Managers get put under the microscope in October, which may be why they are prone to such micro-managing. Deprived of the large sample size rationale that insulates them against failure during the regular season, the perception is that managers do way more to impact playoff games than they do regular season games. All four managers in yesterday’s NLDS games made crucial decisions that helped settle both games.
It’s very hard to find “analysis” that doesn’t simply boil down to “It worked. Good Call, Skip.” Let’s look at the big managerial calls yesterday and see if we can parse the rationale from the results.
Roberts vs. Kershaw
After watching and covering the Dodgers all season, I can tell you that like many managers, when Roberts goes out to the mound, the hook comes out with him. When Clayton Kershaw allowed the single to Espinosa and then the infield single to Trea Turner, Roberts went out to the mound. Based on the brevity of the conversation at the mound, it struck me at the time Roberts was hoping to go out there and hear that Kershaw wanted Harper, and had enough left to do so. He heard it, and Kershaw stayed.
Kershaw’s overall mastery of Harper, the left on left dynamic, and Kershaw’s status as the Dodgers’ “best option” in the words of Roberts, all pointed to leaving him in the game. Again, in a regular season contest, this likely wouldn’t have even been worth a mound trip. However, when you consider that Kershaw was coming back off a lengthy DL stint as well as the reality of a vastly improved Dodger bullpen in 2016, it shouldn’t have been a big surprise that many felt Roberts should not have even let Kershaw start the 7th.
The flaw, however, wasn’t the choice to let Kershaw remain in the game, but rather the choices that came after. First, Roberts brought Pedro Baez, a name synonymous with “gasoline” to many Dodger fans thanks to his struggles in previous postseasons, instead of Joe Blanton, who has literally remade himself into one of the best non-closing relievers in baseball. This year, Blanton posted the 23rd-best FIP in the NL for relievers, as well as the 13th best LOB% at 82%.
Baez, truth be told, had an 81% LOB rate for the year. It’s when you get into the specifics of the situation that Blanton emerges as the better choice. Blanton was better in high leverage situations (.549 OPS vs. .613 for Baez) in 2016. He was also significantly better than in that exact situation (1st and 2nd, two out). Yes, small sample size, but batters had an OPS of 1.881 against Baez (6-8) in that situation, compared to a .748 OPS (14 PA’s) for Blanton. Add the fact that Baez was far more susceptible to a homer (16% HR/FB vs. 7.4% for Blanton), and this was a highly questionable move from Roberts.
The choice became even more puzzling after Roberts eventually brought Blanton in anyway two batters later.
Dusty vs. Reality
Amongst the attributes of the 2016 Dodgers forever branded into our memories, the team’s struggles vs. left-handed pitching have been most prominently seared into our cerebral cortex. Given the improbable gift of a three-run rally to tie the game, Dusty Baker proceeded to immediately re-gift by ignoring what cannot be ignored at this point.
The Dodger win probability entering the 7th was 92%. Even after Kershaw departed two outs later with the bases loaded with Nationals, it was still 88%. For Washington to come back, even with the presence of Dodger-Killer Daniel Murphy, was an incredible feat. With a tie game and nine outs still left to secure, Baker then committed some managerial negligence that cost him the game.
In the bottom of the eighth, Baker allowed right hander Blake Treinen to face five consecutive left-handed batters. He nearly got away with it, but after a Yasmani Grandal ground-out, Andrew Toles, the recently exhumed Andre Ethier, and baseball sociopath Chase Utley strung singles together to give the Dodgers the winning margin at 6-5. Baker let Treinen pitch despite having Sammy Solis, who struck out nearly one third of lefties he faced on the year while limiting them to a .556 OPS, waiting in the bullpen.
Solis had thrown 34 pitches the previous game, but Baker only needed to get one batter. Like Roberts, Baker also undermined any defense of the move by bringing in Solis anyway, after the damage had been done.
Bruce Bochy vs. His Entire Bullpen
Bruce Bochy’s reputation as the Zen Master of Postseason Managing took a hit last night when his bullpen did what his bullpen did all year: combust like a Spinal Tap drummer. Celebrated Dugout Cheerleader Derek Law started the inning by allowing a Kris Bryant single, taking the Giants’ win expectancy from 97.95% to 95.43%. Law’s career vs. Righties: .598 OPS. Kris Bryant’s R/L splits: .896/1.060. Bryant had been 1-3 against Law in 2016. This choice seems fairly reasonable, especially when considering Law was there in all likelihood to get just the one batter.
