Using statistical trends to make profits from betting is my chosen path. The stats behind my bets are a great mental support to me during losing spells. My faith in their long-term profitability carries me through temporary setbacks.

This is especially important to me because my betting strategy is based on profitability rather than strike rate. Suffice to say my strike rate hovers consistently around 25 per cent and this translates into fairly lengthy losing runs.

It also means profits come in sharp bursts. For serious betting, I need the backing of my statistics.

But what is a statistical approach? The best way to explain is by example:

• Do colts/geldings have an advantage over fillies?
• Do blinkers have a positive or negative effect on performance?
• How important is a recent run?
• What advantage, if any, does a previous course winner have?

• Colts/geldings do have a significant advantage over fillies.
• Blinkers have a negative effect on performance.
• A recent run is very positive.
• Course winners are more likely to repeat than non-course winners.

You could probably have guessed the answers but statistical analysis allows you to measure the exact significance of the variables. This measurement is known as an Impact Value (IV) and is obviously useful to know.

However, more important than the IV is the Profitability Measure (PM), which demonstrates to me whether I can make use of the knowledge highlighted by the IV.

An example: The control I use to measure against any other variable is a strike rate of 8.8 per cent and a level stake loss of 29 per cent of all stakes invested. These are the figures you could expect to achieve if your selection method involved a newspaper and a set of darts!

So how much of an advantage do colts and geldings have over fillies? To determine this (UK stats) I looked at the results of all handicaps over a representative period of time. The results:

• Fillies, strike rate 7.65 per cent, Profitability Measure -38 per cent.
• Colts/ geldings, strike rate 9.36 per cent, Profitability Measure -25 per cent.

Clearly, colts and geldings have, in general, an advantage over the fillies. If you backed every colt and gelding in a handicap you would lose 25c in every \$1 bet.

On recent runs: The stats show that a recent run is a vital statistic, and the more recent the better.

 Last run 3 or less days: Strike rate 12.5 per cent PM -16 per cent. Last run 4-7 days ago: SR 11.2 per cent PM -25 per cent. Last run 8-15 days ago: SR 9.8 per cent PM -27 per cent. Last run 16-31 days ago: SR 9.0 per cent PM -27 per cent. Last run 32-47 days ago: SR 7.8 per cent PM -29 per cent. Last run 48-79 days ago: SR 6.8 per cent PM -32 per cent. Last run 80+ days ago: SR 4.7 per cent PM -42 per cent.
Contained with the 80+ days category are horses who last ran 548+ days ago. In the past seven years there have been 1436 such runners in handicaps and claimers (200 per season) and only 27 (4 per cent) have won, resulting in a loss of 72c per \$1 bet.

Which reminds me of a story about a compulsive gambler in America. "I passed the racetrack today, it was closed, so I pushed my money under the gate," he is reported to have said.

Stats for course winners show they enjoy an advantage, but if you rule out poor-class Maidens the stats for no previous course wins show a strike rate leap from 3.9 per cent to 9.6 per cent.

These statistics on their own are not immediately profitable, but they do provide an excellent basis when linked with other statistics. For example, when you link a recent run 4 days or less with 2yo's only, the strike rate rises to 18 per cent and the PM to a remarkable +37 per cent.

Similarly, if you link recent runners with only those horses beaten less than a length, the strike rate is 22 per cent and the PM is +11 per cent. Alternatively, if recent runners are linked with winners only of over one length, the strike rate is 22 per cent and the PM measure is +13 per cent.

This article was published with kind permission of the author. We will have a Russell Clarke special feature on 'betting psychology' in the December PPM.

EDITOR'S NOTE: Contributions for this page are welcomed. They should be about some aspect of racing and betting, preferably with a personalised angle to them. Length should be around 750 words. Payment is \$100.

Send your submissions, with a covering letter containing details about yourself to Brian Blackwell, PO Box 36, Isle of Capri, Qld, 4217.

By RUSSELL CLARKE, a UK professional punter

PRACTICAL PUNTING - NOVEMBER 1997