Typing a Hitter

Tom Tango reviews Radial wOBA, but makes an important point about wOBA in general:

But, is that the most interesting thing we can do here? What would you prefer, a hitter’s batting average or the individual components of 1B, 2B, 3B, HR? How about his slugging, or the 4 individual components? How about wOBA or his 4 individual components?

Some people may suggest they’d prefer the wOBA. And that’s certainly possible and understandable. Because sometimes you don’t want to know more. Whenever you present a single number encapsulation of a set of components, the conversation ends. What you will have done is taken a series of data points, merged them into one, and… present it. It just ends there because you can no longer unravel it otherwise.

I like numbers that draw a picture of a player, which is why I prefer slash lines to wOBA. If I want a single number for a player, I’d rather see something like runs created per game or wins above replacement, because to me, those are more useful numbers. If I have nine players and their runs created, I can approximate how many runs a team should score. If I have WAR, I can approximate how many games should win. wOBA does provide a nice measure of the quality of a player, but it’s a black box quality. I want to know how the player gets there.

Back to Radial wOBA for a minute. Add a second radial component, the directional angle, and you have the basis for a good probabilistic range system.

from baseballmusings.com http://ift.tt/2iayRI6