Carlos Carrasco was having a really nice year until it rained in Atlanta during his start on August 15. There was a 55-minute rain delay, after which Carrasco came back out to pitch the second inning. That’s all he pitched, as he injured his oblique, which necessitated a trip to the IL. Coming into the game, Carrasco was 13-4 with a 3.76 ERA. In his previous seven games, he was 5-0 with a 1.43 ERA. Counting the Atlanta game and his final six starts, Carrasco was 2-3 with a 4.88 ERA.
A contributing factor to Carrasco’s hot stretch before being injured was playing against a bunch of second-division clubs. But he also faced the Padres and Braves in that span and he allowed just 3 ER in 11 IP against those two playoff-bound squads. And his bad closing stretch had two games against both Miami and Washington, along with one against Pittsburgh – teams that combined to lose 300 games last year.
That’s the narrative account of Carrasco’s 2022. The objective case is he finished the year with 29 starts, 152 IP and a 3.97 ERA. Here’s how the computer models forecast Carrasco in his age-36 season:
ATC – 145 IP, 4.04 ERA, 1.28 WHIP, 140 Ks, 19 HR
Marcel – 142 IP, 4.12 ERA, 1.324 WHIP, 137 Ks, 18 HR
RotoCh – 149 IP, 3.81 ERA, 1.33 WHIP, 131 Ks v
Steamer – 150 IP, 3.95 ERA, 1.26 WHIP, 140 Ks, 19 HR
THE BAT – 147 IP, 4.50 ERA, 1.32 WHIP, 138 Ks, 21 HR
ZiPS – 107.2 IP, 4.35 ERA, 1.33 WHIP, 104 K, 17 HR
Things were going nicely for the oft-used line of how the models were in agreement over how a veteran would produce. Until we came to THE BAT, which had an ERA out of step with the others. And ZiPS closed it out with another ERA higher than previous consensus, along with by far the fewest IP of the group. If it’s any consolation, THE BAT had Carrasco with a 5.02 ERA projection for the 2022 season.
While he just missed last year, the fact is that Carrasco hasn’t made 30 starts in a season since 2018. It makes sense that the models have him with the innings totals that they do. If anything, perhaps it’s a surprise that they have him accumulating essentially the innings he did in 2022. ZiPS might be ultra-bearish on IP but is that really more surprising that the others forecasting a repeat of last year’s 29-start output?
It’s easy to see how Carrasco might be a tad difficult to project. Along with the shortened-season in 2020, we have the fight back from leukemia, along with him missing half of the 2021 season with a hamstring injury. And when he did pitch in ’21, he was frequently throwing beach balls to the plate, as he allowed 12 HR in 53.2 IP.
After posting a 2.01 HR/9 in ’21, Carrasco put up a 1.01 mark last year, which was much more in line with his career rate of 1.08 HR/9. Both THE BAT and ZiPS see a jump in HR/9 this year, coming in with marks of 1.31 and 1.42, respectively.
One of the basic principles for the computer models is to see how the player has performed recently. But when two of the past three seasons have big narrative warnings – and the third one has a mini warning, too – how much faith do we put in the models? And to make matters worse, we have little idea how much of the forecasts are depending on less than stellar previous numbers versus the regression built in for a 36 year old.
My totally biased forecast for Carrasco in 2023:
126 IP, 3.64 ERA, 1.298 WHIP, 120 Ks, 15 HR