What should the Twins do with Alex Colomé next season?

Perhaps no Twins player bore the brunt of the blame for this past season going off the rails more than reliever Alex Colomé. His dreadful April included three blown saves and three losses and contributed significantly to the Twins ending the season’s first month with a 9-15 record.

How and when those poor performances happened made April even more painful. Minnesota went 0-5 in extra inning games and 2-6 in one run games in the first month and several of those close losses came in games that appeared headed for Twins wins before improbable late collapses.

Each of Colomé’s early season blown saves came in games the Twins had at least 95% chances to win at one point or another, according to FanGraphs’ win expectancy charts. The first was on opening day in Milwaukee, when Colomé and the defense behind him frittered away a three run 9th inning lead. Then came a Sunday afternoon series finale against Seattle where the Twins had a 6-0 lead in the 6th inning. The third was the bonkers midweek series finale in Oakland coming out of the COVID pause, when leads were blown in both the 9th and 10th innings of a 13-12 loss.

Altogether, Colomé worked in nine games in April, covering 8.2 innings. For the month, he delivered an 8.31 ERA, 1.97 WHIP, and allowed 20 of the 47 batters (42.5%) he faced to reach base safely via hit, walk, or hit by pitch. Colomé’s minus-2.17 win probability added (WPA) was the worst mark of any pitcher for any team in their first nine appearances, according to Aaron Gleeman.

Colomé had signed with Minnesota on a one year, $5 million deal late in the offseason and his contract also included a mutual option for next season valued at $5.5 million that can be bought out for $1.25 million.

After his miserable April, it seemed like a pretty safe bet the Twins’ side of that option would be declined.

But Manager Rocco Baldelli gave Colomé opportunities to work through his struggles in lower leverage spots in May and he mostly rounded back into his previous form the rest of the season. From May 1 onward, Colomé logged 56.1 innings out of the Twins ‘pen with a 3.51 ERA, 1.31 WHIP, and +1.10 WPA. Of the 243 batters he faced in that span, a more palatable 30.9% reached base via hit, walk, or hit by pitch. Once Taylor Rogers went down with an injury that ended his season in July, Colomé even returned to the primary closer role and locked down 15 of 19 save opportunities.

Now, the Twins’ decision about whether to pick up Colomé’s option for 2022 might be a little more complicated than it looked early on.

When Colomé signed with the Twins many viewed it as a shrewd value deal made by the front office. Colomé had been one of the most reliable and effective high leverage relievers in the game over the prior five seasons. If we toss out the shortened 2020 season for small sample size noise concerns, Colomé had a combined 2.78 ERA and averaged +1.76 WPA per year in the four full seasons from 2016-2019. His cumulative +7.04 WPA in that span was the eleventh-most of 257 qualified relievers.

This past season, once he worked his way through his terrible April start, he was, in a lot of ways, the same as he ever was.

Below is a chart plotting Colomé’s run prevention stats from May 1 through the end of the season against the same stats for each full season 2016-2019. I use ERA and FIP a lot in my writing and here I’m also adding two of the leading ERA estimators — xFIP and SIERA — because they are more predictive of future performance than ERA and FIP.

Data from FanGraphs

Here, I’ve highlighted the 2021 data (not including April) in red and used shades of gray for prior seasons so we can easily see how the good part of Colomé’s 2021 looks in comparison to his other full seasons. You can see his ERA after April this past season was still the highest of his career, but in the other three categories his 2021 stats show up in the middle of the dataset. In each of those cases his 2021 numbers are actually improvements on the numbers he had in 2019.

Let’s go a little deeper and do the same for some key process stats. Below is a similar chart, but this time with the core FIP inputs — strikeouts (K%), walks (BB%), home runs allowed (HR%) — and three other important stats: the difference between his strikeout and walk rates (K-BB%), which some studies have shown is even more predictive for future performance than the ERA estimators in the chart above; called strike plus whiff rate (CSW%); and hard hit rate.

Collectively, this group of defense independent stats gives us a good view into the aspects of pitching that the pitcher can generally control (or at least has a good amount of influence over).

