Monday, August 27, 2007

The Fickle Finger of Fate - A Cautionary Tale

After the feast, the famine... Having enjoyed such a tremendous weekend courtesy of a certain racehorse named after a tree, winners have been somewhat more difficult to come by this week.

For followers of TrainerFlatStats or TrainerTrackStats, yesterday exemplified the current frustrating run, with second placed horses at 9/1 and 17/2 both looking like winners before being beaten a neck and half a length respectively.

Which got me to thinking about the margins by which we define success and failure. It is a thought which recurs to me on a regular basis, and is one of the main reasons why I love (and hate) this fantastic sport.

The thing is that we are paid out in a binary system. Its black and white, with no grey. You win. Or you don't win. Simple as that. Trying to win consistently is an exacting discipline, and it doesn't suit all temperaments.

As regular readers will know, mine is generally a systematic and statistical approach to betting.

I don't really believe in collateral form. The notion that Horse A beat Horse B by two lengths, and Horse B beat Horse C by two lengths, therefore Horse A must beat Horse C by four lengths, is - frankly - pseudo-science at its worst.

There are always so many imponderables that its nigh on impossible to conclude with any degree of certainty that Horse A will beat Horse C. Added to this is the undeniable truth that within any horse race a form selection will run inexplicably badly, and another will run inexplicably well. Inexplicable in terms of collateral form, that is.

To use the Silver Birch example, I made this horse about a 14/1 chance. I didn't bet on the day as I was already on. The first I learnt of his huge SP of 33/1 was when it flashed up on screen after he'd passed the post in front.

On collateral form, in all probability, he was more akin to a 20/1 or 25/1 short, but in terms of the statistics he had a much greater chance than that.

This brings me nicely on to my point (and not before time, I hear you cry!). Because not one of us has ever had just one bet, we should not be looking to win every time. Rather, the nature of the game is to stay ahead in the long run. So even if the extremely brave and slightly unlucky McKelvey had bettered Silver Birch in the shadow of the post, I would have remained convinced that my selection policy is correct, and that the fickle finger of fate was raised firmly in the vertical, presenting me with the one (fickle) finger salute.

I would have won no money for my vindication. Indeed, I would have lost money. But I remain convinced that - in the grander scheme of things - I will come out in front.

This is the central pillar of my statistical approach, and its the reason why I don't necessarily think my approach (and that which I champion in my statistical guides) is for everyone. It takes discipline and a broad overall perspective to ride through the inevitable troughs that will occur.

But 'when the battle's lost and won' (apologies to the Bard), I am confident that my cup will be full of cheer, even if it doth not runneth over.

p.s. A quick mention for my trainer Julia Feilden, who sent out her first juvenile runner of the season 'Spirit of Sharjah' yesterday. Backed in from 16/1 to 10/1, it fairly pis...... hosed up, in a time under a minute (that's 12 sec's per furlong, or very very fast for a 2yo debutante in other words!). She reckons she's a couple of others as good as this one, and I've emailed her to find out their names. I'll endeavour to keep you posted in due course.

p.p.s. Still on the subject of Julia, my (bit of a) horse, Rapid City, runs on Saturday at Newbury in the Spring Cup (25,000 to the winner). He's possibly drawn on the wrong side and its a very hot race, as you'd expect for the money. Also, first UK race on turf. We're hopeful rather than expectant of a good run, but Julia reckons he's been crying out for this decent ground and has been running well on the sand, despite not really liking the surface. We'll see....

p.p.p.s. Teofilo Update: currently trading at between 4/1 and 9/2 with betfair. Or you can have 7/4 with William Hill. (Are they for real?!) My money is still on him sidestepping Newmarket, in favour of a shot at the Curragh in the Irish equivalent. Again, we'll see...

Matt Bisogno is a lifelong horseracing and betting enthusiast, and has published a number of statistical analyses of trainer patterns for horse racing betting purposes.