File this one under weird math, unless I’m missing something. David Pinto looks at Denard Span’s PECOTA projections for OBP and sees a man far outperforming expectations at the major league. Is this a sign that Span will regress?
Pinto then looks at Span’s Triple-A numbers earlier this season and sees that they match what he’s doing at the major league level, suggesting he’s taken a great leap forward. So far, so good. This is the essence of player projection, at least the verbal kind where we identify interruptions in the data.
But he then does some math showing that there is virtually no chance that PECOTA’s evaluation is correct, based on Span’s breakout numbers this year. I’m inclined to accept the verbal argument that Span has improved his approach at the plate and so is getting on base more, but have a hard time accepting as proof this year’s sample.
That’s why we compare small sample sizes to larger ones, as a reality check. And when the small sample is out of line with the bigger one, we should be very skeptical.
Notably, this year, Span is walking more than he has in the past, in Triple-A and the bigs, but he’s also getting on base much more on balls in play, and a significant part of his OBP is due to his BA. It looks to me that at the end of the day (and not necessarily this season) he should end up a .360 OBP hitter. His .399 OBP now is as much the result of good fortune as improvements in his game (though it looks like there are some of those, too).