Hi all. I cant seem to get my head around this one. I will try and write it as it was explained to me.

 

Is it possible to work out the Probability of e.g Who will win a race based on an individals odds measured against each others, in decimal.

 

i.e

1. 9.9%

2. 32%

3. 19%

 

Their odds are as above. Looking at that, to me the probability of 2 winning is high. But his odds are based on previous form, and the one we expect to win, doesn't always win.

 

Is it possible to work out the probability based on their previous form and that of those on the field around them.

If I understand you correctly, then I don't think that's enough information to do what you want. Those percentages you give represent the portion of races each has won, right? But it's against different competitors, meaning the numbers are not directly comparable. I could win 100% of the time vs. a bunch of 2 year olds, and you could win 0% of the time vs. a bunch of olympic sprinters, but that doesn't mean I have a 100% chance of beating you.

Sure it is. Use Bayes' Law. Those prior winning percentages are the prior probabilities you need to apply that law. Wikipedia link: http://en.wikipedia.org/wiki/Bayes'_theorem.