1. We don't know if his career average is his true talent
I very much doubt his career average is his true talent, which is what I was trying to demonstrate. The supposition was a hypothetical one.
2. The standard deviation should be a lot higher than the one you used because the sample size of is smaller. If you look at the formula for Standard deviation, N is in the denominator so the lower the N the higher the standard deviation.
The standard deviation I used is taken from a sample of hundreds of thousands of major league at-bats. N becomes pretty much irrelevant at that point. Did you read the article that I linked in my original post? It explains the methodology in detail. I'm not comparing Ruben Tejada to himself--I'm comparing him to the average major leaguer, and trying to determine how likely it is that his BABIP is "real." I think anyone could tell that it was a bit luck-inflated, but I don't think people realize just how unlikely a "true" .387 BABIP is--I didn't until I crunched the numbers.
(Of course, deciding that his current average is probably inflated is quite different from trying to calculate what his true average might actually be, which would indeed bring sample size, regression, confidence intervals, and a host of other things into play. But I wasn't attempting to do that, which I think is where the confusion lies.)
