October 21, 2005
Evaluating the Departed Yankee Pitching Coach
Last week, the Yankees' season ended atypically early and in disappointment, with the team's elimination at the hands of the Angels in a five-game Division Series. As is always the case when a pinstriped season ends short of a celebratory dogpile, speculation about whose head might roll predominates the postmortem analysis, overshadowing even the usual finger-pointing and second-guessing. On Tuesday, manager Joe Torre ended a tense week of silence by announcing that he would return for another season at the helm (he has two years and $13 million remaining on his contract), and GM Brian Cashman is weighing his options as his contract approaches its expiration at the end of this month. But one familiar face has already departed. Pitching coach Mel Stottlemyre resigned in the wake of the team's defeat, with some parting shots aimed at owner George Steinbrenner.
Stottlemyre nearly didn't make it this far. In the season's first month, as the Yankees struggled along with a sub-.500 record, the unimpressive performance of the pitching staff renewed calls for his dismissal that had been heard at the end of last season. At the base of the complaint was an undeniable decline in the quality of the pitching staff's performance, one that appeared to have something to do with Stottlemyre's directive for the team's pitchers to rely less on their ability to strike hitters out in favor of putting the ball in play and subjecting it to the whims of a subpar Yankee defense.
ERA (rk) K/9 (rk) PIP (rk) DE (rk) 1996 4.65 (5) 7.12 (2) .677 (11) .683 (11) 1997 3.84 (1) 7.15 (3) .688 (10) .685 (8) 1998 3.82 (1) 6.67 (5) .697 (9) .713 (1) 1999 4.13 (2) 6.95 (3) .680 (12) .699 (3) 2000 4.76 (6) 6.57 (3) .690 (10) .693 (4) 2001 4.02 (3) 7.85 (1) .672 (12) .684 (10) 2002 3.87 (4) 7.04 (2) .706 (6) .690 (9) 2003 4.02 (3) 6.89 (2) .714 (6) .682 (13) 2004 4.69 (6) 6.60 (6) .707 (2) .688 (7) 2005 4.52 (9) 6.20 (6) .714 (7) .689 (10)ERA and K/9 should be familiar enough. DE is Defensive Efficiency, the percentage of balls in play a team converts into outs. The numbers in parentheses are the relative ranks within the AL. Note that for all of the team's success, the defense has rarely even finished in the upper half of the league in doing so, placing the staff's ability to miss bats at a premium. PIP is the percentage of balls opposing hitters put into play, by the formulas:
BIP = BFP - HR- SO - BB - HBP - SF - SH PIP = BIP/BFPIn Stottlemyre's first eight years with the Yanks, the team finished in the top three in the AL in strikeout rate seven times, and had a top-three ERA five of those times. In the year they finished outside of the top three in strikeout rate, their league lead in Defensive Efficiency helped them win the ERA title. In the three years they finished outside the top three in ERA, their DEs ranked in the lower half twice. In the past two years, the team has posted unimpressive middle-of-the-pack finishes in strikeout rate and DE, with ERA rankings that are subpar compared to what preceded them.
More perceptible than that decline, of course, was the underperformance of a pricey group of starting pitchers and the lack of continuity with the previous rotations. Recall that following the 2003 season, Roger Clemens, Andy Pettitte and David Wells left via free agency, while Jeff Weaver was traded--an unprecedented amount of turnover for such a successful team. The end of 2004 saw more upheaval, with Jon Lieber, Orlando Hernandez and Esteban Loaiza (acquired for the enigmatically disappointing Jose Contreras) departing via free agency and Javier Vazquez sent away in a trade for Randy Johnson. Signed as free agents were Carl Pavano (four years, $39.95 million) and Jaret Wright (three years, $21 million), two injury-prone pitchers who were being paid on the basis of strong 2004 seasons but who possessed rather sketchy track records.
