seventieslord
Student Of The Game
I understand the DTWOM is based on expected goals by each team. But I'm not clear on whether it takes score effects into account. A team that goes down by 2 early and spends the whole game throwing everything at the net could end up 80% on the DTWOM. But I'm not sure a team that does that actually wins 80% of the time.
So I'm wondering, has this model been checked against real results? Surely there have been hundreds of games, if not thousands, where the DTWOM said one team was up at 80%. That's in theory. What about in practice. Do those teams actually win 80% of the time? If not, wouldn't it be a good idea to calibrate these likelihoods to reflect reality?
So I'm wondering, has this model been checked against real results? Surely there have been hundreds of games, if not thousands, where the DTWOM said one team was up at 80%. That's in theory. What about in practice. Do those teams actually win 80% of the time? If not, wouldn't it be a good idea to calibrate these likelihoods to reflect reality?