Are the Pirates Killing their Bullpen?

The Pirates beat the Reds 3-2 yesterday, ending a three-game losing streak, as five relievers preserved the win for starter Jeff Locke. Bryan Morris, Justin Wilson, and Mark Melancon earned holds (Melancon his NL-leading 26th), and Jason Grilli his league-leading 30th save.

This kept the Pirates not only in second place in their division but also have the second best record in the National League and the third best in all of baseball, an improbable turn for a team that famously haven't had a winning record since George Bush was in the White House. The first President George Bush.

The most prominent (and most discussed) feature of this year's Bucs has been its bullpen.

Pirates starting pitchers have a 3.31 ERA, best in the league, but that figure is barely ahead of the Cardinals, Dodgers, and Reds. The bullpen ERA of 2.74 is second to the Braves, but has been the key to the team's success. (It hasn't been the offense: The Pirates have scored the third-fewest runs in the league.)

Here are some random facts about the Pirates bullpen. All comparisons are to their National League counterparts. Yes, there's a difference between the leagues; relievers appear more in the DH-less NL.

• Pirates relief pitchers have made 275 appearances, slightly below the average of 283.Their relievers are second in relief wins, with 21, and second-to-last in relief losses, with 11.
• They have 34 saves, the most by five. With just nine blown saves, they're also first in save percentage, 79%.
• The middle relievers lead in holds, with 55.
• The relievers have inherited 105 baserunners and allowed just 21 to score. The 20% of inherited runners scoring is easily the lowest in the NL, where the average is 28%.

Those are all impressive stats. But here's the dark lining of the silver cloud: Manager Clint Hurdle has relied heavily on his relievers. Setup man Mark Melancon has appeared in 46 games, 11th most in the league. Closer Jason Grilli's been in 43 (21st most). Tony Watson's been in 40 (37th), Justin Wilson 37 (50th), Vin Mazzaro 32 (61st), Bryan Morris 28 (71st). Only Cincinnati has had more relievers appear in 28 or more games, and only Milwaukee has had more relievers appear in 30+. And most Pirates relievers have seen an uptick in innings:

• Grilli: 58.2 IP last year, on pace for 68.2 this year
• Melancon: 45.0 IP last year, on pace for 76.2 this year
• Watson: 53.1 IP last year, on pace for 78.2 this year
• Wilson: 135.2 IP at AAA and 4.2 in the majors last year, on pace for 89.1 this year
• Mazzaro: 67.0 IP at AAA and 44.0 in the majors last year, on pace for 74.2 this year
• Morris: 81.0 IP at AAA and 5.0 in the majors last year, on pace for 67.2 this year

What happens when relievers throw that many innings? To answer that, I looked at every National League reliever from 2010 to 2012 who pitched 67 or more innings. (That may not sound like a lot of innings, but fewer than half of relievers with 30 or more saves logged that many innings. Mariano Rivera hasn't since 2008.) I checked how they did in the second half of the year compared to the first. I used on base plus slugging (OPS) as my metric, since ERAs for relievers are not necessarily reliable. I grouped every reliever into one of five categories:

• Same in each half: OPS within 25 points, plus or minus, in each half
• Better in second half: OPS 25-75 points lower in second half
• Way better in second half: OPS 75+ points lower
• Worse in second half: OPS 25-75 points higher in second half
• Way worse in second half: OPS 75+ points higher

Here's how the 81 pitchers with 67+ innings pitched broke down:

• 14% same
• 20% better
• 23% way better
• 16% worse
• 27% way worse

That's not too scary. A high-innings reliever is equally likely to be better in the second half of the season than in the first. There's a slightly higher chance of a collapse, but that's not a surprise, given that pitchers with a really strong first half are going to be given more opportunities in the second half.

But we can't let the Pirates off the hook just yet. Included among those 81 pitchers with 67+ IP in 2010-2012 are eight Pirates:

• Joel Hanrahan 2010, OPS .641 in first half, .658 in second half (same)
• Evan Meek 2010, .506/.605 (way worse)
• Jose Veras 2011, .578/.705 (way worse)
• Chris Resop 2011.643/.899 (way worse)
• Daniel McCutchen 2011, .701/.776 (worse)
• Joel Hanrahan 2011, .514/.579 (worse)
• Jared Hughes 2012, .654/.706 (worse)
• Chris Resop 2012, .711/.738 (worse)

The Pirates have not had a high innings reliever improve in the second half of the year during this decade. They're clearly an outlier of the wrong kind. That could be a matter of personnel (Hughes is the only pitcher still with the team, and he's been disabled since the beginning of June), or coaching, or managing. But it is enough to make me think that the Bucs need stronger hitting and a stable rotation to stay in playoff contention this year. Because it's probably not realistic to expect the bullpen to remain this good.