Two days, no comments.

This is clearly not a sabremetric site, if no one is reading and commenting on a well written discussion of the amazing forecast that Oliver gives Yu, and the inherent difficulties of forecasting Yu's performance, given the limited sample size of and pool of com parables.

This is clearly not a sabremetric site, if no one is reading and commenting on a well written discussion of the amazing forecast that Oliver gives Yu, and the inherent difficulties of forecasting Yu's performance, given the limited sample size of and pool of com parables.

I just read it, and honestly have no idea what to say.

I don't know enough about building projection systems to really comment besides saying it's a good article and an interesting problem.

This is clearly not a sabremetric site, if no one is reading and commenting on a well written discussion of the amazing forecast that Oliver gives Yu, and the inherent difficulties of forecasting Yu's performance, given the limited sample size of and pool of com parables.

I was just thinking, who cares. It's a back of the rotation pitcher for a team that made the World Series last year. Great, they got Edwin Jackson, what a deal. Don't care what the projection says, the team should be happy if they get 200 innings out of him with an era+ of 100. Anything more is just ignoring history.

After the Wieters fiasco, any time I see a projection in which the player who has no major league experience is projected to be anything other than average, I automatically adjust the projection down to average. I'll let Oliver project ten players who have never been in the majors as above average, and give him 5-1 odds on each one and feel comfortable knowing I'll end up ahead on the deal.