There is an old saying that if you read something enough you will end up believing it. Well, in the example of horses 'action' it is almost certainly the case. ​ Many years ago, I kept hearing about the 'knee action' of a horse and, being statistically minded, I wondered if it was a variable I could exploit and add it to those already used in my modelling. I went about building a database of horse conformation/skeletal structure/physical action and a few other related variabl

Let's start with the basics. Toss a coin and there will be a 50% chance of heads and 50% chance of tails. Over a long period of time the number of heads and tails will level out. Assuming you have single bets if you take Evens (2.0 decimal price) you will break level, take 10/11 (1.9) you will lose, take 11/10 (2.1) you will win. If you grasp that read on! Now using a 6 sided dice, each number has a 16.67% chance or ‘5/1’ (6.0 in decimal odds). In a nut shell, bookmakers will