Has Emmet Sheehan done enough to take on big postseason role for Dodgers?

He's looked pretty good, but the blowup outings are a concern.
San Francisco Giants v Los Angeles Dodgers
San Francisco Giants v Los Angeles Dodgers / Jayne Kamin-Oncea/GettyImages
facebooktwitterreddit

Los Angeles Dodgers rookie Emmet Sheehan shut down the San Francisco Giants this season on Thursday night when he held the division rivals hitless over 4.2 innings of work. The last time against them? He held them hitless through six frames.

But even with the Dodgers' desperate need for pitching and Andrew Friedman revealing the team will have an unconventional pitching plan for the postseason, has the right-hander done enough to warrant a key role on the playoff roster?

He'll more than likely join the team for October after his largely promising audition once rosters expanded, but it's unclear how involved he'll be. When he debuted in June, he was great, tossing back-to-back quality starts in his first two career outings against the Giants and Astros. Then he got the win against the Rockies with five innings of three-run ball.

Then July arrived, and perhaps there was enough tape on him for opposing lineups to figure him out. He was bludgeoned by the Pirates, Orioles and Rangers in consecutive outings, allowing 17 earned runs on 17 hits and 12 walks over 12.1 innings. His next nine innings against the Reds and A's went well (he got a four-inning save against Oakland), but he was sent back down to Triple-A until Sept. 2.

Since then, he's lowered his ERA from 5.63 to 4.92 and his FIP from 5.14 to 4.82 while striking out an impressive 24 batters in 16 innings against the Braves, Nationals, Mariners and Giants (again), but his 4.3 BB/9 and max exit velocity (121.2 MPH) should be concerns for manager Dave Roberts, even after his top-notch outing against the Rockies on Wednesday (6 IP, 2 ER, 0 BB, 10 K).

Sheehan no doubt has a bright future ahead of him, but it still seems as if he's getting his bearings at the next level while working out some notable kinks (chase rate, barrel percentage, ground ball percentage and overall location need to see considerable improvements before we make any sweeping judgements).

Has Emmet Sheehan done enough to take on big postseason role for Dodgers?

Dodgers fans love their up-and-coming rookies, but perhaps starting with Bobby Miller and James Outman on the hype train with some Ryan Pepiot sprinkled in might be the best way to approach this. Sheehan appears to be in between wherever Pepiot and Michael Grove/Gavin Stone/Miguel Vargas/Michael Busch stand.

Don't forget, Sheehan initially got the call straight up from Double-A this year, and when he went back down to Triple-A following his MLB debut, he got knocked around for six earned runs on four hits and six walks in 9.2 innings. Though his 1.03 WHIP and 14 strikeouts were positives, he again struggled with walks and hard contact.

Roberts and the Dodgers are already factoring him into the postseason picture.

"Before Thursday’s game, Roberts said the team sees Sheehan as a reliever in the postseason with the ability to give them length, likely following an opener or pitching the middle innings of a game. It’s a bucket of players that also includes the likes of Ryan Yarbrough, Michael Grove, Ryan Pepiot and Gavin Stone."

Juan Toribio, MLB.com

While Sheehan might have the best pure stuff of any Dodgers prospect right now, he's still probably behind Pepiot and Ryan Yarbrough if we're talking about trustworthy long relief options (though Pepiot might deserve a start in a seven-game series).

Either way, the jury's very much still out on Sheehan, who could be a nice lightning rod in the playoffs if utilized correctly when his stuff is "on." He definitely shouldn't be, at this stage in his career, relied upon to hold a tight lead or give a spot start in the playoffs, but he should see low-leverage innings so he can get the feel and is better for it come 2024.

THe Dodgers are expected to throw enough young arms into the fire once the playoffs begin. Sheehan's 13 games probably aren't enough of a sample size to do that just yet.

manual