For the first time in just about a week, I had a pretty bad day. My overall results were 16-18. My more confident results were 12-13, and the rest were 4-5. On the money line I was 5-6, against the run line I was 7-5, and I was 4-7 on the over/under.
Just really bad all the way around.
I was thinking about the whole exercise of trying to predict the winners yesterday, and something occurred to me. The way I'm treating this, as a mental challenge, is quite different from how someone actually betting would do it. I think about the game, and try to figure out who will win, by how many, and how many runs they will score. My goal is to get as many of them right as possible. When you're betting, though, there is a risk/reward to each proposition. If you could predict with 100% accuracy, then of course just picking the winners would be enough. But, in reality, you need to compare the odds that you think a team will win, or score a certain number of runs with the odds that the bet gives you. Trying to maximize the number of games you pick correctly shouldn't be the goal. If, for example, you went 10-10 picking all favorites, you would lose money. If you went 10-10 picking all underdogs, you would make money. What you really want to do is maximize the expected value on each proposition.
In light of that, I started thinking about how I would change my picks yesterday if I was going to bet. I did this before the games started. And, strangely enough, my picks based on my EV judgements did a little bit better, even record-wise. I would have been 18-18 overall, instead of 16-18. More importantly, on my "more confident" picks, I would have been 15-11, instead of 12-13. That is huge.
So.. anyway.. for now I am going to continue trying to just predict the outcome of the games, but I will try to put notations where I think an EV pick would differ from a nominal pick.
