If you have ever read Baseball Prospectus' Baseball Between the Numbers , or a similarly analytical baseball text, the argument against standard pitching statistics (wins and losses, ERA, and saves) is not new to you. By now, most educated baseball fans understand that all three of those statistics reflect myriad factors far beyond the pitcher's control (and, in fact, relatively few that fall within it).
But while some fans have eagerly accepted the common-sense preaching of sabermetric prophets, others continue to live in hazes of either delusion or obstinacy. This is especially true when it comes to relief pitchers. Though ERA does a particularly woeful job of capturing contributions made out of the bullpen (when so much of the game depends upon context, preferred kinds of outs, and leverage), it continues to be the hallmark of a relief pitcher's worth.
Other flawed metrics used to measure the performance of bullpen men include holds, saves, and record, all of which better reflect their teams' performance than their own.
Unfortunately for the 2009 (and, it appears inevitable, 2010) Chicago Cubs, one fan who fights statistical analysis tooth-and-nail every day sits in the general manager's chair. According to Cubs beat writers Carrie Muskat and Paul Sullivan, Jim Hendry has already begun contract negotiations with southpaw reliever John Grabow, whom he ill-advisedly acquired in July as a (rather foolhardy) attempt to keep Chicago in the National League Central race, from which they were already quickly fading.
How Grabow's ineptitude goes so unnoticed is a question better answered by those who fail to notice it. For my part, I will simply undertake the task of demonstrating his more dire failings.
First, he lacks the three fundamental and repeatable skills of a good pitcher: control, stuff, and the ability to induce ground balls. No pitcher statistics correlate more highly in the same player from year to year than (in order of best correlation coefficient): ground ball/fly ball ratio, strikeout rate, and walk rate.
Grabow looks bad in every category. His walk rate in 2009, 4.98 per nine innings, ranked 19th-highest among Major League relievers. He made marginal forward progress under the (dubious) tutelage of Larry Rothschild, walking "only" 12 in 25 innings as a Cub, after issuing an obscene 28 free passes in 47.1 frames in Pittsburgh.
Still, control is a clear problem: walks go a long way toward explaining his inability to ever post a single-season WHIP under 1.28, and his failure to reach league average in that statistic in 2009 (he surrendered a 1.41 WHIP, compared to the league average for relievers of a shade under 1.40).
Nor can he miss bats with sufficient aplomb to remain consistent. Grabow struck out fewer batters per nine innings in 2009 than he ever had before, at a rather bland 7.09 K/9. Thanks to that troubling lack of punch-out skill, he ranked even worse in strikeout-to-walk ratio (17th-worst among relievers) than in walk rate alone.
At least all of those lacking skills used to have a strong ground ball percentage as a counterbalance. But after beginning his career with at least five seasons of a 1.44:1 or higher ground ball to fly ball ratio, Grabow has been utterly neutral the past two seasons: 0.98 and 1.05, respectively. That may have been a tolerable tendency in relatively neutral PNC Park, but in hitter-friendly Wrigley Field (12 percent more homers and overall runs than Pittsburgh in 2009), it's a fatal flaw.
Defense also masks some of Grabow's problems. He spent years inducing grounder after grounder to the likes of Freddy Sanchez and Jack Wilson, premier defensive middle infielders. Then, over the last two seasons, he simply watched fly balls fall harmlessly into the gloves of Nate McLouth, Nyjer Morgan, and Andrew McCutchen.
Barring something unforeseen (like a trade for J.J. Hardy or Carl Crawford), or something stupid (like signing Mike Cameron or Coco Crisp), Grabow cannot expect to get such help from the Cubs (who feature miserable defenders in left field and at third base, and probably at second, too) in 2010.
Grabow has another point which works strongly in his favor in the myopic eyes of Hendry. He has appeared in 149 games over the past two seasons, the seventh-highest total in the Majors, and sufficient evidence for Hendry to declare him a rubber arm, with whom manager Lou Piniella can expect extra flexibility.
It is all too easy, however, to confuse prolificacy with durability. All too often, managers deploy pitchers based on the manager's perceptions about pitcher durability and usage in general, without regard to an individual's true abilities. So does Grabow actually possess this knack for throwing often, and doing it well?
The numbers say no. In the last two years, Grabow has appeared 39 times on zero days of rest. He logged 36.1 innings in those appearances. He struck out 31, a solid number somewhat higher than his actual rate of whiffs during that span. But he also walked 25, and allowed 27 hits, making his composite WHIP for those games 1.43.
Finally, there is the question, ever-relevant to the examination of an individual pitcher, of how luck influenced his performance. Grabow clearly has been the beneficiary of good luck over the past two seasons, the first two in which he has broken out as a so-called "fireman."
In 2008, he registered a .251 batting average on balls in play, a statistic over which pitchers exhibit relatively little control, and one that places him well below the league average of .304. In 2009, that number rose slightly, to .279, but still registered as well below league average, a red flag that Grabow got rather lucky.
If that weren't enough, Grabow also notched a home run per fly ball rate of just 5.7 percent in 2009, about half the league average, and a highly suspicious figure for a man who has seen his fly-ball percentage rise over the past two years, and who had never before managed a rate of less than 10 percent.
Even stranger, the percentage of his fly balls that stayed on the infield (a fairly repeatable skill, if one trusts the measure of year-to-year correlation as an indicator of consistency) fell by nearly a third from 2008 to '09, from 11.6 percent to a flat eight percent.
Although his demands (three years, somewhere between eight and nine million dollars) make him infuriatingly inefficient, the Cubs will likely re-sign Grabow, and that is okay. I do not write with the intent of picking on the lefty, who will be 31 years old next season. Rather, this is meant as an allegorical piece, from which I hope readers will draw the following broader conclusions:
1. Team contribution and overall quality simply cannot be inferred from top-line, generic, old-school baseball statistics, especially for relief pitchers. Scouting reports, more advanced metrics, and component rates provide better indicators.
2. Teams should, but generally don't, consider the way their defense and ballpark affect their pitchers, and vice versa. If your two best hitters are poor defensive middle infielders (Florida Marlins), aggressively pursue fly ball pitchers. If you play on a mountaintop and have one baseball's best double-play combinations, try to keep the ball on the ground.
If your entire lineup can hit (a little) but not field, you are the Washington Nationals, and your best shot is probably to spend top draft picks on flame-throwing wunderkinds every year until you're halfway decent.
3. If a guy can't throw the ball over the plate, and if he can't miss bats when he does, he is not a multi-million-dollar pitcher, regardless of his other indicators.
4. In many cases, managers exercise absolute influence over relievers. Clubs should pursue hurlers who can succeed in the roles to which the Major League manager is likely to assign them.
5. Platoon advantages do not stop where handedness leaves off. Pitchers who encourage grounders are better against ground ball hitters, and fly-ball pitchers do better versus fly-ball hitters.
The case study of Grabow informs some of these ideas; others bear upon other projected members of the Cubs' pen in 2010. But for now, I will be content with just one objection. Jim Hendry, please, in the name of all that is good and rationally sound, let John Grabow go.