Happy Thanksgiving! Regularly Scheduled Articles Will Resume Monday, December 1
April 5, 2011
Prospectus Hit and Run
NL Bullpen Blowout
Twice during the first four days of the season, ESPN televised matchups between the defending world champion Giants and the Dodgers at Dodger Stadium. Opening night featured a taut pitchers' duel in which Clayton Kershaw got the upper hand on Tim Lincecum, while on Sunday night, the Dodgers outlasted the Giants thanks largely to Aubrey Huff's sloppy outfield play. Even as the Dodgers were winning, both nights gave their fans cause for anxiety when rookie manager Don Mattingly summoned Jonathan Broxton to protect multi-run leads in the ninth. Broxton shut the door both times, but only after yielding solo homers in each instance, to Pat Burrell and Aaron Rowand, lowering his margin for error. (To be fair, he sandwiched those two appearances around a spotless save in the season's second game, out of the national TV spotlight.)
Up until the middle of last season, Broxton had shown himself to be a dominant and reliable successor to Eric Gagne and Takashi Saito, a 300-pound behemoth armed with a triple-digit fastball and a wicked slider. From Opening Day 2008 through June 26, 2010, Broxton struck out more hitters per nine (12.7) and yielded fewer homers per nine (0.3) than any other pitcher in baseball who threw at least 150 innings. He led the league in WXRL in 2009, and was near the top again in 2010. Sure, he'd blown a save in the 2009 postseason, but so did every closer not named Mariano Rivera, and only the fact that both that failure and his previous one—in the 2008 NLCS, when he served up a tater to Matt Stairs—came against the Phillies defined it as a trend.
On June 27 of last year (coincidentally, during a Sunday night ESPN telecast), Broxton suffered a meltdown that came to define his season. Handed a four-run lead against the Yankees, he flailed for an agonizing 48 pitches while manager Joe Torre fiddled. By the time the dust had settled, the game was tied; the Yankees would win in the 10th inning. Broxton had come into the game having converted 16 out of 17 save opportunities, with a 0.83 ERA and 48/5 K/BB ratio in 33 innings. He hadn't allowed an earned run in over a month, or surrendered a single homer all year. From that night onward, he was rocked for a 7.58 ERA, with a 25/23 K/BB ratio and four homers allowed in 29.2 innings; after blowing four out of 10 save opportunities over the next eight weeks, he was supplanted as the closer, with veteran lefty Hong-Chih Kuo and rookie righty Kenley Jansen doing most of the heavy lifting from late August onward.
Beyond noting a dip in velocity—down 2.5 MPH on his average fastball, and 1.6 MPH on his slider—nobody ever adequately explained whether Broxton's woes were rooted in fatigue, injury, mechanical lapses, or simply adjustments made by hitters. Back in mid-2009, inflammation in his right big toe led to a brief midsummer swoon and caused him to sit out the All-Star game, but other than a late July illness, he was never listed as even day-to-day for any injury in 2010.
The Dodgers restored Broxton to the closer role this spring, in part because as dominating as they can be, both Kuo and Jansen come with limitations. Kuo held hitters on both sides of the plate to a combined .139/.211/.192 line in 2010—that's one homer and 73 strikeouts in 60 innings—but he's a survivor of four elbow surgeries, including two Tommy Johns. The Dodgers, not wanting to push their luck, worked him on back-to-back days just four times last year. Jansen was a light-hitting catcher as recently as mid-2009; after converting to the mound, he rocketed through the system and joined the big club with just 56.2 minor league innings under his belt. Batters hit a feeble .130/.259/.163 against him, as he whiffed 41 in 27 frames, but walks were a problem: he issued them at a clip of five per nine, and his short history on the mound means that he's fairly raw and has never really faced adversity in that capacity.
Furthermore, the rest of the Dodgers bullpen is in a state of upheaval. Righties Ronald Belisario and Ramon Troncoso, so crucial to the team's success in 2009, were absent from the Opening Day roster after extending their 2010 woes into 2011. Belisario, beset by personal problems that have complicated his visa status, is currently on the restricted list, while Troncoso, who was ridden hard by Torre, is bound for Albuquerque, where he spent nearly two months last summer trying to recover his previous form. Over the winter, general manager Ned Colletti shelled out three years and $12 million for Twins mainstay Matt Guerrier, who doesn't miss enough bats to be worth that kind of bread, while the voids created by Belisario and Troncoso have been filled by such retreads as Lance Cormier, who was nontendered by the Rays last December, and Mike MacDougal, whose obscene walk rate (5.6 per nine) has kept him shuttling among 11 Triple-A and major league teams over the past five seasons while totaling 161.1 innings at the big league level.
