Relievers are weird. Not only because they do strange stuff like play video games on the side, but they are probably also the most volatile position in baseball. Go look up the best relievers of 2015 and have a good laugh at some of the names that appear. Liam Hendriks? Kevin Siegrist? Luke Gregerson? Josh Fields? The top 10 or so names have been relatively consistent and are still playing at a high level now, but after that, there is little to no guarantee that production will be sustained at all. Contrast that with the best position players the same year and it seems that relievers are about as dependable as the current economy.

Or even just consider the Twins bullpen this past year, who would have predicted that Taylor Rogers would add a slider and become the Terminator basically from July until the end of the season, or that Trevor Hildenberger would continue his great 2017 season for the first half or so until the ghost of Matt Capps took control of his body and he struggled from August until the end of the season, or how about Addison Reed starting off en Fuego and then losing velocity as the year went on until he couldn’t even strike out my grandmother.

The point is, relievers are unpredictable, but what if we could predict them? More importantly, if we could predict if a reliever will bounce back? Ignore the snake-oil-salesman-style question and let me explain my thought process. I want to be able to identify how likely it is for relievers to bounce back, so I looked at data from 2016, 2017, and 2018 to see how many relievers came back from a poor season and how many didn’t. To be exact, I am looking for relievers who had good/great 2016 seasons, noticeably poor 2017 seasons, and then how they fared in 2018. I want a solid basis of talent first, a season that can tell me that this guy has the potential to be legit, which is the point of the 2016 season data. Then I want a poor 2017 season. Since what is “poor” is debatable, I’ll say that if a reliever had either an ERA, FIP, or xFIP that was .75 points higher than the year prior, I will consider that a “poor” season in context with their body of work. I will leave a little wiggle room if the numbers are close enough to give me a bigger sample, but generally, I will stick to the .75 rule. Then I want to see how they performed in 2018 so that I can draw my conclusion and make my closing statement. I will be using their pitching “slash line” of ERA/FIP/xFIP. I also will not be including pitchers who did not pitch in 2017 or threw a fraction of innings that they usually do because there is not enough sample size to draw from.

These are the first 10 players who fit my criteria and we have some interesting information already! About 4 players rebounded in my mind; Dellin Betances, Aroldis Chapman, Seung Hwan Oh, and Jeurys Familia (5 if you include Herrera’s ERA but not his peripherals), while 6 players had 2018 seasons that continued their downward trend (or 5 if you think peripherals are for schmucks). Interestingly enough, these were 10 of the top 11 pitchers by fWAR in 2016 with Kenley Jansen being the only player in the top 11 who had a great 2016 and 2017. Let’s continue!

This list of players requires a bit more nuance than the previous one and I really need to stick to the peripherals to draw a solid conclusion. I see 5 players who bounced back in a significant way through their peripherals; Edwin Diaz, Will Harris, Brad Ziegler, Zach Duke, and Hector Neris. But the interpretation of this data can vary depending on the reader, if you value ERA, then you might see different bounceback players than me, but I believe that the 5 players I mentioned here had a noticeably better 2018 season compared to their 2017 season. Alright, let’s continue!

I’m just going to throw Buchter out of here because he doesn’t seem to really care about what his FIP or xFIP says he should be doing. After that, I see Shawn Kelley, Alex Colome, Sam Dyson, Hansel Robles, Alex Wilson, Ryan Pressly, Tony Barnette, and Jeremy Jeffress as guys who bounced back in significant ways in 2018. That’s 8! The only guy here who didn’t really bounce back was Justin Wilson.

These 30 pitchers were among the top 50 relievers from 2016 and 17 of them rebounded after a poor 2017 season in my eyes, good for a 57% success rate. How can this apply to the Twins? Well, I see this as good news for Trevor Hildenberger, who is looking to have a 2019 season that is more in line with his 2017 campaign, but I see this as bad news for Addison Reed who saw his numbers decrease heavily thanks to another drop in velocity. I also see this as good news for Cody Allen, the player who inspired this article, as there’s a better than average chance that he can have a better year than his awful 2018 campaign. This could also be somewhat neutral news for Taylor Rogers, who had the most dominant season of his career so far but needs another one to really cement himself as a top relief arm. An interesting thing to note is that a fair amount of these players who had bounceback years did so on a different team, so maybe there is some credibility to the idea that sometimes a player just needs a change of scenery.