Postseason Managerial Report Card: Dave Roberts

The Dodgers were supposed to be a team for all seasons. A lineup comprised exclusively of All-Stars? Check. A pitching staff with Clayton Kershaw as a fourth starter? Check. A bullpen with two closers, plus a cornucopia of useful relievers to fill in the cracks? Check as well, particularly with Kenley Jansen turning back the clock.

Injuries had something to say about that, of course. By season’s end, the team lost Max Muncy to a fluke injury and Kershaw to elbow pain that shut him down for the postseason. A lineup with only seven phenomenal hitters? A rotation that stopped at three Cy Young candidates, and had to make do with the perfectly cromulent Tony Gonsolin as a fourth starter? The horror! At least the bullpen was still intact — Jansen and Blake Treinen anchored a top-five unit that kept the Dodgers rolling throughout the season even as their starters dealt with injuries.

All of these factors come into play when evaluating the decisions Dave Roberts made in the playoffs. I’ll stick to the general format of these posts — first hitting, then pitching — but in my opinion, the cascading set of questionable decisions on both sides started with the two injuries that changed the Dodgers’ plans from plug-and-play to figure-it-out.

Lineups/Pinch Hitting

Grade: C
It’s hard to goof up the top of the Los Angeles order. Mookie Betts, Corey Seager, Trea Turner, Justin Turner, Will Smith — you can jam those five in without thinking. That’s what Roberts did in the winner-take-all clash with the Cardinals. He supplemented them with AJ Pollock, Matt Beaty, and Cody Bellinger. Beaty was a contingency plan, a career 104 wRC+ hitter who is a) left-handed and b) capable of playing first base. He was in the lineup out of necessity. Not in the lineup: Chris Taylor, who struggled down the stretch but is still surely a better hitter.

The problem: the Dodgers couldn’t come up with a clear way to put their best position players on the field. The lefty bat they wanted was Gavin Lux, not Beaty, but Lux was blocked by their sterling middle infielders. With only seven professional games in center under his belt, he was clearly a worse option than Bellinger out there. He also hadn’t played first base, another possible landing spot. Meanwhile, Bellinger is enough better than Taylor in center that Roberts chose not to make the other obvious change, sliding Bellinger to first and using Taylor in center over Beaty.

It was clear that this configuration wouldn’t last. Beaty went 0-for-3 before being replaced by Billy McKinney, and the Dodgers used a wave of pinch hitters to scrap through a 3-1 win keyed by Taylor’s pinch-hit home run. Chris Taylor: pretty good at baseball.

In Game 1 of the NLDS, Roberts ran the same lineup back, with some batting order shuffling at the bottom. Beaty went 0-for-3, the Dodgers got shut out, and that was it! Forget what I said about Taylor’s defense; he was in center for the next game, with Bellinger sliding to first base. Taylor mashed — 2-for-4 with a walk and a double — in a Dodgers romp, and it was time for a new plan.

The lineup against lefties was easier — Albert Pujols at first, Taylor and Pollock in the outfield — but the next time the Dodgers faced a righty starter, Roberts tried yet another permutation. This time, he benched Pollock, moved Taylor to left, and put Lux in center, seven games of experience and all. Bellinger, the best defensive center fielder on the team, stood at first base. It was counterintuitive — but potentially the best alignment for a team with strange fit issues.

When Justin Turner was felled by a neck injury, Taylor’s versatility saved the day. He slid to third, Pollock re-entered in left, and the lineup otherwise remained the same until the last game of the season, when Roberts went back to Beaty over Lux. The Dodgers lineup was very resilient most places — but their best defensive center fielder was also their best lefty first base option with Muncy down, and the team’s fumbling to put out a cohesive lineup was a direct consequence.

I don’t give Roberts full blame for that, but his pinch hitting decisions also left something to be desired. The team was very short on depth, particularly after Turner got hurt. There are only so many ways you can flatter your bench of McKinney, Beaty, Pujols, Steven Souza Jr., and Austin Barnes. Roberts did them no favors, though, by picking the wrong spots for his best hitters.

In a close game against the Cardinals, Roberts used Luke Raley as the first hitter off the bench. Lux, meanwhile, was announced and then replaced without taking a cut. In a 1-0 loss to the Giants, the team started Pujols, pinch ran for him with McKinney, then pinch hit for McKinney with Souza. Given that Roberts didn’t seem to trust any of his righty pinch-hitting options, I would have kept Pujols in there. I also would have carried a third catcher so that Barnes could pinch hit, as Roberts clearly wanted him in the game over the rest of his bench.

Some of this comes down to personnel. When Pollock started on the bench, Roberts deployed him efficiently, in the first high-leverage spot he could find. But nothing excuses the last two decisions Roberts made. First, he let Brusdar Graterol bat for himself in a 6-2 game against the Braves with a runner on base. That doesn’t sound weird at first blush, and the Dodgers won comfortably — but he let a reliever bat for himself, then pulled that same reliever before the start of the next inning. That’s right: Graterol’s last act in the game was to strike out swinging.

That’s awful! Even if you think that the Dodgers wanted to save pinch hitters — and sure, whatever, I don’t think that’s a good use of resources but reasonable minds can disagree — they had a bench full of starting pitchers who would have been better options than Graterol. Leaving Graterol in would also have been fine — he had only thrown 14 pitches. Either way, the one thing that did happen was the thing that made the least sense.

Finally, Lux completely disappeared for the last two games after two defensive blunders in center field. He may have been banged up, or otherwise unavailable for reasons that the Dodgers didn’t broadcast, but it sure seemed like he just lost favor and vanished. That’s fine — playing a poor center fielder to get a career .233/.314/.368 bat into your lineup is a weird decision in the first place — but the frequent outfield reshuffling feels desperate to me. I can believe that either option — maxing out on offense with Lux or using Bellinger and a random first baseman to shore up the defense — is correct, but changing back and forth seems harder on the players, and certainly at least a little wrong.

Pitching

Grade: D
It’s easy to be too certain that Roberts’ pitching decisions doomed the defending champions. Mike Petriello recently observed little measurable effect when he searched past playoffs for starters in relief. Rather than reflexively give Roberts an F because Scherzer ended up with a dead arm (a defensible position, to be sure), I thought I’d walk through the games sequentially and see how each move painted him into a corner for successive games.

Against the Cardinals, Roberts was direct and efficient. He got what he could out of Scherzer, then went to an array of dominant relievers. Joe Kelly, Graterol, Treinen, Corey Knebel, and Jansen entered in order. They held the Cardinals to two hits and a walk in 4.2 innings of work. With a bullpen like that, there’s no need to get cute.

Against the Giants, Roberts got cute. After two games where there wasn’t much managing to do, Game 3 was his first test. There, he rode Scherzer for 110 pitches — including skipping a juicy pinch-hitting spot. From there, he opened up the good stuff — Treinen and Jansen combined to pitch the last two frames in a 1-0 loss.

The next day, the Dodgers won comfortably — they were up 5-1 after five and won 7-2. Roberts, though, had turned to Walker Buehler on short rest. That decision could go either way — I would have preferred Gonsolin, but only by a hair. It put a large innings burden on the bullpen, however, and despite a big lead, Roberts didn’t screw around. He went to Graterol, Alex Vesia (excellent this year), and Treinen before the Dodgers really put the game out of reach in the eighth. Finally, he put Phil Bickford in, but the plan was clear — work the big bullpen horses as much as possible.

In Game 5, Roberts shortened the list of trusted relievers even further. Knebel opened, a gambit that paid off despite loud contact. Graterol entered next — a strange decision given that Mike Yastrzemski, a lefty who the Giants like to hide against southpaws, was due up. This was Julio Urías’s game, and it seemed like an okay spot for a lefty, though Graterol also escaped trouble. Then came Urías, but only for 59 pitches and four innings. If Roberts only wanted to use the best of the bullpen, he maybe should have tried to get more than 15 batters faced out of his Cy Young candidate.

That “best of the bullpen” plan didn’t quite work out. Treinen got an inning, and Jansen did as well, but that left one inning remaining. Enter Scherzer, on two days rest after his 110-pitch effort in Los Angeles. He closed out the ninth in an economical 13 pitches, but just…. What was with the double openers? Why use Urías so sparingly? Why not extend Jansen an extra inning given everything that took place beforehand? There were plenty of ways to manage this game. Roberts chose the one — only using his top four relievers, not extending any of them for more than an inning, and keeping an extremely short leash on Urías — that required a bullpen appearance from a 37-year-old pitcher coming off of his highest pitch count of the year.

With his top three starters all burned, Roberts still didn’t want Gonsolin in. He went with a full bullpen game to start the NLCS, with Knebel making his second consecutive start (fun trivia there). He emptied the vaults, and the low-leverage pitchers mostly performed well. Bickford, Gonsolin, and Justin Bruihl combined for four innings, six strikeouts, and one run allowed. Vesia and Kelly each chipped in scoreless innings. Treinen, pitching for the fourth time in six days, gave up a walk-off single to Austin Riley, but the bullpen pieces Roberts was so reticent to use performed just like you’d expect based on their season-long numbers: perfectly well.

Because he only used Gonsolin for five outs, Roberts had completely emptied out the bullpen. That meant Scherzer was pitching on two days’ rest after a high-leverage relief appearance, which was itself on two days’ rest after his longest outing of the year. He was excellent when in, but had a dead arm, and left after only 4.1 innings. That meant Roberts had to turn back to those same bullpen arms for the second day in a row — something he clearly didn’t want to do.

To avoid having to run out everyone again, Roberts got creative. First, he used his two least-disfavored relievers outside of the top group: Vesia and Kelly got the fifth and sixth innings. Then, he went to Treinen again for the seventh. With the Braves sending two lefties to the plate in the eighth, he skipped Bruihl, and also skipped the three trusted righties in his ‘pen: Knebel, Graterol, and Jansen. Instead, he called for Urías, who like Scherzer was on two days rest. Urías scuffled with location and gave up two runs, which could have happened to anyone, but the cost is higher when you bring in a starter to do it.

After an off day, Roberts was forced to go back to the bullpen arms he didn’t trust. Buehler scuffled, which meant Vesia entered in the fourth inning to retire Freddie Freeman. Knebel came in next, but only lasted three batters — single, walk, long fly out. That gave Roberts no choice but to use the Bickfords and Gonsolins of the world — in this case actually Phil Bickford, who gave up a single but then got a clutch double play to end the inning.

From there, the theme recurred: the relievers that so terrified Roberts did just fine. Bruihl mostly did his job against lefties. Kelly backed him to clean up the inning. Evan Phillips, added to the roster for the NLCS, got five somewhat adventurous outs before Gonsolin slammed the door behind him. When the Dodgers roared into the lead with a four-run eighth, the mop-up brigade had done their job again, and Jansen entered for an effortless ninth.

Urías was far from sharp the next day, giving up five earned runs, eight hits, and two walks in only five innings pitched. His stuff didn’t look the same. Perhaps that’s because he had doubled his previous career-high workload, but if that’s the case, maybe he didn’t need to put extra miles on his arm with that high-leverage relief appearance a few days prior. The whole thing felt like present Roberts trying to make up for the excesses of past Roberts — while continuing his own excesses, of course. Roberts emptied the bottom of the bullpen behind Urías, and had Gonsolin soak up the last two innings.

Naturally, the Dodgers had planned a bullpen day the next day, and were missing Gonsolin’s bulk for it. Phillips had another scoreless outing, entering in the first after Kelly was knocked out with an injury that sidelined him for the rest of the playoffs. But even in an absolute laugher of a game — 11-2, with the Dodgers up 6-2 after five — the four core relievers had to throw. Graterol and Treinen handled two innings each, Knebel and Jansen kicked in one apiece.

By the time Game 6 rolled around, things had fallen fully apart. Scherzer’s arm didn’t recover as planned — he might seem invincible, but the workload had been too much for him. Buehler had to make the start on short rest, his second short-rest start of the playoffs. He simply didn’t pitch well enough to win — four innings pitched, three walks, and seven hits, one of which was a homer. That was four runs for Atlanta, which was all they’d need. The high-leverage bullpen brigade poured it on — Graterol, Treinen, Knebel, and Jansen combined for five innings of two-hit, seven-strikeout relief to hold the Braves scoreless — but it just didn’t matter.

There’s a cascading problem here. Roberts didn’t trust huge swaths of his bullpen (and rotation if you count Gonsolin, who never did make a start in the playoffs). One way to compensate for that would be by extending your starters — all of Buehler, Urías, and Scherzer are bona fide Cy Young candidates this year. If you’re going to do that, though, you can’t really do it on short rest — short rest means shorter outings, which means you’ll need to trust the bullpen. There’s just no way to use your ace for only 15 batters, trust only four bullpen arms, and make it through a game.

Over and over again, Roberts leaned on two default ideas. First, he only trusted three starters and four relievers, and wanted those seven players to pitch as many innings as possible. Second, he was willing to use those starters on a schedule other than once every five days, which meant shorter starts when they did pitch, as well as stamina-sapping throw day relief appearances.

These two ideas are at odds! If you want your starters to mostly go deep into the game to minimize bullpen use, you can’t use them on short rest. Even if you think that using starters in relief doesn’t blunt their effectiveness, it certainly limits their ability to go deep into games, and it’s strange that Roberts would accept that tradeoff given his strong desire to have only the very best relievers throw. Much as you might like to, you can’t make the entire airplane out of the black box. Every time Roberts tried to, it only made things worse in the future, which put him in worse binds. He tried to get out of those worse binds by doing more of the same. It’s a vicious cycle, one that ended with the Dodgers home for the playoffs.

Maybe this model would be rational if the “bad” parts of the Los Angeles bullpen were truly awful. They weren’t! The top four relievers were excellent — they had a collective 1.50 ERA in 30 innings of work, with 40 strikeouts and only four walks. The rest of the ‘pen? They had a 3.28 ERA over just under 25 innings, with 31 strikeouts and seven walks. Exclude Gonsolin, who was never going to be a high-leverage relief arm, and the other fill-ins had a 1.74 ERA and 27:6 K/BB ratio.

The Dodgers didn’t hit well enough to make the World Series this year. Their offense, which was productive throughout the season, sputtered at the worst possible time. It’s hard to overcome an offense that scored two runs or fewer in half of their postseason contests. But by contorting his roster to feature only the pitchers he trusted most, Roberts put those pitchers — particularly the starters, who he used aggressively on short rest and out of the bullpen — in tough situations that could have been avoided with different planning.

To me, this is more of a philosophical failure than a tactical one. Many of the tactical decisions that Roberts made were no-brainers, forced on him by decisions made earlier in the playoffs. How would this team have done if the players were used normally? We’ll never know. But Buehler and Urías were asked to do more than they’d ever done, after shouldering career-high workloads during the first season since the shortened 2020 campaign, and both faltered. Scherzer was asked to carry a huge burden — a season-high pitch count in the playoffs, then high-leverage relief, then another big start — and his arm suffered for it. All three workhorse starters wore down, in one way or another.

Could all of them have overcome those difficulties and put together a dominant October performance? Sure. But I strongly disagree with the concept of ignoring the Dodgers’ only plus source of depth this year: their impressive bullpen. This was a team that was well-suited to throw each of its best starters on full rest, use a short-leash Gonsolin with a fully operational bullpen behind him, and challenge other teams to deal with it. Instead, it became a needlessly complicated juggling act. Maybe that was planned — but it still earns my scorn.