September 25, 2008
Prospectus Hit and Run
The Mets Bullpen Failures
If the Mets miss the playoffs this year, they'll remember August 2 as the day their season irrevocably turned. In the ninth inning of a game against the Astros, interim manager Jerry Manuel summoned Billy Wagner to protect a 4-2 lead. Things quickly went awry, as the closer allowed four of the first five hitters to reach base, with Geoff Blum hitting a two-run single to knot the game. It was Wagner's seventh blown save on the year, and it would be his last. Within a week, the 36-year-old lefty went onto the disabled list with what was reported as a recurrent strain in his forearm. A month later, just as he'd returned to throwing at full velocity, he was discovered to have a torn ulnar collateral ligament that required Tommy John surgery, which cost him the rest of 2008, and will also cost him most of 2009.
That the Mets remain in the hunt for a post-season berth despite Wagner's injury is something short of a miracle, given the other obstacles they've had to overcome. Despite the drama of last year's collapse, they entered 2008 with the league's highest payroll and top PECOTA forecast, and they've largely lived up to their promise:
On the whole the Mets are third in the league in run differential and tied for third in wins. For them not to be in the playoff hunt would be an upset.
However, Wagner's injury exposed the team's Achilles heel, their bullpen, and if the Mets don't make the playoffs, they can hang a good share of the blame on a unit that hasn't come close to holding up its end of the deal. The Mets have lost 28 games in which they were either ahead or tied after six innings, the most in the majors. That's partly a function of opportunity (they've played more games in which they were ahead or tied after six than any team) and partly a failure to execute (they're 5-15 when tied after six). They rank dead last in the league and 25th in the majors in Reliever Expected Wins Added (WXRL), unsurprisingly the worst among the dozen teams that retain a mathematical chance of playing in October. Updating Joe Sheehan's chart from a week ago:
Rank Team WXRL 1 Rays 15.3 2 Phillies 14.3 4 Angels 13.1 5 Dodgers 12.1 7 Astros 11.0 10 White Sox 9.6 11 Cubs 9.0 12 Twins 7.7 13 Red Sox 7.4 16 Brewers 7.2 20 D'backs 5.8 25 Mets 4.7
The ranking is by overall standings in the majors, and make no doubt about it where the Mets are concerned-that's ugly. In fact, if the Mets reach the playoffs without much improvement from their bullpen, their WXRL total would rank among the 10 worst for a playoff team since 1988, a year chosen to represent the dawn of the one-inning closer era:
League Year Team WXRL Rank Won 1997 Mariners 0.3 14 Division 2005 Red Sox 0.5 14 Wild Card 2005 Braves 1.1 16 Division 1995 Yankees 1.8 13 Wild Card 1990 Red Sox 2.3 12 Division 2003 Red Sox 3.9 11 Wild Card, Division Series 1988 Red Sox 4.7 8 Division 1992 Braves 4.8 8 Division, Pennant 2002 D'backs 4.9 13 Division 1997 Astros 5.5 7 Division
Not surprisingly, the teams that made the playoffs despite such bullpen problems didn't last long; only two out of 10 won a playoff series, and nobody won any championships. Limiting the discussion to teams like this year's Mets who were in the bottom quartile of their respective leagues in WXRL-as I did in performing a recent postmortem of the Diamondbacks-narrows the list to seven, with one series win. This just in: a bad bullpen isn't part of a winning recipe.
The sad fact is that through the first half of the season, even with Wagner performing at a level far below his peak, the Mets actually had one of the league's better bullpens; their 6.1 WXRL through the All-Star break ranked third in the NL behind the Phillies and Dodgers. Since then, they've been a league-worst 1.4 wins below replacement level as a unit. Replacement closer Luis Ayala (-0.04 WXRL since coming over from the Nationals on August 17) is an obvious culprit, but he's hardly the only offender.
A quick peek at the individual numbers informs us that it's not hard to recognize a systemic combination of overuse and ineffectiveness. Of the six relievers Jerry Manuel has called upon most frequently, five have second-half ERAs above 4.90: Ayala (5.54, including his Washington stint), Pedro Feliciano (6.38), Aaron Heilman (6.75), Duaner Sanchez (6.00), and Joe Smith (4.91); Scott Schoeneweis (4.50) is the exception. Excluding the late-arriving Ayala, that bunch has combined for 152 appearances in the 63 games since the break, a breakneck 78-game pace for each over the course of a season. Feliciano (83 games), Ayala (80), and Smith (79) represent three of the six major league pitchers stretched to that exhausting plateau over the full season, with Heilman (77) not far behind. Overall, the Mets rank second in the league since the break with 227 relief appearances, an average of 3.6 per game.
Driving such a frenetic pace is a massive platoon split that has Manuel chasing the "right" matchups, following a single-minded La Russa-style tactical orthodoxy at the expense of more important strategic imperatives such as conserving bullpen arms over the course of the long season. When they have the platoon advantage (righty on righty or lefty on lefty), Mets relievers have limited hitters to just .225/.299/.325; ranked by OPS, that ranks an impressive fourth in the majors. However, when they don't have the platoon advantage, they've been tagged at a .294/.375/.479 clip, worst in the majors. The 227-point OPS difference between situations is the highest by a wide margin; second-highest are the Brewers at 188 points, and they just whacked a manager over his platoon-related shenanigans and bullpen mismanagement. The take-home message is yet another reminder that chasing matchups can easily backfire on a skipper, either by exposing lefty specialists like Schoeneweis (.333/.421/.509 versus righties) or Feliciano (.357/.453/.561) to the point where they face more righties than lefties, or by shunting a heavier workload to the second- or third-tier pitchers in a bullpen.
Yet for all of those woes, things might be different if Wagner were still around. Despite a superficially tidy 2.30 ERA, the five-time All-Star had accumulated just 1.5 WXRL in about two-thirds of a season, after compiling 3.8 last year and 5.9 in 2006 (second in the league). Depending upon which model of Billy Wags you use as a benchmark, that's anywhere from one to four wins missing from his ledger. Even at its lowest, that margin may easily be the difference between a club playing their way into October and adding another season like their now-infamous 2007 collapse to give them a matched pair of late-season meltdowns.