Racing: Figuring out winners the Whitley way

Punters are turning to a Yorkshire-based computer analyst in the quest for winners.

THE THICK, smartly-bound book which dropped through several hundred letterboxes last week is not - if you will excuse the pun - the most obviously racey of reads. Computer Racing Form, an annual production by the organisation Racing Research, is an analysis of the 1998 Flat season, crunched into 467 pages of closely-printed names and numbers. Michael Crichton it is not.

To its devotees, however, the annual, and in particular the daily ratings and racecards produced by John Whitley, Racing Research's founder, are as essential a part of the process of betting as a pen and a blank slip. Whitley, a mathematics graduate, was an employee of Timeform for nine years, until he left in 1983 to develop his own ideas about handicapping and speed figures with the help of some serious computing power, in Brighouse, West Yorkshire.

Many punters believe that finding winners is an art rather than a science, a matter of instinct and emotion rather than cold analysis. Whitley begs to differ. "Our methods are entirely objective," he says. "We have a database which is basically all British horseracing results, and programmes operate on top of this database, which in the case of jumpers goes back to 1993, to produce collateral form ratings as rapidly as possible. We do a complete re-handicap every day, usually overnight because the programmes take several hours to run, even on modern computers."

For those who struggle to set their video, a rough translation goes like this. Each day, Whitley feeds the results of every race in Britain into his computer. The machine then looks at what it already knows about the horses concerned, and all those that have raced against them since 1993, in the light of this new form. It then decides how relevant the information is - recent form, clearly, being more important than old - and reassesses its opinion of every horse in training accordingly. The same process would take a human with a pencil and paper years, probably decades.

The end results are speed and form ratings for every run by every horse, which are then supplied to Whitley's clients in the form of racecards, for a fee varying from pounds 6 to pounds 3 per meeting depending on the length of your subscription. Some of his clients have taken every racecard since Racing Research was founded 15 years ago.

"We look at it as raw material for people who are already racing experts," he says. "We don't say there's anything magical about them. They are a starting point before taking into account all the usual sophistications which relate to any race."

In other words, considerations such as distance, going and track must still be borne in mind. Nor, this being racing, are there any guarantees. There are, however, many racing professionals - including bookmakers - who always turn to Whitley's ratings before they open the form book. At the five-day stage before the Tripleprint Gold Cup at Cheltenham in December, for instance, they found that Northern Starlight, who was available at 20-1, had a much stronger chance according to Racing Research than any of the other entries, including his better-fancied stablemate, Cyfor Malta. On the day, Northern Starlight won - Cyfor Malta was a non-runner - at 15-2.

Last Saturday's Tote Gold Trophy also looked very different to Whitley's subscribers. "According to our figures," he says, "Decoupage should have been the favourite, and Tiutchev [the actual and red-hot favourite] should not have been in the first three in the betting."

Other subjects to come under the cold scrutiny of Whitley's computer analysis are the effects of the draw and the relative merits of jockeys. Kieren Fallon, for instance, was regularly rated among the best in the country, long before he became the champion jockey.

Whitley, though, is never entirely happy with the numbers his computers give him. "My interest is in developing the mathematical side of the systems," he says. "We add algorithms to the ones we already use, which improves the sophistication of what is already there." Racing and betting will never be an easy way to make a living, but for those who can afford it, Whitley's analysis at least ensures that they will walk into Tattersalls with several thousand pounds-worth of silicon on their side.

Racing Research 01484710 979.