Here is what the Yanks got from their starters, along with how that performance compared to the market value, using a formula recently unveiled by Nate Silver:
GS IP ERA VORP SNLVAR WARP Sal Val Dif Johnson 34 225.2 3.79 44.1 5.7 6.8 16.00 12.56 -3.44 Mussina 30 179.2 4.41 23.3 3.4 4.7 19.00 6.58 -12.42 Pavano 17 100.0 4.77 -1.3 1.0 1.0 9.00 0.61 -8.39 Brown 13 73.3 6.50 -9.5 -0.5 0.4 15.00 0.19 -14.81 Wright 13 63.2 6.08 -9.8 0.1 0.2 5.67 0.09 -5.58 Wang 17 116.1 4.02 17.3 2.1 3.6 0.32 4.20 3.89 Chacon 12 79.0 2.85 25.1 3.1 3.9 0.94* 4.80 3.86 Small 9 76.0 3.20 22.1 1.7 3.6 0.32 4.20 3.89 Leiter 10 62.1 5.49 -1.7 0.7 0.8 0.15* 0.46 0.31 *pro-rated shares for players acquired via tradeAll salaries are in millions of dollars and include prorated signing bonuses, but not incentive bonuses, and no allowance has been made for deferred money. Shawn Chacon's and Al Leiter's salaries are pro-rated shares for midseason acquisitions; the rest of Chacon's $2.35 million contract and nearly the entirety of Leiter's deal were picked up by their previous teams (salary info provided via Cot's Baseball Contracts, Hardball Dollars, and the USA Today Baseball Salary Database, three very useful sources).
Beyond Johnson, who at least pitched his share of innings at a level most mere mortals would be jealous of, the rest of the Yankee staff had trouble just making it to work, let alone pitching to the high standards expected of an expensive hurler on a contender. The quartet of Mike Mussina, Kevin Brown, Pavano and Wright made just 73 out of the 120 starts that could have been expected of the #2 through #5 pitchers, about 61 percent. All four missed significant time due to injury, and were a combined 2.7 runs above replacement level according to VORP, all for the low, low price of $46.7 million--more than the entire payroll of the Cleveland Indians, who narrowly missed the postseason, not to mention four other teams. Mussina was the game's highest-paid pitcher, and ranked fifth in overall compensation behind Alex Rodriguez, Barry Bonds, Manny Ramirez, and Derek Jeter. As a unit, the starting five (including the biggest Unit) performed at a level that, per Silver's formula, was $44.6 million short of the expected performance at that market value.
Some of the blame for this catastrophe can be pointed at management (presumably not just Cashman but the multi-headed hydra that resides in Tampa, whispering sweet nothings in Steinbrenner's ear while undermining the Cashman/Torre/Stottlemyre axis) for failing to properly assess injury risks and provide adequate backup. The 2005 season was the first in recent memory that the Yanks didn't have a spare starter (Orlando Hernandez, Jose Contreras, Sterling Hitchcock, etc.) waiting in the wings amid some controversy, expensive but effective allegiance to the dictum, "You can't have too much pitching." In the end the Yanks were able to call on rookie Chien-Ming Wang and a pair of pitchers, Chacon and Aaron Small, whose track records foreshadowed little of the high-caliber contributions the team got from them. That trio saved Stottlemyre's bacon, and with it, the Yanks' season.
