Today at Baseball Prospectus, I've got a look at an oft-misunderstood metric, quality starts. Invented 25 years ago by Philadelphia Inquirer writer John Lowe (now at the Detroit Free Press), a quality start is one in which a pitcher tosses at least six innings and allows no more than three earned runs. Many a mainstream writer — such as Murray Chass, nemesis of postmillennial knowledge — has complained that the stat potentially rewards a start in which a pitcher's ERA for the game is 4.50, a level to which the major league ERA has risen just eight times since 1950. Eager to find just how often such cases occur, and how the value of a quality start has changed over time, I dug in. The take-home points:
• Among major league starter, the frequency of a quality start is heavily tied to overall scoring levels. It has ranged between 45.8 percent in 2000 (when MLB scoring levels reached 5.0 runs per game for the first time since 1936) and 62.6 percent in 1968 (when they fell to 3.4 runs per game, the lowest level since 1908). Historically, it hovers around 50 percent, but last year (when scoring dipped to its lowest level since 1992) was just the second time it was above 50 since 1993.
• Those 4.50 Cases historically constitute just 5.9 percent of quality starts and 3.0 percent of all starts. Last year they made up 8.5 percent of quality starts (the first time below 10 percent since 2005) and 4.5 percent of all starts. They're a small slice of the pie.
• Starting pitcher ERAs in quality starts are around 2.00 and have never been higher than 2.13; last year they were 1.99. By comparison, starting pitcher ERAs in non-quality starts are generally around 7.50 and have never been lower than 6.99; last year they were 7.58. These are similar splits to what we see among all pitchers in wins (1.96 last year) and losses (7.69). Collectively, pitchers making quality starts are pitching to an elite level and those who aren't are getting the snot knocked out of them.
• Teams receiving quality starts win about two-thirds of the time, ranging over time from a low of about 64 percent to a high of just above 70 percent and not surprisingly intertwined with scoring levels, though not as tightly as the frequency of quality starts are.
The unevenness of last year's Yankee rotation is quite apparent when one looks at their rates of quality starts:
Pitcher GS QS QS%
CC Sabathia 34 26 76%
Andy Pettitte 21 14 67%
MLB Average 53%
Phil Hughes 29 15 52%
A.J. Burnett 33 14 42%
Javier Vazquez 26 10 38%
Dustin Moseley 9 3 33%
Ivan Nova 7 1 14%
Sergio Mitre 3 0 0%
Team Total 162 83 51%
The big man ranked eighth in the majors and fourth in the AL in quality start rate. The Mariners' Felix Hernandez (88 percent) and the Marlins' Josh Johnson (82 percent) topped their respective circuits, with David Price (81 percent) and Jered Weaver (79 percent) the other AL pitchers higher than Sabathia. Meanwhile, Vazquez had the eighth-lowest percentage among the 126 pitchers with at least 20 starts, and Burnett was 18th-lowest.
One of the interesting facets of the Yankees' ongoing rotation race is that Freddy Garcia actually fared quite well in this department. With 18 quality starts out of 28, his 64 percent rate ranked 39th, matching Zack Greinke, Cole Hamels, John Lackey and Cliff Lee, and edging Dan Haren, Jon Lester, Colby Lewis, and well ahead of three-fifths of the Yankees' regular starters including Hughes. Garcia's swollen ERA (4.64 overall) resulted from a wider-than-normal split in his starts; he posted a 2.73 ERA in his 18 quality starts, and a 10.57 mark in his non-quality starts. In the four worst of them, he combined to allow 25 runs in nine innings. As bad as this sounds, this is actually the way to go, minimizing the number of lost causes and staying competitive in the rest. On a team level, those that tend to concentrate their beatings are the same ones who outperform their Pythagorean records by a wide margin.
That performance wasn't exactly a fluke; Garcia owns a 58 percent career mark, the same as Andy Pettitte despite a career ERA 0.3 runs higher (4.13 to Pettitte's 3.83). Barlolo Colon is at 56 percent for his career despite an ERA 0.03 lower than Garcia's, and he hasn't been anywhere close to 50 percent since hurting his shoulder in 2005. It's all just one more reason to think that Garcia's got a spot waiting for him in the rotation so long as he stays health.