Bullpen Management

Every year, I seem to pick up a pet cause. In 2009, it was Cole Hamels; last year, it was Ryan Madson. This year, that cause is bullpen management. I’ve made no secret that I think Charlie Manuel‘s bullpen management leaves a lot to be desired. In particular, I’ve stressed that J.C. Romero needs to be used exclusively against left-handed hitters as there is such a disparity in performance depending on the handedness of the batter. Additionally, I have been critical of Manuel’s willingness to leave his starters in the game unnecessarily (particularly Roy Halladay).

The last two days have seen some bullpen meltdowns, both literally and figuratively. If you recall, a “meltdown” is a statistic that goes hand-in-hand with a “shutdown”, measuring the effect a reliever had on his team’s chances to win the game. From FanGraphs:

In short, if a player increased his team’s win probability by 6% (0.06 WPA), then they get a Shutdown. If a player made his team 6% more likely to lose (-0.06), they get a Meltdown.

The Phillies entered tonight’s game with the second-fewest meltdowns in baseball with 18, trailing the Cleveland Indians by two. They also had the ninth-most shutdowns with 54. On the whole, the bullpen had been doing a great job in the relatively few innings required of them thanks to the impressive starting rotation.

The last two games, however, have been a different story. After Kyle Kendrick was removed from the game following a rain delay, the Phillies bullpen slowly brought the game back to the Cubs. Eventually, the Cubs would tie the game at 3-3 on a Geovany Soto ninth-inning home run against Ryan Madson. They very nearly took the lead when Tyler Colvin hit a fly ball to right field that was touched by a fan. Initially, it was ruled a home run but was overturned upon video review. David Herndon allowed the fourth and game-winning run in the 11th when Placido Polanco made a poor throw to first base. Herndon and Madson earned meltdowns with -.172 and -.208 WPA, respectively.

Tonight, the Phillies staked Roy Halladay to a 7-0 lead and it looked like smooth sailing going into the eighth. Manuel removed Halladay for pinch-hitter Ben Francisco, figuring his bullpen was adequate for the job. The combination of a not-quite-right Jose Contreras and J.C. Romero (who faced two right-handed hitters) allowed five runs in the eighth inning and were lucky that was the extent of the damage. Michael Stutes came in to get two outs in the eighth and two outs in the ninth before Bastardo finished the game with a strikeout.

In the last two games, the bullpen has pitched ten innings and allowed nine runs. Yet, despite the results, I was mostly pleased with Manuel’s bullpen management. Let’s take a look at the specifics:


– Used Danys Baez after the rain delay. Conservative; smart. Asking Kendrick to go back out after a 77-minute rain delay would have been completely unnecessary.

– Used Romero (LH) to face Carlos Pena (LH) in the sixth inning with a runner on second base and two outs in a 3-1 game. This is how you use Romero properly.

– Used Stutes to pitch the seventh inning. Nothing wrong with that.

– Used Bastardo to pitch to the left-handed pinch-hitter Brad Snyder. Cubs manager Mike Quade countered with right-handed pinch-hitter Lou Montanez. Nothing you can do about that, since the manager of the batting team gets the last “move”. Besides, Bastardo is somewhat good against right-handed hitters, unlike Romero. Manuel left Bastardo in to start the eighth inning against Kosuke Fukudome. Smart.

– Used Contreras to finish out the eighth, as three of the Cubs’ next four hitters were right-handed. Unfortunately, Contreras didn’t have his best stuff and allowed a run on two consecutive two-out doubles.

– Used Madson in the ninth. It sucks that he allowed the tying run and almost allowed the go-ahead run, but Madson had been lights out up to this point. Absolutely nothing to complain about here.

– Used Herndon, his last reliever in the bullpen, to pitch the 10th and 11th innings. You manage the game assuming regulation. In retrospect, it would have been nice to have additional depth going into extra innings, but if you manage in anticipation of going extra innings, you will unnecessarily lose more games in regulation. This is essentially what I was asking for when I criticized Manuel earlier in the season when Madson was left sitting on the bench against the St. Louis Cardinals.


– Took Halladay out after seven innings and 107 pitches. Smart move. No need to leave Halladay in there with such a high pitch count in a game that the Phillies were 95% to win. To leave Halladay in there would imply that the Phillies’ bullpen has the collective skill level of a 31.50 ERA pitcher (seven runs in two innings) or worse. I will take selective match-ups in the final two innings than Halladay after 107 pitches, with all due respect.

– Used Contreras to start the eighth inning. Again, he didn’t look sharp. Fortunately, it was a low-leverage situation — he started the inning with a 0.06 leverage index — so he had room to work through his struggles. He faced five batters, four of them reached base on two walks and two hits; two of them scored.