After Bryant’s single, Bochy yanked Law to let The Ghost of Javier Lopez take on the struggling Anthony Rizzo. Lopez vs. Lefties this year: .626 OPS, up from .572 in his career. That’s not that much worse than his career numbers. Rizzo vs. Lefties: .832 OPS. That’s pretty good for a platoon advantage, let alone one’s weaker platoon split. Rizzo is a tough out, recent struggles notwithstanding.
What made this decision questionable is the availability of Will Smith. However, I am willing to bet all the money in my pockets (it’s not much) against all the money in your pockets that Bochy was deterred by the head-to-head numbers. In 11 career PA’s vs. Lopez, Rizzo had an OPS of .586. In 14 career PA’s against Smith, he had an OPS of 1.318. This is a statistically significant difference. The problem is that those are numbers from 2011 to 2016, so they haven’t taken into account the rigor mortis that may have set in on Lopez.
This season, both pitchers held Rizzo to one hit each in 7 combined PA’s (4 Lopez/3 Smith). However, Rizzo’s one against Smith was a three-run double, while his hit vs. Lopez was a harmless single. The point here is that it’s dangerous to put too much into head-to-head history, especially with such small samples. However, it sure looks like that’s what Bochy did.
Romo relieved Lopez after the Rizzo walk, which dropped the Giants’ win expectancy to 89.945% and created a high leverage situation. How did Romo do in high leverage situations in 2016? Very well, as a matter of fact. He limited hitters to a .348 OPS in such moments. That all went for naught when Ben Zobrist smoked a double into the rightfield corner, but it’s hard to blame Bochy for this choice.
Up came Willson Contreras to face the aforementioned Smith. Contreras singled up the middle to tie the game. Jason Heyward then bunted terribly to Smith, who threw to Brandon Crawford, who uncharacteristically fired the ball into the photographer’s well. That put Heyward on second, ended Smith’s night and brought on Hunter Strickland.
After all this calamity, the Giants’ win expectancy plummeted to 44%, which means Strickland was put into a tough spot. Had he gotten Contreras, however, the Giants’ expectancy goes to to 54%.
What’s the point? Ultimately, trying to parse Bochy’s choices last night is like trying to do an autopsy on somebody shot, set on fire, and tossed out of a high rise building only to land in the jaws of a shark. What killed him? It all killed him. The bigger takeaway is that once more, selective results-based analysis still rules. After 23 seasons and 3,400 games, Bruce Bochy’s winning percentage as a manager is .505. That’s it.
Yet many choose to focus on the 62 postseason games he has managed to canonize him. Bochy is a smart guy, and has forgotten more about baseball in the last five minutes than most of us will ever come close to knowing. However, maybe it’s time to reconsider how we evaluate managers and coaches. Last night’s ninth could have happened at any point in the 2010, ‘12, and ‘14 postseasons. Would you look at Bochy differently if it had?
Joe Maddon vs the PA Announcer
Once the meat of the Cubs order had gotten him back into the game, Maddon was going all in, regardless of having played 13 innings and having his closer blow up in his face the night before. It’s very rare, even in the regular season, to burn two players for one PA, but that’s exactly what Maddon did in pinch hitting Contreras for the pinch hitter Coghlan.
The first choice is being held up as bold because Coghlan was to hit for Addison Russell, who had 95 RBI’S on the season. Oy. We’re not going to digress into why that is a horrendous way to evaluate a hitter. We’re just going to point out that Russell’s OPS vs. righties in 2016 was .715. Coghlan’s OPS vs. righties was .627. Did Bochy know this? Did Maddon?
Bochy sure knew that Coghlan’s OPS vs. lefties was .427. So when Maddon doubled down and announced his second pinch hitter, he ended up with Contreras, who has an OPS of .854 vs. lefties, vs. the southpaw Smith. The risk of course is that you have now burned through two players the night after a game went 13 innings and was now tied in the ninth. Maddon didn’t blink, got what he wanted, and Contreras got the hit that put the Giants under.
So Maddon comes off as Obi Wan and Bochy the hapless Stormtrooper because the move worked. But he should really be praised for his rationale and willingness to aggressively pursue victory instead of robotically avoiding defeat. Maddon engineered a better situation for his team, and was rewarded for it.
Roberts engineered a confrontation for the ages, then mismanaged his choices after Kershaw walked Harper, and got away with it. Baker either went to sleep, or didn’t believe his best choice could get one batter, right before he then went to said best choice after he was now down a run and looking up at a win expectancy of 14%.
Ultimately, all these moves could have (and in the regular season, assuredly did) backfired or gone the other way. The big takeway is to fight the urge to judge these guys on individual moments, but if you must, try to understand the why of what they did, not just the what of the result.