Nonetheless, Pavano, Wright, Brown and Johnson all joined a list of dubious company: high-profile pitchers who fared significantly worse than their career norms upon coming to New York during Stottlemyre's tenure. Here is an admittedly selective list of pitchers who bombed in the Bronx from 1996-2005, one guaranteed to produce groans among Yankee fans while evoking memories of Ed Whitson:
Years IP ERA Car. ERA* Kenny Rogers 96-97 324 5.11 4.13 Denny Neagle 00 91 5.81 4.16 S. Hitchcock 01-03 140 5.84 4.68 Jeff Weaver 02-03 237 5.35 4.20 Esteban Loaiza 04 42 8.50 4.60 Javy Vazquez 04 198 4.91 4.12 Kevin Brown 04-05 205 4.95 3.16 Randy Johnson 05 226 3.79 3.11 Carl Pavano 05 100 4.77 4.27 Jaret Wright 05 64 6.08 5.17 * besides listed seasonsA variation of that list was used in a piece that I wrote encouraging the Yanks to find themselves a new pitching coach back in April, one that made its way into just about every Yankee-themed discussion site on the Internet. In doing so, I also pointed out the fact that the Yankees had failed almost completely to develop any pitchers of consequence from within before trading them away. Both Pettitte and Mariano Rivera arrived in the majors in 1995, the year before the Torre/Stottlemyre regime did, and while they went on to flourish under Stottlemyre, the only other Yankee product to do so has been Ramiro Mendoza. Every other pitcher of lasting note during Stottlemyre's tenure arrived in the Bronx a finished product, for better or for worse. Many of them did fare for the worse, and when they did, Stottlemyre appeared powerless to pull them together. Meanwhile, traded Yankee prospects such as Eric Milton, Jake Westbrook, Zach Day, Ted Lilly, Brandon Claussen and Damaso Marte have had a handful of seasons in which they were both effective and inexpensive, though by 2005, many of them had shown that their success was fleeting.
Again, the Yankees' management is somewhat culpable for this shortcoming, particularly the player development side. But some level of accountability must reside with Stottlemyre and Torre, despite the obvious positives of four World Championships, six pennants, nine division titles and 10 playoff appearances on their watch. Unquestionably, many pitchers didn't perform as well for the Yankees as they did elsewhere, and some of them clashed with the coach. But leaving aside anecdotes and personal issues, how do we measure that? Gauging the success or failure of a pitching coach is a difficult thing to do directly. Aside from accounting for the staff's collective performance relative to the league as above, perhaps the best way is to compare the performance of the pitchers during the interval at hand to their performance elsewhere.
To do that, I culled every Yankee pitcher who threw more than 50 innings in a single season for the Yanks during Stottlemyre's tenure, creating a pool of 39 pitchers who through 2005 have thrown nearly 59,000 innings and racked up 3,912 wins. To adjust for the changing venues and scoring environments in a set of pitchers whose careers go back as far as 1984, I used ERA+, the ratio of the park-adjusted league average ERA to the pitcher's ERA (indexed so that 100 is average and that 110 represents 10 percent above average), along with actual ERA and the standard peripherals of strikeout rate, strikeout-to-walk ratio, homer rate and batting average on balls in play. The initial data would appear to support the premise that pitchers didn't perform as well under Stottlemyre as they did elsewhere:
Group Age IP ERA ERA+ BABIP K/9 K/BB HR/9 non-NYY 29.0 46169 3.75 115 .289 7.24 2.37 0.85 NYY 96-05 31.5 12703 4.02 111 .293 6.99 2.42 0.94In terms of ERA+, the pitchers were four percent worse relative to league average while wearing pinstripes. Their strikeout rates were down, while their homer rates were up (though some of that is a park effect). They also got less support from their defense on balls in play. The performance gap becomes magnified when we split the non-Yankee performance into pre- and post-NYY groups:
Group Age IP ERA ERA+ BABIP K/9 K/BB HR/9 Pre NYY 28.0 37879 3.67 118 .289 7.38 2.41 0.82 NYY 96-05 31.5 12703 4.02 111 .293 6.99 2.42 0.94 Post NYY 33.4 8290 4.12 104 .293 6.60 2.23 1.03Aside from the fact that the progression of the decline appears to follow a symmetric path of seven percent per interval relative to the league average, what's emerging here is a story of a set of very good pitchers who come to New York and perform well, but not quite up to their career standards. Once they leave, however, they aren't worth a heck of a lot more than league average, suggesting at the very least that the Yankees managed to extract their value out of those pitchers. One thing not accounted for in this study is the steady rise of the league-average strikeout rate over the 22 seasons covering the pitchers in this study. In 2005, the AL strikeout rate was 6.16 per nine innings, down from a high of 6.74 in 2004; in 1984, when the careers of Clemens and Jimmy Key began, it was 5.14.
