Before the 2010 season, I got familiar with the idea of regression to the mean. This concept illustrates the phenomenon of an outlying statistic in one season (such as an extraordinary ERA) coming back down to earth in the next, simply by chance.
My study examined how the best pitchers (in terms of ERA) fared in the following season, dating back to 2000. To measure this, I (somewhat arbitrarily) choose 2.50 as the cutoff for ERA.
Here’s what I found:
On average, each pitcher’s ERA increased by 1.01 runs per nine innings in the season following their sub-2.50 campaign. Also, each pitcher averaged nearly 40 innings less the following season.
I revisited this last offseason, and found similar results.
Including the totals from 2010, there have been 20 instances of a sub-2.50 ERA season since 2000. Eleven of those performances came between 2000 and 2005. During the 2006, 2007 and 2008 seasons, there were none. 2009 produced five. The year of the pitcher (2010) yielded four.
Of the 20 sub-2.50 ERA seasons since 2000, only five times has a pitcher lowered his ERA in the following season. Josh Johnson is one of them, though he only pitched 60.1 innings this season (1.64 ERA) after posting a 2.30 ERA in 183.2 innings in 2010.
The other four?
Randy Johnson (01-02), Pedro Martinez (02-03), Felix Hernandez (09-10) and Roy Halladay (10-11).
Here’s the full list:
Six more pitchers posted a sub-2.50 ERA in 2011: Clayton Kershaw (2.28), Johnny Cueto (2.31), Roy Halladay (2.35), Cliff Lee (2.40), Justin Verlander (2.40) and Jered Weaver (2.41).
Based on recent history, we should expect an average increase in ERA of about one among these starters in 2012.
Last year I called for a regression from Felix Hernandez. I got slammed for it, but it happened (2.27 ERA to 3.47). I’m not suggesting that any of these pitchers will fall flat next season. Simply put, a regression to the mean is likely.
Of the six pitchers who posted sub-2.50 ERAs in 2011, Cueto, Weaver and Verlander are the ones I expect to fall the farthest in 2012.
Cueto, for example, improved on his previous-best season ERA (3.64 in 2010) to post a 2.31 ERA/3.45 FIP/3.90 xFIP in 2011. His .249 BABIP (while supported by 53.7 ground-ball rate) is simply unsustainable.
Although I’ve been tooting Weaver’s horn for a while (here and here), he is likely to be overvalued on draft day next season at a price you won’t want to pay. Despite his stellar 2.41 ERA in 2011, Weaver’s FIP (3.20), xFIP (3.80), BABIP (.250) and strand rate (82.6) suggest a regression in 2012.
Verlander’s 24 wins in 2011 will make him seem more attractive than what he actually is next year. Consider his 2011 FIP (2.99), xFIP (3.12), BABIP (.236) and strand rate (80.3). Don’t get me wrong, he had a fantastic season and is definitely the A.L. Cy-Young award winner. But as Dave Cameron points out, if you ignore the win total, Verlander’s 2011 season wasn’t all that historic.
I’m less worried about Halladay and Lee because they’ve proven to be consistently awesome, and luck doesn’t appear to have played a role in their 2011 success. It’s possible that Kershaw has reached his statistical peak at age 23, but I doubt it. A huge improvement in walk rate (3.57 in 2010, 2.08 in 2011) catapulted him into the elite ranks. If he can continue to minimize his walks, a fourth-straight sub-3.00 ERA season is likely.
Image courtesy of: Mark Sobba
Tags: Adam Wainwright, Andy Pettitte, Chris Carpenter, Clay Buchholz, Clayton Kershaw, Cliff Lee, Felix Hernandez, Jake Peavy, Jason Schmidt, Jered Weaver, Johnny Cueto, Josh Johnson, Justin Verlander, Kevin Brown, Mark Prior, Pedro Martinez, Randy Johnson, Roger Clemens, Roy Halladay, Tim Lincecum, Zack Greinke