Pittsburgh Pirates: Setting Mitch Keller’s Expectations for 2020


Pittsburgh Pirates top prospect Mitch Keller will be entering the season with high expectations. Can he meet those big expectations?  Well, signs point to yes.

Mitch Keller is one of the most unlucky pitchers in baseball’s recent history. Keller is the Pittsburgh Pirates top pitching prospect. According to FanGraphs, Keller ranks as the 34th best prospect in baseball.

In 2019, Keller, in just 48 innings, allowed a batting average on balls in play of .475. No pitcher in the history of the game has ever pitched at least 40 innings and allowed a higher BAbip. That combined with the low amount of innings of work, you should take his surface numbers with a grain of salt.

Keller was outstanding at Triple-A last season. In 103.2 innings, Keller had a 3.56 ERA, 3.60 FIP and 1.24 WHIP. All of those rank among the top-5 at all levels of Triple-A, including his FIP being the lowest (min. 100 innings pitched). He also was second in HR/9 (0.8), and first in strikeout rate (28.2%). It’s not like her over performed either. His xFIP was 3.87, the only sub-4 mark at all Triple-A levels in 2019.

At the MLB level, the young right-hander may have had a 7.13 ERA, but you need to look past that. Keller struck out 65 of the 227 batters he faced, giving him a 28.6% rate. He also walked just 16 men, resulting in an outstanding 7% walk rate. Keller also gave up home runs at a solid 1.1 per 9 rate, or 0.3 better than the league average. Nationals’ lefty ace Patrick Corbin had almost identical peripherals with a 28.5% strikeout rate, 8.5% walk rate, and 1.1 HR/9 just for comparison. This all resulted in a 3.19 FIP for the 24-year-old. But that isn’t the only ERA estimator that said he was extremely unlucky. Keller also had a 3.78 SIERA, 3.47 xFIP and 4.18 DRA.

In terms of expected stats, you should see massive improvements. Keller had a .265 expected opponent batting average. In reality, that was .343, impart due to that .475 BAbip. There were also major discrepancies in his expected slugging to his actual opponent. A 135 point difference to be exact (.546 actual, .411 expected).

While it wasn’t great, his 4.35 xERA shows a run difference of nearly 3 (2.78). Another expected stat that Keller way under-performed was his weighted on base percentage. Despite it sitting at .392, he had an xWOBA of .314. Overall like with his peripherals, there was a major league pitcher with almost identical numbers. That was Mike Soroka. Last year, Soroka had a .264 xBA, .304 xWOBA, and .395 xSLG. He finished 2nd place in National League Rookie of the Year voting, and placed 6th in NL Cy Young voting.

Mitch Keller is above average in terms of exit velocity. Last year, he sat at the 69th percentile with an 87.2 MPH exit velo. That ties him with Dodgers’ ace Clayton Kershaw, but Kershaw had a great defensive team behind him in 2019. He was also allowed hard hit contact 35.9% of the time. That also sat above the league average.

Keller has outstanding movement on his slider and curveball. Both sit above average. His slider has a 60 grade because of it’s vertical movement (35.9 inches) and horizontal break (4.1 inches). However his curveball has some serious drop to it. With 56.9 inches of drop and 9.2 inches of break, it’s extremely comparable to Justin Verlander’s curve (57.4/7.6). But unlike Verlander, he averages over 80 MPH on the pitch (81.1 MPH).

While it’s not his best pitch, Keller’s fastball averages out at 95.4 MPH and was given a 55 grade per FanGraphs. He also throws a change up every now and then, and it has shown improvement throughout the years, but right now it sits around average to below average.

Just by going off of pitch usage, Keller was predictable. Keller was using his fastball at a 59.5% rate. Verlander is very comparable in terms of velocity to Keller. He used his fastball only 49% of the time. Another comparable pitcher in terms of movement and velocity according to Baseball Savant was Lance Lynn who used his fastball 54% of the time. So one of the easiest adjustments Keller could make to improve his game is mixing up his pitches more.

But one thing that is out of Mitch’s control is the team’s defense, which has massive potential. By the end of 2020, Ke’Bryan Hayes and Cole Tucker will likely be regulars who saw plenty of time at the MLB level. Tucker will also push a below average shortstop in Kevin Newman to second base, a position he excelled at in 2019. Both are some of the best defenders in the Pittsburgh Pirates system.

The Pittsburgh Pirates also made a major stride in center field. In 2019, former All-Star Starling Marte ranked last among all center fielders in DRS (-9), UZR (-7.6), UZR/150 (-8.3), range runs above average (-9.5). In all of them except DRS, the guy above Marte outdid him by nearly double. Replacing him will be Jarrod Dyson. Dyson has never once posted a negative mark in DRS, UZR, UZR/150, and range runs above average. Last year, he ha +14 DRS, 1.3 range runs, 7.9 UZR and a 12.3 UZR/150. He also had 6 outs above average whereas Marte had just 2.

Plus, with the likely possibility of the designated hitter being part of a shortened 2020 season, both Josh Bell and Colin Moran, two guys who combined for -34 defensive runs saved, will see most of their time at the DH spot. That opens the door for more playing time at first for Jose Osuna and possibly Will Craig, both of which grade out as way better defenders than Moran or Bell.

Next. A Day in the Life of Aaron Shackelford. dark

In all honesty, it would not be a surprise to see the Pittsburgh Pirates top prospect not only won NL Rookie of the Year, but also got some Cy Young votes. In terms of peripherals, he was almost identical to Corbin. Based on expected stats like xWOBA, xBA, xSLG, you’re looking at 2nd place Rookie of the Year finisher Soroka. In terms of pitch velocity and movement, you could confuse him with Verlander. ERA estimators make him similar to Zack Greinke (3.74 xFIP, 3.96 SIERA) or Hyun-Jin Ryu (3.32 xFIP, 3.77 SIERA). In total, Keller should easily post an ERA in the low-3’s with similar ERA estimators to 2019.