Finding the next breakout pitcher

Over the past few fantasy seasons, we’ve seen a number of pitchers increase their fantasy value. Some of these improvements have been flukes (hello, Oliver Perez), and some of them have showed that their new talent level was legitimate and warranted a pretty high draft slot. If I were to name some of those folks offhand, I’d come up with guys like C.C. Sabathia, Aaron Harang, Dan Haren, and James Shields. These are pitchers who had what I’d call “breakout” seasons: seasons in which they improved dramatically and sustainably.

It would be an enormous advantage to have some insight during a draft into who might be a breakout candidate, as before their coming out party they’d likely still be available by the 150th or 200th pick. And going forward, they won’t be around that late for many years.

Defining “breakout”

If we’re going to objectively determine who might be a breakout candidate in 2009, we first need to objectively determine what a breakout season is. Going back to 2002, I downloaded statistics on all pitcher-seasons in which a pitcher amassed over 100 innings. It seemed like a good number in order to minimize small sample size issues, include only starters, and allow for some variability in pitcher health over the course of the analysis. Choosing 200 innings would give me seasonal lines in which I was more certain of true pitcher talent, but it would eliminate some pitchers who managed as many as 30+ starts—by all accounts, a total that any GM should be happy to get out of a starter.

How can one best measure whether a pitcher’s true talent has changed from season to season? There are a number of statistics one could choose from, but I decided on FIP. It’s important to note that just because a pitcher had a significant change in FIP from year to year, does not mean his fantasy value changed. A bad FIP one year with a few lucky wins is certainly as good or better than a solid FIP the next year with some tough losses and no-decisions. On the whole, though, pitchers who show a repeatable improvement in FIP will be good bets to have significantly higher fantasy value as well.

Knowing that I wanted to look at FIP, I decided to select all pitchers who met the following criteria:

1. pitched 100+ innings in three consecutive years
2. showed a FIP improvement from the first year to the second of at least .30
3. showed a FIP improvement from the first year to the third of at least .30
4. maintained a FIP of 3.99 or better during the second and third year.

In plain words, the first qualification ensures the pitchers are starters with at least a minimum measure of durability. The second qualification ensures that pitchers improved by a substantial amount, and the third qualification ensures it wasn’t a fluke and that they maintained this improvement the year after. The final qualification ensures that the improvement put the pitchers among the elite in the big leagues, and can be a big asset to a fantasy team going forward. A pitcher who meets all four qualifications is one that I’d define as having had a “breakout” season during the second of those three seasons. They showed lasting, noticeable improvement that put them at the highest level among their peers.

Examining the breakouts

Applying those criteria to the aforementioned sample (seasons since and including 2002), I get the following list of players:

Year	Player	Previous Year FIP	Breakout FIP	Next Year FIP
2003	Brad Penny	4.52	3.92	3.59
2003	Carl Pavano	4.98	3.76	3.54
2004	Ben Sheets	3.98	2.65	3.31
2004	Doug Davis	5.13	3.57	3.98
2004	Jake Peavy	4.99	3.14	2.89
2004	Johan Santana	3.47	2.92	2.80
2004	Roger Clemens	3.60	3.11	2.87
2004	Roy Oswalt	3.67	3.18	3.16
2005	Aaron Harang	4.68	3.67	3.69
2005	Brandon Webb	4.41	3.51	3.18
2005	Chris Carpenter	3.85	2.90	3.44
2005	Erik Bedard	4.26	3.49	3.59
2005	Jeremy Bonderman	4.30	3.90	3.29
2005	John Lackey	4.08	3.10	3.35
2006	CC Sabathia	3.69	3.30	3.14
2006	Chien-Ming Wang	4.26	3.91	3.79
2006	Scott Kazmir	3.76	3.36	3.45
2007	Dan Haren	4.12	3.70	3.01
2007	James Shields	4.39	3.86	3.82
2007	Josh Beckett	5.12	3.08	3.24
2007	Tim Hudson	4.55	3.46	3.83

Not bad, right? I actually think the criteria I established ended up working pretty well. The four examples that came to my mind are there, and with the exception of Pavano and maybe Doug Davis, I think this is a very good list of pitchers who have turned themselves into stars over the last few years. C.C. Sabathia would actually appear twice, showing us that he’s perhaps not once, but twice establish new levels of excellence. However since his first breakout year looks like a result of one down year in 2004, I decided to only include data from his second breakout, in which he reached a substantially lower FIP than he’d ever had even before that down year.

So in examining this list, I am reminded that there are two primary ways to be an elite starter in the majors: strike a bunch of guys out and keep the walks down, or keep the ball on the ground and in the park. To account for this, I’ll divide the group into two subcategories: those who get more than 50 percent ground balls, and those who get less. This will help me examine how these players managed to get better.

Our ground ball group of Webb, Wang, Carpenter, and Hudson show an average of a 5 percent smaller strikeout rate in their breakout year, and merely one additional percentage point of batted balls being ground balls. What is the source of their improvement then? It appears to be their walk rate, which improved by 26 percent in their breakout year.

As for the classic strikeout pitchers, their breakout seasons show a 14 percent increase in strikeout rate, a 16 percent decrease in walk rate, and no meaningful change in rate of ground balls.

Conclusion

Over the past six years, there have been two types of breakout pitchers: high ground ball pitchers, and high strikeout/low walk pitchers. Groundball pitchers, though a small group, have reached new levels of performance by reducing their walk rate by over 25 percent. The rest of the breakout pitchers showed significant improvements in both their strikeout rate (14 percent increase) and their walk rate (16 percent decrease). In the second and final installment of this series, I’ll take a look at the collective pitching lines of these pitchers in the year prior to their breakout. I’ll then match that up to pitching lines from 2008, and try to come up with a list of potential fantasy sleeper picks and breakout candidates for 2009.