Arithmetic Value System using Favourite-Longshot Bias

[Joe]

Since I've added a little bit of analysis to my website, I though I might as well proof this system properly. Basically, I'm going to do what Betbrain and Tip-ex do by way of odds comparison, but do it properly. They provide information on value bets (in addition to the much rarer sure bets). The way they compute these, however, is flawed because they don't take into acount the pricing bias in the odds which I strongly believe exists. The upshot is that my analysis will avoid throwing up loads of high priced away wins as possible winning edge bets since back testing quite obviously confirms that these are not profitable bets in the long term. My home page will be doing one best value bet every weekend, but I might add a few more for proofing here. The long term yield is not going to be large if the system is accurate (in the region of 1 to 3%), but I'm more interested in finding out if my hypothesis is correct. I believe that the system I proofed on here last year (if any of you remember) which basically did a similar thing demonstrated considerable viability in this respect. Anyway, to get the ball rolling..... Juventus to beat Como, 1.65, Expekt. P.S. There is absolutely no match analysis whatsoever here, simply odds comparisons. Obviously the more bookies one uses the better, although I think if you have too many silly bookeis in there it might bias the results. So therefore, I'm going to stick to about 10 fairly reputable firms, since I think this should be an adequate number to demonstrate success or otherwise.

[Guest teeshappy]

Re: Arithmetic Value System using Favourite-Longshot Bias Hi Joe, Interesting stuff, I'll be tracking your results. A quick question, do you base your results on working out the Favourite-Longshot Bias by using the historical odds data for each bookie? Or is it just an average of all the bookies? Cheers dave b ps/ I heard Juventus are still recovering from the flu! But they won anyway!