Despite the questionable assemblage behind Broxton, Kuo, and Jansen, the Dodgers bullpen projects to be the NL's second-strongest unit according to PECOTA. To determine this, I tallied the top five playing-time adjusted WARP totals among each team's relievers, with a few caveats: relievers with less than 25 projected innings were excluded from the sample, as were swingmen who projected to pitch out of the bullpen in less than 40 percent of their appearances; also, in one case I swapped out a low-inning injury case in favor of a low-projected closer (John Axford). PECOTA-wise, most teams are into the sub-replacement realm by the time you drill past the fifth man, and on top of the already difficult challenge of predicting a class of players who get relatively little playing time in the first place, it's even tougher to predict how long teams will stick with potentially subpar performances given the availability of other options. While the resulting WARP totals are relatively minuscule given Colin Wyers' tweaks to our valuation system, we can get a feel for which bullpens the system expects to do well and which... not so much.
With Luke Gregerson and Mike Adams setting up Heath Bell, the Padres project to have the league's strongest bullpen, one that's largely the legacy of former general manager Kevin Towers, who traded for all three righties. Each member of that trio is forecast to be worth at least 1.2 WARP while posting ERAs below 3.00 and strikeout rates above 9.0 per nine. LOOGY Joe Thatcher and swingman Wade LeBlanc round out the quintet in respectable fashion.
The Dodgers' quintet—filled out by Guerrier and swingman Vicente Padilla, who's quickly working his way back from nerve entrapment surgery—ranks a solid second behind the Padres', with Kuo's 1.4 WARP forecast edging that of Broxton (1.3), and Jansen (1.1) not far behind. Of course, not even PECOTA can see inside the mind and body of the big Broxton to know whether he'll remain such an enigma. When he re-signed for just $2 million plus incentives, Padilla was mentioned as a potential fill-in closer should Broxton remain deposed; in his short time back in the NL (134.1 innings since late August 2009), Padilla has ratcheted his strikeout rate to 8.2 per nine while walking 2.4 per nine, compared to 6.1 and 3.2 prior to coming to L.A.
The Giants, Reds, and Rockies occupy the next tier. For the defending champions, the surprise is that Sergio Romo (1.3 WARP, 2.80 ERA) forecasts as more valuable than closer Brian Wilson (0.9, 3.11), but that's a function of the former's outstanding walk rate, which dipped to 1.7 unintentionals per nine last year, and the latter's starting the year on the DL due to an oblique strain. Wilson is forecast for 66 innings, after throwing 74.2 innings last season plus another 11.2 in the postseason. Those aren't unreasonable amounts, but it's worth noting that aside from Mariano Rivera, recent series-winning closers such as Keith Foulke, Bobby Jenks, Jason Isringhausen, and Brad Lidge have hardly been immune to injuries and/or significant declines in the year after. The remaining unit identified via WARP includes Dan Runzler, Ramon Ramirez, and Jeremy Affeldt, while omitting Santiago Casilla, who's coming off a career year after getting rocked for a 5.11 ERA over parts of six seasons in Oakland, but who may have turned a corner.
More skepticism is merited regarding the Reds' ranking. I've already pointed out that PECOTA isn't terribly sanguine about the Cincinnati rotation even without knowing that it's Dusty Baker calling the shots, and the same goes for closer Francisco Cordero (0.4 WARP, 3.86 ERA). Count me as somebody who will take the under on Aroldis Chapman's projection for 112 innings and even 11 starts; it's his 1.8 WARP that's driving this ranking, more than he'll contribute if he's confined to a setup role. Bill Bray for 63 innings as their third-most valuable reliever? He's managed just 75.1 innings over the past three seasons, and has never topped 47.
The good stuff from the Rockies' ranking rests entirely on the projections for Huston Street and Rafael Betancourt, forecast for 1.3 WARP apiece with similar ERAs (3.36 and 3.31, respectively). PECOTA doesn't think much of Matt Belisle (4.75 ERA, 0.0 WARP) despite last year's breakout, and after he tied for the league lead in relief innings, it's fair to wonder whether he'll be able to repeat his success with a similar workload.
