As rosters are finalized and spring training stands just around the corner we’ll start to see many projections. The first overall one I’ve seen with wins and losses is Clay Davenport’s. Go see his here.
They all look pretty reasonable to me, at least insofar as how each team finishes in its division. I’m not a numbers guy so I can’t pick nits or say things like “Only 85 wins?! Bah! I think they win 89!” Go read Clay’s methodology. It’s complicated.
And I’m glad people like Clay do these because it helps to put actual past performance and numbers and things on top of all of our offseason feelings and expectations. Aging curves and past performance being the best predictor of future performance are concepts that are easy (and frankly, fun) to forget when your team is having press conferences putting its jersey on its new free agent.
My favorite thing about projection season, though? How mainstream baseball writers will deride them as silly as they come out. They’ll drop barbs on Twitter and in their columns about how, since someone projected wins and losses for the league “there’s no reason to play the season now!” or some such. The disdain for projections and statistical analysis in general comes through pretty clearly in this stuff.
And then, come March, these same guys will run “predictions” columns. Based on far less data and far more unverifiable and unfalsifiable conventional wisdom and pure gut feeling. And rather than predict things like “generalized results over a sample of 2000+ games,” they’ll predict who out of 30 teams will win a seven-game series six months in the future. And claim their own expertise as the basis for taking them seriously.