from Amazon
Tuesday, April 27, 2010
With or Without You - At the win/loss game level
This is for the NHL, but the idea is the same: what is the team’s W/L record with and without the player playing. This is not a new concept of course. And, you guys know I use it for fielding. The question is to ask when should you use WOWY and when should you not.
When you look at a team’s W/L record, there is an enormous level of noise in there, especially when we are talking about 5 or 10 “without” games. Simply put, the pool of games “with” and the pool of games “without” will differ by far more than just the player being targetted. The rest of the team, the opponents and the sample size will conspire against you.
Where I first used WOWY was on the catchers-pitchers pairing on baserunner events (SB, CS, BK, PK, WP, PB). In this case, pairing Gary Carter with Steve Rogers and all other catchers with Steve Rogers gives you two pools that are very similar (outside of the targetted player). You have first of all the pitcher, who is one of the two huge influences. We have no expectation that the runner quality will be biased one way or the other. And we have a huge sample size of baserunners to begin with.
So, when you do WOWY, you are looking for these things. Going back to hockey, the obvious place to look at it goals scored and allowed when the player is in the game or not. But, you can look at other events, like shots (or shot-location). And, you can drill even deeper, and instead of looking at whether the player was in the game, you can see if the player was on the ice. In hockey, you still have a pairing issue, since teammates are not distributed randomly each time a player is on the ice.
In any case, the point is that the noise and uncontrolled parameters have to be handled, somehow. The linked piece above, while interesting as a story, is limited to being just that: a story with no conclusions possible.
Recent comments
Older comments
Page 1 of 344 pages 1 2 3 > Last »Complete Archive – By Category
Complete Archive – By Date