Of the three teams on the next tier, I've discussed the lack of depth behind Carlos Marmol and Kerry Wood recently, and lauded the prospect of the aforementioned Towers turning the Diamondbacks' historically awful bullpen into something less awful than that via the acquisition of J.J. Putz. The headscratcher is the Braves. Manager Fredi Gonzalez has begun the year with rookie Craig Kimbrel and high-velocity lefty Jonny Venters as co-closers, but PECOTA hates the latter, forecasting a 4.75 ERA and -0.5 WARP even after his strong rookie campaign (83 innings, 1.95 ERA, 101 K/9, 1.3 WARP). Venters missed most of 2008, and his work as a starter in 2009 wasn't so strong, so you can sort of see why the system sees regression ahead, but it's reasonable to believe he'll be much better than that. In any event, his total isn't included in the Braves' ranking, while Kimbrel, Peter Moylan, Eric O'Flaherty, George Sherrill, and Scott Linebrink make up the quintet of choice.
As for the Brewers, the system's very down on Axford, forecasting a 4.62 ERA and -0.2 WARP, but then 27-year-old rookies aren't guys who are likely to have impressed PECOTA anytime in the recent past. Ageless Takashi Saito carries by far the best forecast of the remaining bunch (1.0 WARP, 3.05 ERA), with Zach Braddock, Mitch Stetter, and Sean Green within a run or two of replacement level. PECOTA isn't too wild about the Cardinals, either, with Jason Motte's 0.4 WARP forecast heading the list, and Mitchell Boggs' -0.8 (not included in the tally) looking particularly bloody. Closer Ryan Franklin comes in at 0.2 WARP with a 3.93 ERA, but he's been in the league's top seven in WXRL in each of the past two seasons, and his career .273 BABIP defies our forecasting system. Expect both NL Central contenders to have better bullpens than shown, though not as good as the Reds'.
Near the bottom of the table is some grim news for the Phillies, for whom Lidge is forecast to toss just 25 innings (one reason I chose that number as a cutoff). His projected 0.1 WARP owes much to him gritting out an awful 2009 camaign, though let's face it, replacement level could be optimistic for a guy with a Grade II rotator cuff strain that may sideline him until the All-Star break. PECOTA remembers replacemen closer Jose Contreras (-0.1 WARP, 4.37 ERA) less than fondly from his days as a starter, though it's worth noting that the ancient Cubano took to a relief role surprisingly well, striking outa batter per inning and setting a career low for walk rate (2.5 per nine). Ryan Madson (0.5 WARP, 3.60 ERA) is another key to the equation; if he can climb back to the 75-80 inning level he occupied in 2008-2009, instead of the 59 innings we've forecast for him, that will help as well. On the other hand, when LOOGY types Antonio Bastardo and J.C. Romero, projected for a combined 79 innings, qualify as the upside while concealing Kyle Kendrick, Danys Baez, and even lesser lights (if such a thing is possible), you can understand why the Phils' pen rates as a distinct minus at the moment.
Given that, and the four-game spread between the Phillies, Braves, and Marlins in our Depth Chart projections, it's worth noting that the Fish in the bullpen are looking green around the gills. Leo Nunez (0.3 WARP, 4.04 ERA) is nobody's idea of a classic closer, and the verdict is decidedly split regarding the pair of former Padres (Edward Mujica and Ryan Webb) who arrived via the Cameron Maybin deal. Mujica's forecast (0.4 WARP, 3.92 ERA) is the sunniest among the Marlins' relievers, while Webb's (-1.1 WARP. 5.42 ERA) is so bad that he falls out of the quintet of choice, with Clay Hensley and Michael Dunn (-0.3 and -0.6 WARP, respectively) only slightly less underwater.
In any event, using PECOTA to pick on the pens is admittedly an imperfect exercise, but this does suggest something close to bullpen parity in the three divisions, with the West featuring the by far the best ones, the East the worst, and the Central… somewhere in the middle, as usual. Provided Broxton comes around and does what the Dodgers and our projection system expect them to do, their bullpen should be able to support one of the league's top rotations, and aid their quest to return to the postseason.