October 28, 2015

Playoff Prospectus

Assessing the Managers' Moves in Game One

by Matthew Trueblood

Printer-friendly

Contact Author

As we’re slowly learning, the single most important tactical step a manager can take to win a playoff game is to minimize his starting pitcher’s exposure to the opposing lineup for a third time—if not eliminate that exposure entirely. I say we’re slowly learning, of course, though it’s really a thing many saber-savvy fans and analysts learned five or 10 years ago. It was a major theme of The Book. The slowness of teams’ embrace of this concept has made it maybe the last issue on which outsiders have a substantial, obvious advantage over the median big-league manager.

We’re on our third or fourth October of talking about the progression of offense each time the lineup card turns over, and at least our second of assertively calling for managers to get proactive, not reactive, about the removal of their starters. The compression of the regular-season schedule and the relatively low leverage of many regular-season fifth and sixth innings conspire to make the traditional model of starting pitching a viable strategy over the full 162, but here, in the World Series and even before it, managers have to get over the mental blocks that prevent them from lifting starters at low pitch counts, while they’re pitching well, with 12 or even more outs left to get. Neither manager could summon the creativity or gumption to break the mold on Tuesday night, though, and it backfired on both.

Edinson Volquez retired the Mets in order in each of the first two innings. In the third, though, he hit a batter, and walked the next one. In the fourth, he allowed three hits. Through four frames, he’d faced 17 batters. It probably never even crossed Ned Yost’s mind to remove his starter that early, and maybe it ought not to have. If it had, though, and if Yost had obeyed that imagined impulse, here’s what he could have avoided: two innings, three hits, two runs, and one strikeout. If Kris Medlen and Danny Duffy couldn’t have gotten the Royals that far, they didn’t deserve to win the game.

Matt Harvey, meanwhile, cruised through five innings. The only run he gave up was on what was, even if the official scorer didn’t have the heart to call it one, an error—a major defensive gaffe on Alcides Escobar's first-pitch inside-the-park-homer née flyout . He’d only fanned two Royals (of 19 faced), but if he couldn’t dodge their bats entirely, he was, at least, missing their barrels. Harvey wasn’t pitching poorly, his pitch count was low, and his stuff was good (though not quite his best). In response to the well-documented success the Royals have had against particularly hard fastballs, he leaned away from his heat, mixing his secondary pitches effectively and keeping Kansas City off balance.

Terry Collins had a somewhat more desperate bullpen situation about which to worry than did Yost, but he also only needed 12 more outs to win a game New York led 3-1 by the time the bottom of the sixth began. With Ben Zobrist, Lorenzo Cain and Eric Hosmer due, it was a very good time to go to a high-leverage reliever disguised as just another arm. Instead, Collins left Harvey to try to find his way through that dangerous Royals lineup again, and a half-inning later, the game was tied.

There was no obvious occasion for either manager to remove his starter; there was just the whisper of opportunity. One of these days, a manager will storm into October with 12 honest-to-God pitchers and start lifting his starter 10 outs into games. That manager will be laughed at by Harold Reynolds, questioned fairly sharply… and then praised, because he’s going to win a lot of games.

Let’s move off the third-time-through thing, though, because some other interesting decisions were made on Tuesday night, and several of them helped shape the marathon game. Arguably, the biggest real mistake either skipper made was one Collins made before the game even began. It came when Collins decided to keep his outfield alignment the way it always looks when the Mets face right-handed starters: Michael Conforto in left field, Yoenis Cespedes in center, Curtis Granderson in right. In so doing, he left Kelly Johnson as the obvious choice to DH (and batted him ninth), and hamstrung his defense early in the game. That’s why, when Harvey inexplicably elected to throw Alcides Escobar a first-pitch fastball in the strike zone to lead off the Royals’ half of the first, it was Cespedes in center field. It was Cespedes getting the crummy break, Cespedes miscommunicating with Conforto, Cespedes not even getting his glove up, and Cespedes watching it zig away.

Juan Lagares catches that ball, and Lagares isn’t so much worse than Johnson at the plate that choosing Johnson’s bat over Lagares’ glove made any sense. Conforto should have been the DH, with Lagares in center and Cespedes in left. Johnson could have pinch-hit for Lagares if a critical offensive situation arose late in the game, and Kirk Niewenhuis could have gone in to play defense thereafter, if needed. Instead, the Mets cycled three players through the DH slot in six plate appearances, and the only time they got on base was the time Volquez completely lost the release point on his curveball and planted it in Johnson’s thigh. They could have had Lagares’ glove and Conforto’s bat for much longer if Collins hadn’t so prized offense.

Yost should have gone to his bullpen sooner, but everything else about his use of it was absolutely perfect. He should use Danny Duffy to push Johnson out of the game on the second turn again on Wednesday, if he has a chance, and then to get Granderson, just the way he did Tuesday night. Collins probably erred slightly in the Royals’ eighth, when Ben Zobrist led off with a double and Tyler Clippard struck out Lorenzo Cain. At that point, the next three hitters due were Eric Hosmer, Kendrys Morales and Mike Moustakas. Morales’ lack of power from the right side and the left-handedness of Hosmer and Moustakas made Jon Niese the man for that moment, and Niese was up behind Clippard in the bullpen. Instead, Collins let Clippard face two more batters, then called on Jeurys Familia to end the threat.

Normally, using a closer for more than three outs is an automatic plaudit in October. Not so much here. It was unnecessary; Niese could just as well have navigated out of the same trouble, given the matchup. Familia probably gives up that home run to Alex Gordon even if he’s not asked to get the last out of the preceding inning, but there’s an argument that pitching so briefly, then sitting down and coming back, had some effect on him. It’s a tricky variable to whose vagaries Collins left himself exposed by trusting Niese too little.

On the positive side, for Collins: Lagares singled twice after entering as Conforto’s defensive replacement. On the first, with Kelvin Herrera (on whom runners have not been afraid to steal, this year; his TRAA was 4.64 percent) on the mound, Lagares took off and stole the base. That’s what allowed him to score the go-ahead run on Eric Hosmer’s error. The second time Lagares singled, though, it came against Ryan Madson, and Madson (with a -5.48-percent TRAA this season) is much harder to run on. Collins’ decision to have Wilmer Flores bunt Lagares over remains a sketchy one, but it makes sense in light of the hurler on the mound—just as the decision to send him in the eighth made sense.

Long games like Game One are exercises in universal balancing out. One manager might make a mistake, but if the game stretches on, the other will probably make one, too. That’s how this one went, for the most part, but while the players decided the final outcome, Yost slightly outmanaged Collins.

Matthew Trueblood is an author of Baseball Prospectus. Follow @MATrueblood
Click here to see Matthew's other articles. You can contact Matthew by clicking here