Horseracing Betting: Why creating ratings is fun

Betting Strategy / Jack Houghton / 16 May 2009 / 1 Comments

Jack Houghton reflects on a formative trip to Market Rasen and explains why creating ratings is essential and fun...

Growing up, Boxing Day presented a simple choice: Huntingdon or Market Rasen. Choosing the latter meant spending hours rounding the Wash before traversing the Lincolnshire Steppe. So Huntingdon usually won out. But one year, inexplicably, the Rasen was chosen.

And despite, as a then ten-year-old, having already gone racing hundreds of times, the day stuck in the mind. On the journey up, Dad gave me a tenner and said: "See what you can do with it."

It can't have been the first day I'd had a bet. But it's the first I remember, for whatever reason. Maybe it was because it was also the first day I used ratings to make selections. Nah. The more likely answer is: I turned a profit. So on the journey back, on the flat surface of the down-turned glove box door, in our beat-up, dark red, Volvo estate - registration PCC 74S - I counted and recounted piles of coins and notes, and the obsession was set.

However, the whole ratings thing started that day too. Although the details are now sketchy, I know I'd been reading a book - presumably one of Dad's - that talked about the importance of ability and recent form. To quickly assess any race, the book suggested, you should look at the last two runs of each horse and give them a number of points according to their finishing positions. The exact points system has deserted me, but I remember the maximum any horse could score was 13: five points for each win, two points for a previous course victory and one point for a previous distance win.

Armed with this new mathematical approach, the Rasen card that day was at my mercy. Simply find the highest scoring horse in each race, and back accordingly. And although, looking back, I cringe at the obvious shortfalls in logic of this rating system, it worked for me that day, and set me off on a ratings road for life.

Now for the point of this blog: create ratings. Even if they don't help your punting at first - and, in my case, taking Market Rasen as the start point, I guess it took nearly 20 years to settle on something that worked consistently - the process of creating them will provide an intellectual stimulation far in excess of anything Suduko can muster and, with luck, the profits will soon follow.

But forget about the potential prize. Just focus, Zen-like, on the process. An ex-colleague, a two-year-old specialist who creates his own vast array of ratings to guide his punting, and has been profitable each and every season post-decimalisation, once told me he enjoyed the intellectual game of race prediction first and foremost; his profit and loss was merely a way of keeping score. As a mindset for profitable punting, this would be hard to top.

Creating ratings is fun, honest. Take my snooker ratings. In the 2004/05 season I spent hours putting together a spreadsheet GCHQ would struggle to decode, concatenating the results of the previous five seasons. I was using an adapted Elo rating system to assess the ability of players and a new-fangled Monte Carlo simulation model to create tissue prices for each and every match and tournament. During early season paper-trading, the ratings performed exceptionally and, come the 2005 World Championships, confidence was brimming as I started betting with real money. I lost a lot of money that year.

And so began a four year process of remodelling and tweaking: changing assumptions to favour recent form, then changing them back again; considering the margin of victory in each match crucial, then stripping it out as unimportant; incorporating a revolutionary measure of tournament importance, before enacting counter-revolution; each time confident of having found the answer, each time being ultimately disappointed.

The result? At the close of this year's World Championships, totalling all my snooker betting driven by ratings, I am showing a loss of £32.63. And I had an epiphany when O'Sullivan was knocked out this year. Elo is the wrong approach for snooker. The pool of top players is too small and they don't play each other regularly enough. It should have been obvious: Elo was never going to work! So I'm starting again with something completely different. And I'm brimming with confidence.

You see, I told you this ratings thing was fun.

Comments (1)

1. Christopher Small | 26 June 2009

   I think you'll find your Market Rasen ratings wer based on Clive Holt's Fineform Formula. Published I think as Winners back Winners.