– Manuel went to Romero to face Blake DeWitt (LH). Romero walked DeWitt. That should have been the end of his night right there, as the Cubs had the right-handed Geovany Soto (career .407 wOBA vs. LHP) due up. The Cubs had the bases loaded and one out, and were clearly not going to pinch-hit for Soto. Perfect opportunity to use Michael Stutes. However, Romero stayed out there and promptly allowed an RBI single. Quade pinch-hit for Tyler Colvin (LH) with Montanez (RH), who sliced a two-run single to right field, bringing the score to 7-5.

– Finally, Stutes was brought in and the game was calmed down. He recorded the next four outs. Bastardo (LH) was brought in to face Pena (LH) and struck him out to end the game.

Overall, the bullpen was managed optimally by Manuel. The only hiccup was leaving Romero in the game against Soto. It is unfortunate that Manuel’s good decision-making was punished, as it may deter him from making these decisions as often in the future. Despite the results, I was encouraged as it shows an evolution in Manuel’s managing style. Of course, it could also be completely random, but I’m betting on someone — maybe even Charlie himself — influencing his bullpen management.

Phillies Just Aren’t Clutch!

Last night’s extra-innings loss to the Chicago Cubs was about as frustrating as they come. The offense once again did not contribute — a Raul Ibanez hit in the 10th inning was the team’s first since the fourth inning — while the bullpen finally had a meltdown. Adding to the frustration was a rain delay lasting longer than an hour and the prospect of having to use another position player on the mound as the Phillies were ill-prepared for a long extra-innings game. At the Good Phight, FuquaManuel recaps the game about as well as one can expect after that sordid affair.

While tweeting during the game, I came across an interesting stat on Baseball Reference. Also depressing, but interesting.

Coming into tonight, the Phillies had managed just a .452 OPS in extra innings, worst in the NL. League average is .701.

With two walks and two singles in the 10th and 11th innings last night, the Phillies actually raised that OPS all the way up to .473, still last in the league.

LAD 5 46 .417 .512 .667 1.178
HOU 2 16 .333 .500 .583 1.083
WSN 10 82 .304 .380 .565 .945
CIN 7 87 .260 .372 .425 .797
NYM 5 49 .279 .354 .395 .750
ARI 7 68 .298 .385 .351 .735
ATL 13 110 .266 .355 .362 .717
MIL 6 50 .179 .292 .410 .702
SFG 10 84 .257 .329 .338 .667
SDP 10 83 .217 .341 .304 .646
PIT 5 51 .222 .300 .333 .633
STL 9 70 .161 .319 .304 .622
FLA 11 72 .265 .292 .309 .600
CHC 4 21 .150 .190 .350 .540
COL 4 58 .173 .246 .250 .496
PHI 6 94 .171 .242 .232 .473

* Phillies stats updated for last night’s game; other teams were not updated.

The Phillies also rank last in the league in OPS in situations deemed “late and close”. They have a .573 OPS, well under the league average .693 and still well behind the 15th-place Washington Nationals at .627.

In tie games, the Phillies have posted a .634 OPS, which ranks 15th out of 16 in the NL, under the .710 league average.

They are great front-runners: when they are ahead, they have the league’s best OPS at .770, more than 70 points higher than the .699 league average. When they’re behind, they leave no hope as their .645 OPS is 14th out of 16.

While the stats would indicate to a layperson that the Phillies aren’t clutch, the real problem lies with the personnel. As mentioned several times here on the blog, Ryan Howard is the team’s only legitimate power hitter left now that Jayson Werth is gone and Chase Utley is on the mend from patellar tendinitis. Few players have a significant on-base skill, such as hitting for a high average or drawing a lot of walks. Most of the players are slow; Jimmy Rollins and Shane Victorino have accounted for more than half of the team’s stolen bases just by themselves.

The goal in baseball, in terms of offense, is to get on base and to advance around the bases as efficiently as possible. In the past, the Phillies had the personnel to do this. In 2007, when the Phillies paced the National League with 892 runs scored, four of eight regulars had an on-base percentage at .370 or higher. Five of them had a slugging percentage at .500 or higher. The Phillies’ current on-base percentage leaders are Placido Polanco and Carlos Ruiz at .362 and .361, respectively. The slugging percentage leaders are Ryan Howard at .479 and Shane Victorino at .472.

If the Phillies want to start scoring runs again, they need to make personnel changes. Whether that involves making a call to Lehigh Valley or making a trade for a bat remains to be seen. We do know, however, that if the Phillies stand pat, they will continue to struggle to score runs. They are currently on pace to score 640 runs. With the great starting rotation, that may just be enough, but it’s a risky proposition.