Note also within the data that there's an inherent selection bias at work:
Group AGE IP ERA ERA+ Pre NYY 28.0 37613 3.67 118 NYY 96-05 33.3 7824 4.14 107 Post NYY 35.0 6774 4.05 109So the 30 remaining pitchers actually fared worse in the Bronx than they did either before or after arriving. That might be seen as an indictment of Stottlemyre, particularly of his ability to work with veterans. But it also suggests that when given pitchers relatively inexperienced at the big-league level and perhaps more inclined to listen to Stottlemyre's advice, the man had some very real success:
Group AGE IP ERA ERA+ Pre NYY 23.5 266 4.81 95 NYY 96-05 28.5 4879 3.84 119 Post NYY 31.2 1516 4.46 84Of course, time will tell with this group. Relative to the other groups in this study, small-sample caveats apply, particularly when one considers that Pettitte and Rivera account for 48 percent of the Yankee innings here. The continued success of Contreras, Pettitte, Rivera and Wang beyond 2005 may bolster the "Post" group (if you're still optimistic that Lilly will blossom, give J.P. Ricciardi a call) and close the gap. Still, this has to be considered a minor feather in Stottlemyre's cap.
One last way of slicing and dicing this data is to note that in Clemens and Johnson, the "Pre" group features the work of two pitchers who came to the Bronx sporting no fewer than 10 Cy Young awards on their mantle, two pitchers who are arguably the best righty and lefty, respectively, of the past two decades if not an even longer timeframe. Some amount of dropoff was inevitable given their late arrivals (Clemens in time for his Age 36 season, Johnson for Age 41). Controlling just for the contributions of those two, and returning strikeout rates to the chart:
Group AGE IP ERA ERA+ K/9 Pre NYY 27.5 31236 3.82 112 6.85 NYY 96-05 30.7 11473 4.03 111 6.84 Post NYY 33.9 7864 4.21 102 6.50Talk about coming out in the wash! Aside from those two venerable aces, we see almost no change in the comparative performances of the "Pre" group and the Yankee group. Even the pronounced drop in strikeout rate is accounted for once the duo who ranks second and third on the all-time strikeout list is removed from the study. Furthermore, it's worth noting that the duo still outperformed the other Yankee pitchers as a group; Clemens' ERA+ as a Yankee was 114, Johnson's was 117.
Taken together, the data suggests that while the Yankee pitchers under Stottlemyre didn't live up to the collective promise they had shown before coming to the Bronx, much of that performance gap is explained by the late-career declines of two all-time great pitchers who were merely very good as Yankees. Smaller segments of the data suggest that inexperienced pitchers had more success under Stottlemyre than their veteran counterparts, pointing perhaps to a failure of the Yankee front office to deliver the kind of hurlers--whether homegrown or acquired via trade--who played to the pitching coach's strengths. Still, the broader trends show that the Yankee pitching staff had been moving backwards for the past couple of years relative to the league, and if Stottlemyre was in fact advocating a more contact-centric approach, he was doing so on a team ill-suited to withstand more balls in play. It may be unfair for him to have endured so much criticism given his track record, but it's also apparent that the time for change had arrived.
In any event, the Yanks are already busy trying to fill Stottlemyre's shoes. First on their list was the Braves' legendary pitching coach, Leo Mazzone, though according to the latest report, he's already rejected their overtures. Credit the Yanks for aiming high; a vaguely similar study in BP 2005 (and another one at Sabernomics) found nearly a 13 percent improvement in pitcher ERAs under Rockin' Leo's tutelage, though once the big guns (Greg Maddux, Tom Glavine, and John Smoltz) were controlled for, the advantage was a more modest five percent. Other names that have been thrown about include White Sox pitching coach Don Cooper, current Yankee bullpen coach Neil Allen, Triple-A Columbus pitching coach Gil Patterson, and former Yankee ace Ron Guidry. One thing is for sure: the new pitching coach might have a tough time lasting long enough to match the uneven track record of his predecessor.
Peter Quadrino and James Click contributed research to this article.