Point well made. I just wanted to clarify that what I've learned from sabrematricians is that steals in it of itself is not something that leads directly to winning baseball. The steal will only work for a team in terms of winning when it comes at a high percentage success rate. I actually like the argument. But when talking about "speed", that is an entirely different matter than steals (which is a residual effect on the plus side of speed in baseball). I don't think Beane is against speed as much as the misuse of it to win ballgames (ie: steals).
One problem is that traditional statistics don't reflect speed as well as they could. It would be very interesting to see extra bases on hits and advancing on outs tracked for base runners.
Another problem is an example of taking an argument (wait for it, Noe)
out of context. When books about statistical analysis were first published, one of the arguments raised was that teams were hurting themselves by favoring speed at the top of the order over all other characteristics, such as getting on base. This was not an argument that speed was meaningless or bad
per se, but that speed was not very useful if the player possessing it couldn't reach base very frequently. Over time, the context of the argument was removed, and it just got shorted to
speed is not very useful. The condition to the argument,
if the player possessing it couldn't reach base very frequently, was omitted, and the argument ceased to have any logic to it.
