As we wind down this year's series of Pitching by the Numbers, let's focus on the pitchers who performed best in the stats that tend to swing dramatically in random directions. This means that the pitchers who perform far better than career/league norms can reasonably be called lucky.
These stats are HR/FB rate, where the league average tends to be about 10 percent of fly balls resulting in homers; Left On Base%, where the league average tends to be around 70 percent in stranding baserunners; and BABIP, where the league average is about .300 on generating hits on balls put in play (excluding homers and strikeouts). The BABIP average this year is just .294, I know. But more on that later.
I'm not going to put career number in here for everyone. For one thing, in a short career such as Jeremy Hellickson(notes) it's not especially relevant. In those cases, just expect a regression to league-average rates until you see otherwise. And secondly, you really have to eyeball the yearly numbers because the career numbers always include this year's arguably lucky rates.
I understand that the people who are fans/owners of the guys below are going to come up with reasons why their guy wasn't lucky and why this level of success in these categories will not regress. That's human nature. There's a small chance they are right. But it's not really something that you can bet on. These are the type of low probability things that you want opponents to wager on in the 2012 fantasy baseball drafts.
Generally with established players like Justin Verlander(notes), you should rank them where you had them ranked last year. Why is Verlander significantly different based on the six-month season? The burden of proof is on you. With these guys, instead of paying for last year's stats, pay last year's price. You won't get them, I know. But the point is that there's going to be very little room for profit.
Cain's rate for his career is 6.6 percent, which very good. But it was higher prior to this year, which means that instead of sitting on seven homers allowed it should be about 14. That's about 10 more earned runs right there. So his ERA right now should be 3.38 and not 2.86 and I bet that 3.38 was right around his consensus projection. Usually, the plus-minus above or below consensus is 100 percent good-bad luck. I understand the Cain owners (I was one in the Yahoo! Friends & Family League) want to think they were geniuses for calling Cain, but come on. Show you're a genius by not believing that luck will repeat in 2012.
Masterson was below average and is HR/FB for his career and is 4.3 percent this year. So expect his rate to at least double in '11. Instead of six more homers, he should have yielded somewhere in the 12-15 range. At about 1.5 runs per homer, you can do the math with his adjusted ERA.
And, of course, Weaver has been lucky, too. More lucky than Verlander because Verlander has a higher baseline as a better/more dominant pitcher. His career HR/FB is 7.4 percent including this year.
Beckett's career rate is 72.3 percent, Weaver's 76.6 percent and Romero's 74 percent. That includes this year's rates. I understand that you can argue that good pitchers strand more baserunners naturally because they are better than average. I buy that, but not beyond career rates for pitchers with a significant sample size. Guys can and do suddenly get better. But it's safer to never bet on that because it's rare and there's no iron-clad way to tell when they do. We have theories, that's all.
Cueto is now .280 BABIP for his career, Verlander .287. Hellickson is at .234 with not enough career data, but come on. League average this year is lower than normal at .294 and that's a fluke too. The instinct with the data for the year is to say – "That’s too big a sample to be a fluke." But if the expectation is .300, you would expect to have a year where it's .294 even with all this combined data. It's gone from .299 to .297 to now .294, so that's a trend, right? No, it just appears to be a trend. Just like flipping coins in some pattern only appears to be a trend. Think of it this way, if you had a 1,000 years of BABIP data where the average was .300, would you think twice if there were multiple .294s or a .306s in there? Of course not. This is just one of those years.
Michael Salfino writes and edits the SNYWhyGuys blog that projects player and team performance for New Yorkers. He's also a quantative sports analyst whose writing regularly appears in the Wall Street Journal.