How about this, if you were standing on a corner with a jar filled with Jelly Beans an you wanted to find out the number of jelly beans in the jar, the best method is to ask as many ppl as possible to make a guess, you would then find the average of all guesses,the answer will will be pretty close to the actual number of jelly beans.

Now take it another step, the Tote is the group consensus of the chances in each race, which in reality is pretty close to what happens, when you make adjustments for breakage the strike rates in each percentile group is as good as it gets.

Now you compile your own probabilities/ratings an you find there is a variance between what the Tote probs(adjusted for breakage) an the actual.

You now measure the variance between yours an their's, but now you break it down further by Post Position an by percentile group(based on your probs).

So now when you complete your Ratings you apply the above variances to your ratings an you now have a set of probs based on the past history of group consensus.

So now i'm sitting at home in front of my computers and i now see in an upcoming race a runner who's offical tote payoff is considerably lower then the group consensus payoff(your ratings adjusted by variance), which is lower again than my original ratings.

Now what do you do?

Answer it correctly an you will start swimming with the sharks




[ This Message was edited by: woof43 on 2004-11-01 16:44 ]