Perspective with the Bullpen

With the Phillies 9.5 games behind the first place Washington Nationals, there has been plenty of blame to go around. Whether it’s been injuries, the recent incompetence of the starting rotation, or the historically-bad situational hitting, the Phillies have found a way to fail in new and interesting situations. The most consistently-bad aspect of the team, though, has been the bullpen. After Jonathan Papelbon and recently Antonio Bastardo, there simply hasn’t been a reliable arm that Charlie Manuel can call on in important situations. The bullpen’s collective 4.44 ERA is the third-worst in the National League, thanks to the repeated failures of Chad Qualls, Jose Contreras, Joe Savery, and Michael Schwimer, among others who have had a smaller share of innings.

As a fun game, I compiled four stats for each Phillies reliever (min. 10 innings) this season and compared them to the relievers (min. 20 innings) of another very successful MLB team. The stats chosen are K%, BB%, K/BB, and xFIP because they are the quickest to stabilize and the most elucidating.

Go through the stats and make a selection. At the end, you will learn the identity of the mystery team if you don’t put the pieces together before then. Click on only one link for each set.


K% 30% 32%
BB% 4% 12%
K/BB 7.25 2.63
xFIP 2.78 3.01

Select: Player A | Player B


K% 35% 20%
BB% 11% 7%
K/BB 3.10 3.55
xFIP 3.18 3.55

Select: Player A | Player B


K% 29% 20%
BB% 2% 15%
K/BB 13.00 1.37
xFIP 2.94 4.34

Select: Player A | Player B

Four Relievers (ordered by IP)

K% 16% 17%
BB% 6% 10%
K/BB 2.57 1.80
xFIP 3.65 4.31

Select: Player A | Player B


K% 15% 11%
BB% 5% 6%
K/BB 3.00 1.79
xFIP 4.21 4.32

Select: Player A | Player B


  A B
K% 27% 16%
BB% 4% 13%
K/BB 5.00 1.20
xFIP 2.79 4.70

Select: Player A | Player B


K% 14% 19%
BB% 14% 12%
K/BB 1.00 1.53
xFIP 5.58 4.92

Select: Player A | Player B


If you haven’t figured it out yet, all of the B players come from the World Series champion 2008 Phillies. The ’08 bullpen led the league in ERA at 3.22. If you chose correctly according to the stats, you should have picked mostly 2012 Phillies relievers above, and that should surprise you. If the ’12 relievers are better, then why are they near the bottom in ERA and the ’08 bullpen at the top?

  • Grounders: The ’08 bullpen featured three relievers with a ground ball rate above 50 percent: J.C. Romero (61.5%), Clay Condrey (54.3%), and Ryan Madson (51%). The 2012 ‘pen features two, and neither at the height of Romero: Chad Qualls (54.5%) and Jose Contreras (51.4%). As a team, the 2012 ‘pen has a 40.7% ground ball rate compared to the 48.8% rate of 2008.
  • Home Runs: Grounders are good because when they are hits, the extent of the damage is usually a single, and a double in rare circumstances. A fly ball, obviously, can travel beyond the outfield fence for a home run. This is why the average for the two batted ball types are nearly identical, but there is a huge disparity in power (.021 ISO for grounders, .371 for fly balls). The 2008 bullpen had one member allow more than one home run per 10 fly balls: Romero (14.3%). The 2012 Phillies feature four: Qualls (28.6%), Joe Savery (13.6%), Jonathan Papelbon (11.1%), and Jose Contreras (10%). As a team, the 2012 Phillies are at 12% compared to the ’08 Phillies at 8.5%.
  • Sample Size: Two-plus months doesn’t even begin to approach an acceptable data set when you consider we are talking about relievers who individually account for fewer than five percent of his team’s total innings. The bullpen leader in innings is Chad Qualls at 26. Comparatively, starter Cole Hamels reached 26 innings on April 25. Would you use Hamels’ month of April for anything but the most general of observations? It is possible that the bullpen’s collective true talent includes a HR/FB of 12%, but we don’t know that with two-plus months of data. It is more likely that the relievers individually regress towards the league average 10%, which means Qualls won’t give up homers at a 28.6% rate going forward, for example.
  • Defense: The Phillies remain one of baseball’s elite defensive teams, but the data includes a large majority of games involving Freddy Galvis and Placido Polanco. Galvis was recently injured and will be out for “a significant amount of time“. Galvis came out in the fifth inning on June 6, which was the first game that Polanco was absent as well. In the seven games that followed, the Phillies committed 11 errors, including eight by infielders (Ty Wigginton, 4; Mike Fontenot, 2; Jimmy Rollins; Michael Martinez). There were several more that weren’t scored as errors as well. Of course, errors are a very rough way to measure defense, but there’s no game-by-game UZR (and even if there was, it’d be just as questionable). Generally speaking, though, the Phillies’ defense has been much worse recently than it was to start the season.

Based on the things a pitcher has the most control over, the bullpen has actually done a decent job. Their collective 2.89 K/BB is second-best in the league behind the Cincinnati Reds (2.93) and fourth among all 30 teams. Their 3.75 xFIP places them in the top-ten in the Majors as well. Using SIERA, which better accounts for batted ball profile, the Phillies move into the top-five at 3.16. The bullpen isn’t nearly as bad as they have looked through 64 games, and that is an important realization because if the Phillies are able to make up some ground in the next month and go into the deadline looking to add players, they should abstain from paying for relievers based on two months of volatile bullpen data.