This is a series of evaluations that will be done this offseason on every player that closed the season on the 40-man roster for the Minnesota Twins throughout the winter until each player has been evaluated. The plan is to start with Mr. Belisle and move all the way through the pitchers, then to the catchers, infielders, outfielders and finally those listed as designated hitters on the club’s official MLB.com roster. That means we’ll wrap it up with Kennys Vargas sometime before the season starts.
- Name: Jose Berrios
- 2017 Role: Second-best starter on the Minnesota Twins.
- Expected 2018 Role: Second-best starter on the Minnesota Twins.
- MLB Stats: 3.89 ERA, 3.84 FIP in 145.2 innings, 8.6 K/9, 3.0 BB/9, 1.23 WHIP, plus-2.8 fWAR.
- MiLB Stats: 1.13 ERA, 2.63 FIP in 39.2 innings at Triple-A Rochester, 8.9 K/9, 1.8 BB/9, 0.81 WHIP.
- Contract Status: Free agent after 2022, arbitration-eligible after 2019.
2017 Lowdown:
It was pretty easy to separate the wheat from the chaff as far as baseball minds were concerned as it pertained to Berrios. Though not many had written him off entirely, it was still true that a fair chunk of Twins fans expected absolutely nothing out of him in 2017 and some had downgraded his future in their personal rankings. It’s a nebulous concept, but it wasn’t hard to find if you looked for it.
Flash forward a year, and it’s hard not to be enthused about a guy who’ll start next season at the tender age of 23. Despite having a career ERA of 5.07 — the same ERA Kyle Gibson has posted each of the last two seasons — it was a tale of night and day from last year to this for the uber-talented righty.
It can’t be understated, but we’ll try anyway: Berrios halved his ERA from a season ago.
Yes. Halved.
Everything that could go wrong for Berrios in 2016 did. He posted an ERA of 8.02, allowed nearly two baserunners per inning and overall permitted a slash line of .310/.409/.523. In other words, he basically made opposing batters look like Nolan Arenado (.932 OPS) in 2016. In 2017, that line dropped to a meager .239/.313/.380 — or roughly the equivalent of Phillies banjo-hitting middle infielder Freddy Galvis.
The common theory for Berrios’ improvement is his improved fastball command, though proving that exactly can be tricky. It’s not as simple as his walk rate being cut, but that’s one easy thing to point to as he walked just 3.0 batters per nine after walking nearly double that in 2016.
An improved zone percentage — from 38.3 percent to 44 percent — also helps, but isn’t necessarily the indicator it seems to be either. Throwing pitches in the strike zone is a two-way street; tickle the corners of the zone and you’re golden, live in the zone and you’ll die by the zone. Pitchers like Berrios have the kind of repertoire that will induce swings on pitches out of the strike zone, so the improvement in this respect is nice, but certainly not a be-all, end-all.
Looking at contact rates can help a bit. His swinging-strike rate jumped 1.2 percent — 8.2 percent to 9.4 percent — but still lagged a bit behind the AL average of 9.9 percent. On the positive side, Berrios saw a 3.6 percent jump in swings at pitches out of the zone — probably a better indicator for a pitcher like Berrios — and a 6 percent decrease in contact rate on those pitches. When hitters are chasing, that means Berrios is in counts where he can work out of the strike zone, be it with purpose fastballs or breaking balls that mess with the eye levels of hitters.
It’s a good mix.
His first-pitch strike rate was also up just a tick under 4 percent, though still a bit behind (1.5 percent) the AL average of 60.6 percent.
In short, this all seems to suggest A. the line between good and bad work from a young pitcher is narrow and B. there’s still ample room for improvement, which shouldn’t be surprising.
The best spot to check for location improvement on Berrios’ fastball is in the PITCHf/x tables. Last year, Berrios allowed an obscene .339/.465/.543 line in his four-seam fastball. For most pitchers, their fastball numbers won’t be pretty even if they’re a solid hurler. For instance, Berrios took a huge step forward on that pitch in 2017, and still allowed opponents to hit .259/.354/.420 on it. That’s a 230-plus-point reduction on the OPS allowed on the pitch, yet it’s still at .775 OPS.
Does that make sense? A well-located fastball at 93.8 mph on average still don’t be a world-beating pitch, but it’s a massive jump from a poorly located one at 94.2 mph (his 2016 average velo on it).
One of the most marked improvements Berrios made was on his curve. The swinging-strike rate — typically the best indicator of dominance — was just 9.3 percent in 2016, but that mark jumped nearly 4 percent in 2017. That won’t make anyone forget about the best curves in the game, but it’s a steady pitch with room to grow. Opponents absolutely peppered it last season (.809 OPS), but were virtually helpless against it this year (.550).
That seems to be a trickle-down effect with improved fastball usage and command. The rest of a pitcher’s repertoire plays up when a pitcher can get his foundation down right — which is nothing too earth-shattering.
He also mixed in considerably more sinkers. After throwing just 161 in 2016, he fired 732 sinkers in 2017 and the results were pretty good. Opposing batters hit just .217/.324/.380, and it induced a groundball rate of 46.9 percent. Berrios has largely been a fly ball pitcher in his career — and to some extent, that’s OK with the outfield defense he has — but anytime he can burn some worms, that’ll help keep his home run rate in check, too.
He also keeps a changeup in his back pocket, but it’s an unremarkable offering at this point (.816 OPS against, 6.6 percent whiff rate). It did, however, see a 200-point decrease in OPS against on it.
The long and short of it is this: Berrios had a really nice season, and there’s quite a bit more left in here. He’s a very, very special pitcher.
Grade: A-. It wasn’t a perfect season, but what an incredible turnaround. It’s hard to believe he’ll only be 23 when next season starts.
Listen to Brandon on Midwest Swing
