What I Learned in 2015: Relievers Don't Grow on Trees
I know it’s2016 and this is totally gauche now, but since we didn’t have a proper website the last week of 2015 I’m calling a mulligan to talk about the biggest baseball thing I learned in 2015. That relievers, at least actually good ones, aren’t nearly as fungible and limitless as I and many like me believed.
Before the season I was a strident believer that relief pitchers could be found on trees. All one had to do was sign a bunch of near-MLB quality starting pitchers and let them have a Battle Royale in spring training to find the seven best guys to come north. With failed starters like Wade Davis, Andrew Miller and Dellin Betances dominating in the majors as relief pitchers and hometown guys like Tyler Clippard and Craig Stammen upholding that tradition, it seemed like a slam dunk. Unlike other positions, relief pitchers did not have to be groomed for years in the minors, they could just be plucked on the cheap from the ever-swelling ranks of not quite good enough starting pitchers.
That all changed after I watched the trainwreck that was the 2015 Washington Nationals bullpen. The Nats rotated through 20 different pitchers and 2 position players getting at least one relief appearance. 12 different pitchers threw in at least 10 games, 14 threw at least 10 innings and pretty much all of them were terrible. Of the pitchers with at least 10 innings in relief, only five had an ERA below 3.35: Matt Thornton, Felipe Rivero, Doug Fister, Taylor Jordan and Jonathan Papelbon. The bullpen as a whole had 82 meltdowns (when a relief pitcher lowers his team’s win expectancy by at least six percent), the seventh most in MLB, with the six teams ahead of them all either in last or second-to-last in their division.
Now I’m not saying that there isn’t a real positive effect when a pitcher moves from being a starter to being a reliever. In fact, last season there were 13 pitchers who threw at least 50 innings as a starter and at least 20 innings as a reliever. On average, these 13 pitchers saw their K/9 rise by 1.45 strikeouts, their BB/9 rise slightly by 0.14 walks, their ERA drop by 1.20 runs and their FIP drop by 1.09 runs when pitching from the bullpen.
Of course the key point we’re missing from this chart and most analysis of the effect is that just because they improved it doesn’t mean they were good. In fact, six out of the 13 pitchers still had an ERA above 3.70 and seven out of 13 had a FIP above 3.70. While most pitchers can see a bump in their effectiveness when moving from the rotation to the bullpen, that doesn’t mean they become supermen. Getting guys out at a Major League level is hard no matter how many times you have to face them.
There’s a reason why when I detailed the Nationals bullpen’s failure I specifically pulled the meltdowns stat. Because, by many other measures they seemed perfectly average: 10th in ERA, ninth in FIP, 12th in fWAR, 23rd in WPA and 12th in RE24. By these numbers, which smooth out individual triumphs and failings, the bullpen doesn’t look like it was all that bad. Of course anyone who watched the Nats all season could tell you it was, mostly because of the numerous times where they solely cost the Nationals a game they had in hand, as seen in their meltdowns.
This is another thing that many miss in modern analysis of bullpens and relievers: the unique stress and pressure that a reliever endures. Few notice the number of times a reliever warms up before they actually go into the game, or how they only got five warmup tosses in before being thrown into an emergency, and most of the above stats (besides WPA) don’t account for how backbreaking that one run in the eighth is compared to one run in the second inning. It’s laughable to think that a starting pitcher could easily and quickly adapt to become a late-inning reliever after being able to warm up at their leisure because they knew exactly when they would start pitching, in the first inning, where there is little pressure.
So while relief pitchers don’t grow on trees, they do still come up in a non-traditional manner compared to starting pitchers and especially position players. And with the small season samples relievers work in it’s hard to get an accurate idea of how good a guy is for a while, outside of the truly dominant. This can make it appear like they aren’t worth a large investment compared to other positions, however the Nationals have proven that can’t be further from the truth and other teams like the Royals, Yankees and Red Sox have seemingly figured this out as well.
I predict as the years continue we see more teams at the top of the spending pile try to form super bullpens of two to three closer quality relievers. Of course, that is no guarantee of success either, as a lot more lays on the feet of the manager to keep all parties happy, like the Nationals failed to do when they brought in Jonathan Papelbon to team up with Drew Storen. Investing more money into dependable middle and late relief should also catch on, as the Nationals showed with their spending spree on Shawn Kelley, Yusmeiro Petit and Oliver Perez. As they learned, the seventh inning can be just as important as the eighth and ninth and having a good reliever for that situation can be the difference between the playoffs and disappointment.