from Amazon
Friday, December 17, 2010
Reliability of minor league data
Bill James said this:
Do you still think that minor league stats are as reliable as major league stats in predicting future performance, or have you revised your view? Asked by: Anonymous
Answered: December 16, 2010My best answer, I think, is that I haven’t studied that issue in the last ten years, so I have had no reason to re-consider my opinion. Also, for the sake of clarity, I think you left out two caveats:
1) If interpreted correctly, and2) Assuming that the player receives major league playing time.
I replied to him:
minor league stats: perhaps this is semantical, but are you suggesting, with your caveats, that 500 PA from a minor league line is as reliable as 500 PA from an MLB rookie in forecasting the following season’s MLB season? You do have to be careful that there is going to be a selection bias, and I think your caveat introduces that. Put simply: if a rookie gets 400 or 500 PA, then he probably had a pretty good season. If he had less, then either he was unlucky or he wasn’t really that good to begin with. So, your caveats can actually introduce a selection bias rather than imposing a fair test condition.
Taking a guess, I would say the reliability of minor league data would be equivalent to MLB performance from one prior season. So, I would say the reliability of 2010 minor league data in forecasting 2011 MLB performance would be as reliable as 2009 MLB data. Or even perhaps 2008 MLB data. (Given the same age of players.) The 2010 minor league data simply can’t have the same reliability as the 2010 MLB data in forecasting 2011 because there is a larger uncertainty around the 2010 minor league data.
Now, maybe Bill meant that is was pretty close and I’m splitting hairs here. And that’s fine. I’m more interested in the topic itself rather than debating the correctness of Bill’s point.
Recent comments
Older comments
Page 1 of 344 pages 1 2 3 > Last »Complete Archive – By Category
Complete Archive – By Date