Originally Posted by

**twinsfan34**
Do you have some examples of some 'fact based models'...? Perhaps you have a working definition that should be shared. It's fine to point out someone is wrong, but if you don't teach or show, then what value have you really offered?

Can you provide some sort of statistical analysis to aid his case? How about a simple linear regression model based models and optimization based on similar catchers.

Do you know how lucky Salty was this year? His BABIP was .372. So when he actually got the ball in play (roughly 70% of the time, 30 K%) he did pretty well. But do you know how likely he is to put that up? I mean, without even using Bayes Theorem or some other likelihood of events reoccurring formula.

In 2012, his BABIP was .265. In 2011 it was .304. For 2009 it was .320. But, remember, he doesn't put that ball in play that much. 30% K rate...

So Salty swings and misses. A lot. But, he also only plays 120g a year, well only twice (sorry, not a fact based model, just a fact). And by definition, he hasn't had a full season in 7 seasons based on plate appearances (501 needed). I struggle with the idea that someone could give a guy 12.5%+ of your payroll (for an $80M payroll, e.g. Twins) for the next 3 seasons with that kind of production or cost. If he cost 5% of your payroll ($150M total, e.g. Red Sox), sure, it makes more sense in terms of payback ROI.

I do stats for a living. I've taught AP Calculus for years before switching into industry to do analytics. Instead of saying something arbitrary that has no real data points or relevance. Please show me your statistics points versus your opinions and projections on me. This isn't SABR where everything has to be a statistic either. I'm a member, so if you are too, maybe we should take our SQL, Excel, R, and head over there instead?

And then, just acknowledge what *did happen*. Salty did sit for 4 of the 6 World Series games. I don't know how that's not an important stat. As Herm Edwards said, "You play to win the game!"