horses would hold form on Cushion Track because of less kickback or slower pace.

It is true. Turf horses have performed admirably. There were 73 Cushion Track runners the first two weeks whose previous start was on grass. They won 11 races (one winner per 6.6 starters), which is higher than random sample (average Cushion Track field size is 8.0 starters per race).

With minor past-post tweaking, the turf-to-Cushion Track angle jumps sky-high. By eliminating high-odds runners that were overmatched or off form, "well-meant" turf runners produced an unusually high win rate the first two weeks.

There were 46 turf horses that started at odds of 9-1 or less in Cushion Track races, and all 11 turf-to-synthetic winners fell into that odds range and produced a $2.70 return on investment for each $2 win bet.

The trend continued Wednesday. Three more winners went turf-to-Cushion Track - Top of Atlantra ($3) in race 3; Riley's Life ($12.40) in race 5; and Jump Rope Girl ($6.60) in race 6.

Two weeks of data is too little to conclude that all turf runners will like Cushion Track. Nor does two weeks provide sufficient evidence that locally-based runners have an edge, or that bias-aided runners such as Harvard Avenue should be wagered against when they run back.

Perhaps the only lesson from early in the Cushion Track era at Hollywood Park is that any rush to judgment can get a bettor in a heap of trouble.