Tango: Four young aces destined to regress

Clayton Kershaw (2.79 ERA), Felix Hernandez (2.49), Jair Jurrjens (2.60) and Tommy Hanson (2.89) had their breakout seasons in 2009. Can we expect them to repeat those performances in 2010? If history is a guide, the answer is no.

To clarify, I'm defining a breakout season as one in which the pitcher allows runs at less than 80 percent of the league average. For example, John Danks broke out at 72 percent in 2008, and Matt Cain (78 percent) and Chad Billingsley (71 percent) broke out in 2007. What happened to these young pitchers the year after their breakouts? Billingsley kept pace at 73 percent, while Danks regressed to 83 percent and Cain to 84 percent.

From 1920 to 2008, 93 pitchers had their breakout season at age 23 or younger, the same as our 2009 quartet. Their average breakout level was 72.5 percent. Those 93 pitchers, in the season after their breakout, gave up runs at 88.5 percent of the league average.

Historically speaking, in the year after their breakout seasons, young pitchers will increase their ERA by about 20-25 percent on average. In fact, only 11 of the 93 improved from their breakout level, with Dwight Gooden leading the way. Gooden had his breakout season in his rookie year in 1984 and followed with one of the best seasons of all time the next year. However, another 19 pitchers gave up more runs than the league average in their postbreakout year, with the 2004 versions of Ollie Perez and Zack Greinke being the worst of the bunch.

Do young pitchers worsen? No. This is a result of a phenomenon called regression toward the mean. Basically, you are never as good as the results show, and you are never as bad as the results show. If a group of young pitchers gave up runs at far below the league average, they benefited from more good breaks than bad breaks. As a group, the breaks will even out for them the next season.

Don’t mess with regression

As the numbers show, it’s hard to repeat a breakout season.

Getting back to our 2009 quartet, expect their ERA as a group to rise by 20-25 percent, with a 10 percent chance of one of them improving his ERA and a 10 percent chance of one's ERA worsening by more than 34 percent.

Tom Tango writes for Inside The Book.