April 10, 2017
Do You Know This Man?
Chances are that you have not allocated much of your life’s attention to right-handed and occasionally-mustachioed starting pitcher John Burkett. And for good reason: Our temporary protagonist (and current professional bowler) threw his last pitch nearly 14 years ago, after a decade of plying his trade for five different teams, usually rounding out rotations, eating innings, waiting to be replaced.
It would be difficult to design a more invisible major-league starter. Burkett’s career ERA+ stands at 99, his record at 166-136, thanks to the quality of the players behind him. He didn’t walk many batters, sported a league-adjusted K%+ of 115, slightly above average. He made two All-Star teams in 14 years, never reached a World Series. Not once, but twice, he was acquired by a team in the offseason and then released before the end of spring.
His one great year (because even John Burketts have one great year), with the 1993 Giants, he won 22 games and earned six percent of the NL Cy Young vote; it proved to be one win shy of what the team needed to reach the postseason. A decade later, the Red Sox replaced him with Curt Schilling after a 2003 season that saw him post a 5.15 ERA and nine earned runs in nine post-season innings. He retired quietly.
This is perhaps not the kind of accolade one looks for in a letter of recommendation or a comment in a middle school report card. It’s the kind of encouragement you’d expect to hear of a parent with newborn twins, perhaps, or an above-average Minesweeper player.
Burkett's final surface-level totals:
Not a bad career, all told. You can think of him as Jeff Suppan with a few more strikeouts, or if it works for you, Steve Traschel with a little better control. Your basic third or fourth starter with good health, the guy who always seems to be pitching instead of the ace when you go to the ballpark.
Or you could think of him as a borderline Hall of Famer.
The story of John Burkett is essentially the opposite of the story of another young San Francisco Giant: Matt Cain. Before the talk of Matt Cain became “isn’t it depressing how bad Matt Cain is” and “I wonder if Matt Cain will even make the Opening Day roster," the big story was always about his peripherals. Cain was the first tough test of FIP, a fly-ball pitcher who consistently outperformed projection systems by keeping one-eighth of them from escaping the infield. Through 2012, over his first 1,500 innings, Cain posted an ERA of 3.27 and a FIP of 3.65. Since that cutoff, mediocrity has leveled those numbers out a bit.
Burkett, meanwhile, got the opposite treatment from his fielding-dependent pitching. In fact, he was unlucky in many ways. He chose the peak of the steroid era to hit his mid-30s, which exaggerated his decline. (No combination of technology and wokeness will ever truly succeed in making people accept the idea that a 5.62 ERA in 1999 would result in a mere 90 ERA+.) Pitchers were struggling all over the place, but the enigmatic Burkett, with no real standout virtue, looked particularly cooked.
Burkett also holds the third-worst ERA-FIP differential in modern baseball, behind perennial failed fantasy sleeper candidate Ricky Nolasco and perennially underrated Mariner and Red Erik Hanson. But both pitched barely more than half of Burkett’s innings, and as we saw with Cain, Nolasco’s decline phase will probably erode his lead. The next pitcher on the list with as many innings is no. 18, Javier Vazquez, who like Burkett was left off the Hall of Fame ballot entirely, although to perhaps somewhat more public chagrin.
This is fine, but it leaves us with two major and very different questions. The first: if FIP gets us to 44 wins above replacement level, how does DRA get him to 55? Forty-four wins is a surprising number for a guy whose rookie card is worth zero cents, but it’s not methodology-challenging. It’s John Lackey level, A.J. Burnett level. It’s enough that someone who wasn’t around him would probably say, “Huh, I guess I just never really got a chance to watch him,” and go about their day.
But 55 WARP pushes his legacy to the edge of credulity. Fifty-five wins puts Burkett, admittedly with 200 extra innings, a half-win ahead of Justin Verlander, 2.5 wins shy of Juan Marichal. (He’s also six back of poor Javier Vazquez.)
One factor in the difference is just inflation: BP simply rates its starters more highly, or its replacement level lower, than FanGraphs. The average pitcher on that 1,500-plus innings list saw a 10 percent higher WARP than fWAR. So in Burkett's case, that’s three of our 11 wins right there. The other eight are locked within Burkett’s DRA, which at 3.73 is a shade lower than his FIP. As it turns out, there was something we hadn’t seen before:
As it turns out, Roger Craig not only wasn’t lying, he was underselling things. Not only did Burkett have good control, he had excellent (and remarkably consistent) command. Burkett was the quintessential Roger Craig pitcher: taught the manager’s trademark split-finger fastball, Burkett lived by the ground ball and picked up called strikes on the bottom of the zone, avoiding the meatball-prone center and top of the zone.
Burkett doesn’t have the pinpoint reputation that a contemporary like Bob Tewksbury did, because he was willing to walk a batter now and then; better to throw a pitch on the edge, with a 50 percent called-strike probability, than an easy pitch to hit. This skill was worth nine wins, and went unappreciated in his time. Walks, in the 90s, were the only way we had to evaluate command.
But there’s also the second question: whether any of this matters. Deserved or not, Burkett did give up all of those runs. With Cain, the common sentiment was that in his case, Baseball Reference’s rWAR (which uses a simpler RA9-based algorithm to calculate value) did a better job of summarizing his value than the more complicated systems. The argument: every value-seeking system has to make some balance between recording what happened and what should have happened. rWAR leans far more heavily toward the former, while fWAR and WARP do the latter.
But after 1,500 innings, what was left to predict? People were tired of waiting for a regression that never seemed to arrive. At some point, Cain had to be as good as his numbers, because they were the numbers he earned. Should we take the opposite stance with Burkett, disregard the defenses and the ballparks and the sequencing and just say, “If he could have gotten it done, he would have?”
I don’t think there’s a correct answer. On the one hand, if you have a statistic that declares John Burkett a minor legend, it throws doubt on your numbers. At the same time, if you dismiss any system that doesn’t arrive at the conclusions already held, maybe the metric isn’t the problem.
We don’t want the Hall of Fame to be a WAR leaderboard. Burkett doesn’t belong in the Hall, even if he probably did belong on the ballot. But personally, I love the opportunity for reassessment that the WAR variants sometimes force on us. We may go through the numbers and decide that it’s a fluke or a flaw, that Burkett did something that the formula loved, but that didn’t actually prevent the other team from scoring runs. Or perhaps the opposite: that he was doing his part, handling the elements of winning that were under his control. But at least we have to stop and form an opinion about the player’s career.
For a guy like John Burkett, an anonymous fourth-starter star, it may be the first and last time he receives it.