Please check out the Hippeaux's weekly posts at SNY affiliate, It's About The Money.

Wednesday, March 31, 2010

Indicators & Illusions: Matt Cain

Baseball Tonight is just one of several places where the ongoing and somewhat irrelevant antagonism between sabermetrics and scouting is being played out for public consumption.  The fact is, unless you work for a stubborn backwater organization like the Royals or the Astros, you know that the battle has already been won, defiantly, by the sabermetricians.  Successful franchises pay handsomely for a wide variety of statistical metrics, some of their own design, and frequently combine them with economic analysis in an effort to directly link player performance and profit.

Sabermetrics are so in fashion that the heroes of their rise to power, guys like Bill James and Rob Neyer, now find themselves in the unusual position of frequently reminding their neophyte followers that there is actually a human element to the game (an x-factor) that cannot ever be fully accounted for, but must be recognized.  You cannot purely play the numbers.

For instance, Neyer recently discussed the "closer by committee" situation in Minnesota.  The role of relievers has long be a subject of debate for sabermetricians, many of whom believe, with good cause, that your best reliever should not necessarily be reserved for the ninth, and that the best approach to the late innings to to play match-ups and not identify roles like closer, set-up man, LOOGY, long-man, etc.

However, while excellent in theory, the "closer by committee" hasn't really caught on because the few teams that have tried it haven't had much success.  James recently switched his stance on the issue largely due to the "x-factor," recognizing that having a defined role allowed relievers to prepare physically and psychologically for the moment they took the mound, a routine which was critical to having consistent success.

Whether the connection is accurate or not, sabermetric analysic continues to be linked to fantasy baseball.  There is, apparently, a geek quotient that makes them natural bedfellows.  Thus, the Luddites like to accuse sabermetricians of following fake baseball.  Again, however, as men like Trace Wood and Jason Gray have long argued, the numbers cannot exist in a vacuum and they work best when they are used to ratify observations made during actual games.  Anybody who follows political debates know that statistics can be used to mislead.  Though the intent isn't as devious, the same is true in arguments about baseball.  Here is but one example of conflicting statistical indicators this spring.

Matt Cain had a fantasic, some might say "breakout" season in 2009, winning fourteen games and posting at 2.89 ERA at 24-years-old.  The basic stats, combined with his age, would make him appear to be a pitcher who should be considered among the ten or fifteen best in baseball, with the potential to develop into a perennial Cy Young candidate.

However, experts consistently rank Cain outside the top twenty among starting pitchers, behind guys like Ricky Nolasco and Scott Baker, players who are older and have never had a season rivaling what Cain did in '09.  They justify their rankings by pointing to a rising home-run rate (from 0.79/9 to 0.91), a declining strikeout rate (from 7.69/9 to 7.07) and a uncharacteristically low BABIP (.268 compared to .278 for his career).

Anybody who watched Cain consistently in 2009 would be very surprised by (in fact, would probably disbelieve) these suggestion that he was somehow becoming less dominant instead of more dominant as he moved into his mid-twenties.  Counting myself among them, I entirely agree.

Cain slightly altered his pitching philosophy in 2009.  Although still capable of getting strikeouts when the situation demanded it, Cain was far more content to pitch to contact.  The effort, I expect, was prompted largely by his desire to pitch deeper into games and factor into more decisions.  In '07 and '08 Cain had been profoundly unlucky, posting a 15-30 record over the course of those two seasons, despite the fact that his ERA (3.71), WHIP (1.31), and strikeout rate were all significantly better than the league averages.

Cain had an obvious and much-publicized run support problem, but also worked himself very hard, making it difficult for him to pitch deep into games, as is expected of a true Ace.  In 2008, Cain completed seven innings in less than half of his starts and pitched into the eighth on only five occasions.    Compare that to his teammate and Cy Young winner Tim Lincecum, who finished the seventh 60% of the time and pitched into the eighth 30% of the time, and you can understand how Lincecum won eleven more games, despite be saddled with the same poor offense.

In 2009, Cain made eighteen starts that lasted seven innings or longer, including a league-leading four complete games (as many complete games, by the way, as he had in previous four seasons combined).  He pitched into the eighth inning 25% of the time.  We can see, statistically, what made this possible.  Cain's walk rate went from 3.76/9 to 3.02.  The percentage of groundballs he induced went from 33.2% to 38.9%.

Cain's commitment to keeping the ball in the strike zone naturally led to more contact and thus more long balls and less strikeouts, but it also made for a lot more easy innings.  Conveniently for our analysis, Cain threw exactly the same number of innings in '08 and '09 (217 2/3), but in '09 he did it with one less start and more importantly with almost 250 fewer pitches, facing 47 fewer hitters (which explains his career best WHIP).  That's essentially two games less wear and tear on his young arm.

It wouldn't surprise me terribly if Cain returned to a more normal BABIP (~.280-.285) in 2010 and thus saw a slight regression in ERA.  However, it also wouldn't surprise me if he intensified his commitment to the philosophy that served him so well in '09 and saw a further reduction in walks and an increase in innings and wins, cementing him as one of the National League's elite pitchers.

No comments: