The General Chapter of the Baseball Bloggers Alliance, comprised of those of us who don’t write about a specific team, fantasy baseball, or baseball history, has released its ballot for the Goose Gossage Award, which goes to the top reliever in each league. While I respect those bloggers who only cover one team, I put a little more stock in the General Chapter’s ballot than the Alliance’s overall ballot, as we’re the bloggers who spend the whole year analyzing every team. As such, I couldn’t contain my pride in the chapter when I learned that we named Mariano Rivera the American League’s top reliever in 2011.
Why am I so excited? I named Rivera third on my ballot, and my top two choices, Jonathan Papelbon and David Robertson, finished fourth and third, respectively. Rivera pitches for my least favorite team, and has been agonizing me for a decade and a half, so I have no personal attachment to him. I’m excited because I’m proud to be a member of a group of baseball writers who, for the most part, where not fooled by Jose Valverde’s save percentage.
Let’s get this out of the way first: Valverde had a great season. In 72 1/3 innings, he struck out almost a batter an inning and had an impressive 2.24 ERA. His primary job is to preserve leads handed to him in the ninth inning, and he did so every time he was asked to.
Why, then, should Valverde not win the Goose Gossage Award? Because awards don’t evaluate players in the context of their own expectations. Awards compare players’ performance to that of other players. And several other relievers were better than Valverde.
There are two reasonable ways to measure a relief pitcher’s success: run prevention and true outcomes. In terms of earned run prevention, Valverde was excellent, his ERA ninth best among qualified AL relievers. In addition to his 18 earned runs, he also gave up three unearned runs, for which he shouldn’t be completely absolved. His 2.62 “RA” would rank 12th in the AL- still impressive, but not necessarily Goose-worthy.
Run prevention isn’t always the best way to evaluate relievers, since some runs are given up while one pitcher is on the mound but charged to another pitcher. Relief ERA is also subject to small sample size, as relievers typically throw just 50-75 innings in a season. I prefer to evaluate relievers based on true outcomes- that is, strikeouts, walks, and home runs allowed, which both evaluate talent and predict future results.
Fielding Independent Pitching, which is scaled to ERA but only recognizes outcomes within a pitcher’s control, ranks Valverde 25th among qualified AL relievers. His 3.55 FIP suggests that with average fielders behind him and normal luck, his ERA would have been about a run higher. He walked 4.23 batters per nine innings, 11th worst among AL relievers.
How, then, did Valverde convert every save opportunity this year? For starters, the league batted just .247 when putting the ball in play against him. Those walks didn’t hurt him much either, as he stranded 82.9% of his baserunners and gave up home runs on just 6.4% of fly balls he coughed up.
There are two different elements of fortune at play here. Some combination of good fielding, rising to the situation, and luck kept his ERA more than a run lower than expected. Furthermore, the runs Valverde gave up were mostly in non-save situations. He lost four games in which he entered with the score tied, and he gave up three runs in save situations in which he entered with a two- or three-run lead. The average closer in the American League blew one save for every 4.47 runs surrendered this year, so with a normal distribution of runs throughout his appearances, one would expect that Valverde would have blown four to five saves based on his RA and six to seven based on his FIP.
Of course, Valverde didn’t blow those saves, and pretending he did doesn’t evaluate his season any more properly than ignoring the non-save situations in which he did give up costly runs. But to endorse Valverde’s candidacy based on his perfect save record, one would have to ignore (or at least discount) his performance in non-save situations. As we’ve seen in the postseason, Valverde’s ability to lock down a one-run lead is immensely valuable, but when he’s called upon in other situations, like the extra-inning tie in Game 4 of the ALCS, Valverde’s wildness and hittability can be damaging.
But enough about Valverde. As I noted earlier, choosing a Goose Gossage Award winner is not just about picking apart Valverde’s season; it’s about comparing Valverde to other relievers.
In terms of run prevention, David Robertson (1.08 ERA) was the league’s best reliever, while true outcomes suggest that Papelbon (1.53 FIP) was best. But the General Chapter of the BBA tabbed Rivera, so let’s look at his numbers.
Rivera pitched 61 1/3 innings and gave up 13 runs, all earned. His 1.91 ERA was good for fourth among AL relievers (tops among closers, for what it’s worth). Rivera struck out roughly the same number of hitters per inning as Valverde, but walked less than a third as many. His 2.19 FIP placed third in the AL. Rivera distributed his runs somewhat less fortunately, blowing five saves in 49 opportunities, and he also took the loss in two games. But in terms of preventing runs and the underlying numbers that typically help pitchers prevent runs, Rivera was better than Valverde in 2011.
I’m glad my peers noticed.