Odds Better at Start of Season?

[Polydamas]

I've been working on a crowd-sourcing predictions project with a friend, using football to test our results (the initial system was taken from political science). Basically, we built an app / website that gets people to put in the favourite's chance of winning, from 0 to 100%, and then averages that to get the crowd prediction. Comparing that to available odds tells us whether to back or lay the bet. We ran a test for the World Cup last year, and for the first few months of the last Premier League season. In both cases the same thing happened. 1) We had great returns for the first couple of weeks 2) The returns flattened out after that I think there might be two possible causes for this: A) The quality of our predictions dropped off as our player-base decreased (the games had a lot of teething and design problems, making it boring for anyone outside the Top 10) B) Bookies' odds are less accurate predictors at the start of a tournament or a new season, when punters are working with less information, but there's less opportunity as the season progresses. We've decided to start up the test again for the new season, hopefully with some changes that keep the player-base from falling off as much, but I think it's likely that we'll see a similar pattern to last year. We seem to be able to avoid the big losing bets, but the more profitable bets seem to dry up after a while. As neither of us bet regularly, I was wondering if anyone here had a view on B above. Is the start of the season the best time to make money off the bookies, and does that decrease after a couple of months? Any advice is much appreciated.