Making My Betworm Handicapping Tool

Traditionally the British seem to be incredibly skeptical of anything that they even remotely perceive to be a system. I’m guessing this was because lots of unscrupulous people tried to sell a load of rubbish in the past and nobody had access to the extreme level of information we have today. Systems range from just backing the ones with the big ears all the way through to complex handicapping algorithms. 

Whilst I was watching an American podcast featuring @tomandtomracing I noticed something interesting. Many of the tipsters featured on the show were using an algorithm (a complex mathematical formula) to establish the winners of the races. I was completely taken aback. In Britain this would be seen as very unusual as most tipsters are research and intuition based. 

It occurred to me that as an ex-bookmaker with a full understanding of the subject of horse racing I should be able to produce something similar myself. 

My idea was to create a formula that thinks like I do! 

Because I have never been one to make things easy for myself, I decided to try and design it for National Hunt Jumps racing. I have never heard of anyone designing anything quite like it for jumps racing but I'm very sure I'm not the first judging by some of the unusual betting patterns that I see. 

There are many things to consider when making something like this. Which stats to use? In what way do I use them? How do I tie them together. How do account for progression? I wanted to know that the formula would have an answer for all of my questions before I put it to use. 

After many mental gymnastics I had a version that I used for Cheltenham races with great success. After Cheltenham I made a modification to deal with novice races and the strike rate improved. I tested Betworm over the course of many races and discovered it was best at handicap chases. There are reasons for this I am not going to go in but I think I know how to modify it to understand stakes races. I am now on version two which takes extra information regarding the ground and the course in to account. 

I like that the betworm tool does not know what the odds are when it makes the selection. It has no memory of an enthusiastic trainer saying this was one to watch and it never takes notice of the hyperbole. Betworm looks at the facts. It does not do whimsy, conjecture or procrastination. It is designed to handicap the race it is presented with. The tool looks at the amount a competitor will have to improve to beat the favourite. 

I am happy that Betworm has a good enough strike rate to tip and back the horses it recommends.  

I am currently running both versions of Betworm together as I know that the original is good at what it does. This allows me to compare the results to firmly establish one is better than the other. 

I imagine there will be many more versions over time. I really enjoy working on the Betworm Project. 

I will be writing more on this subject so stay tuned. 

Why not check out some of my previous blogs and see if anything catches your eye!