clock menu more-arrow no yes mobile

Filed under:

What would an average bullpen mean for the Mets?

The Mets' bullpen has been terrible the past few years. If the bullpen could be merely not terrible, what would that mean for the 2014 Mets?

Can Gonzalez Germen help the Mets bullpen be not terrible?
Can Gonzalez Germen help the Mets bullpen be not terrible?
David Richard-USA TODAY Sports

It's no secret that the Mets' bullpen has been pretty poor under the Sandy Alderson regime. In fact, "pretty poor" doesn't quite do justice to just how bad the bullpen has been. "Ridiculously terrible" is probably a bit more accurate, or we could go with "piss poor" if we want to stick to alliteration.

How bad have they been? From 2011 through 2013, the Mets' bullpen ranked second to last in ERA- and FIP-, which are park- and league-adjusted versions of those stats. The only team the Mets have been better than is the Astros, and they've basically been losing on purpose the past three years.

Another metric they've ranked second to last in is fWAR, Fangraphs' version of wins above replacement, a stat that needs no introduction to regular readers of this site. We're fond of context-neutral stats like WAR and FIP- because they're stats that get us closer to identifying a player's true skill level. If I tell you that the Mets bullpen has been below replacement level over the past three years, you understand it means they've been ridiculously terrible.

But as good as these context-neutral statistics are, they don't tell us exactly how much impact a player actually had on the game. When you're up by two and you allow a three-run homer, that's much worse than a solo home run allowed in the sixth inning of a blowout loss. Context-neutral stats don't capture this distinction.

A stat that does capture this distinction is win probability added (WPA). It's a counting stat that's calculated on a per-play basis based on win expectancy. The solo home run allowed in the sixth inning of a blowout loss barely moves the needle on a stat like this because a team's win expectancy barely changes as a result. The three-run home run allowed in the bottom of the ninth of a game your team led by two will have a much larger negative impact on your WPA.

We can use WPA to measure just how bad the Mets' bullpen has been. Here are the values the bullpen has posted over the past three seasons; -3.60, -5.55, and -2.77. That's an average of -3.97 per season, which basically means the bullpen has cost the Mets four games a year during the Alderson era. That is a serious headwind to have to fight every season.

WPA is a counting stat and thus is not centered around zero. There don't have to be a specific number of teams with a negative WPA. Over the past three seasons there have been an average of nine teams with negative WPAs. A middle of the pack bullpen adds anywhere from one to three wins per year.

So let's dream big and hope the Mets bullpen can become one of the 20 or so units that actually add to their win probability. The difference over the typical Alderson bullpen would not be insignificant: about four actual, non-hypothetical wins. And if they can turn that headwind into a tailwind, it could be the difference between a non-contender and one that flirts with a wild-card.