Data from FanGraphs and Baseball Savant

Here, the picture is maybe not quite as rosy as it was with the ERA estimators. Colomé’s strikeout rate this past season was the lowest of the five seasons in the data. That, naturally, made his K-BB% also come down to its lowest level in the sample. Also on the bad side of the ledger was that his hard hit rate this season was the highest in the sample. But, his marks in those categories were not wildly different from before — they were worse, but not dramatically so. And his rates of home runs, walks, and called strikes plus whiffs were all in line with prior years.

Beyond the statistics, Statcast lets us look at the characteristics of Colomé’s pitches and determine if anything meaningful has changed with his stuff. A review of those metrics reveals that he did not change his pitch usage from the prior couple of seasons and again worked about 70% with his cutter and 30% with his four seamer. The velocity of the cutter was in line with the past (right around 90 mph), while the four seamer was a tick beneath 94 mph instead of a tick above like it had been previously. There were no changes to the spin rates on either pitch. The vertical movement of the pitches was almost identical to measurements from 2018 and 2019 and, if anything, Colomé had a bit more horizontal break on his cutter this year than his earlier successful years. For the most part, the stuff he was throwing this season was also in line with the past.

As a quick aside, one cannot help but notice on his Statcast page that his four seam fastball has always been much more hittable than his cutter. For his career, opposing hitters have hit .298 and slugged .463 against his four seamer and just .196 and .292, respectively, against his cutter. Now, with the four seamer perhaps losing some speed, that gap was even wider and batters teed off on his four seamer last season (.371 BA, .532 SLG). I wonder if Colomé would benefit from adding or switching to a two seam fastball. That’s a trick that other cutter heavy relievers — Mariano Rivera and Kenley Jansen most notably — have pulled later in their careers to great success.

OK, back to the point. For me, all this data presents signs that Colomé got things figured out after that terrible first month and pitched the rest of the season much like the pitcher the Twins were expecting when he signed. He was more or less the same pitcher for most of 2021 as he was earlier in his career.

The Twins’ decision about whether to pick up their side of his option is made more complicated not only because he pitched better down the stretch of the season, but also because they are in dire need of pitching on next year’s roster.

Outside of Tyler Duffey and Caleb Thielbar the remainder of the pitching staff has a lot of question marks. Randy Dobnak and Taylor Rogers will need to return to health. Kenta Maeda is likely out for the season after elbow surgery. Jorge Alcala, Bailey Ober, and Joe Ryan should continue in their current roles, but they are still largely unproven over long stretches. John Gant likely figures to factor in too, but in what role? Juan Minaya and Danny Coulombe had decent seasons and are still on the roster (for now), but I’m not sure anyone wants to have to count on them to fill important roles. Familiar names Lewis Thorpe, Devin Smeltzer, and Cody Stashak remain under contract, but also have injury, role, and performance questions to settle.

Given all that, the Twins could decide that Colomé’s second half was more representative of what he will be like going forward and decide to try retain him for another season. At $5.5 million the option year is not exactly cheap for a reliever, but it’s not terribly expensive either. If you factor in the cost of the buyout ($1.25 million) and the cost of replacing him with another free agent reliever (perhaps in the $3 million range, which might be on the lower end), letting him go would not clear much financial room.

So, despite the narrative around Colomé’s time in Minnesota, the decision on what to do with him in 2022 is not clear cut (at least to me). I’m pretty sure I know how this will go given the tone of our comments and posts regarding Colomé this season, but I’ll ask anyway. What would you do?

I think that I probably lean toward option B right now. That might be a little bit of a contrarian and unpopular point of view, but reliever performance can be volatile season to season and there are quite a few pitchers playing important roles on the playoff teams now that have washed out of other teams’ bullpens previously. With the data above, I think I’ll make the bet on Colomé’s larger body of work going forward. That said, I would not blame anyone for moving on and I would not be surprised in the least if the Twins decided to do just that.

John is a staff writer for Twinkie Town with an emphasis on analysis. He is a lifelong Twins fan and former college pitcher. You can follow him on Twitter @JohnFoley_21.