Sabermetrics and analytics have taken over baseball. We've known this for a few years, and their hold on the game has only grown stronger as more and more teams have embraced the philosophy. Never was it more apparent than in the World Series, when strict adherence to analytics arguably cost each team a game.
Among the biggest changes in the game that were the result of the analytics revolution are the "opener" and over-sized bullpens. Those are both trademarks of the Tampa Bay Rays that have been copied by virtually every other team in the Majors. And, as the use of openers and over-sized bullpens have become more prevalent, so have predetermined pitching plans. Teams annoying stick to their planned script no matter how well the pitcher currently in the game is doing--a strategy that proved costly for Tampa Bay in Game 6.
Game 6 is perhaps the most important game in any best-of-seven series. It's always an elimination game, so the pressure's on to either avoid or cause a Game 7. So, even though they trailed the series 3-2, the Rays had to feel good entering Game 6 with their best pitcher, Blake Snell, on the mound. And, since it was going to be his final start of the season, you'd figure that you ride your horse as long as he's still dealing. Which Snell certainly was.
He was great in Game 2, when he allowed just two hits and struck out nine in a Tampa Bay win. In Game 6, he was even better. Snell was untouchable for five innings, striking out nine and not walking anybody. The Dodgers only had two base runners--on singles by Chris Taylor and Austin Barnes. However, after the Barnes single, Rays manager Kevin Cash popped out of the dugout to take his ace out of the game. Two batters later, the Dodgers had the lead and the second-guessing began.
For the most part, managers have been hesitant to let starting pitchers face the lineup for a third time in the sabermetrics era. The numbers do give some credence to that argument. Snell's ERA and batting average against were dramatically worse the third time through the lineup than the first and second times. So, the Rays had determined ahead of time that Snell wouldn't face any hitter a third time. And, since Barnes was the Dodgers' No. 9 hitter that meant, single or not, he was Snell's last batter.
Here's the problem with sabermetrics, though. There are things they don't take into account, so you can't rely on them exclusively. For years, managers weren't bound to pages of statistics spit out by computers. They were allowed to read the game situation and trust their gut. And, had Cash trusted his gut, there's no doubt he would've left Snell in, especially considering the Dodgers 1-3 hitters were a combined 0-for-6 with five strikeouts against him to that point. At the very least, he'd finish the inning before going to the bullpen.
That's another point about the over-reliance on bullpens in the sabermetrics age. Tampa Bay was eager to get to the bullpen because it's the strength of the team. But...going to that well over and over again counts on those guys being "on" every night. And the more exposure the opposing lineup gets to those arms, the less likely they are to be "on."
During the regular season, when the longest series you're gonna play against a team is four games, you can get away with using going back to the same relievers all the time. In a seven-game series, though, the more a reliever pitches, the more likely it is the opposing hitters will figure him out. That, combined with the fact that starters frequently get a quicker hook in the postseason means you're counting on the bullpen to get far too many outs night after night. And that's a lot to ask of any bullpen, no matter how good it is.
Contrast that to how the Dodgers handled their bullpen in Game 6. Their closer, Kenley Jansen, had his issues all season, and those struggles carried over into the playoffs to the point where Dave Roberts wouldn't commit to using him at the end of games. Jansen pitched the ninth inning of Game 3 (in a non-save situation) and gave up a home run to Randy Arozarena. Then with the Dodgers up one in the ninth inning of Game 4, he blew the save after giving up a single, a walk and that crazy walk-off sequence.
Nevertheless, you would've figured that the Dodgers' all-time saves leader would get the ball in the ninth inning with the chance to close out the World Series. Except he never even got up! Julio Urias had retired all four batters he faced in the seventh and eighth, so Roberts stuck with the hot hand in the ninth. And he promptly retired the Rays in order to clinch LA's first championship in 32 years.
Maybe the Dodgers learned a lesson earlier in the series. In Game 2, they did the "bullpen game" and used Tony Gonsolin as an "opener." Generally speaking, doing a bullpen game doesn't work in the World Series. And it didn't. It cycled back to Gonsolin again in Game 6, but this time Roberts vowed to use him as a traditional starter. Of course, Gonsolin ended up only going 1.2 innings, but he went in with a completely different mindset.
Roberts received his share of criticism during the World Series, both for the staff game in Game 2 and for his usage of Clayton Kershaw in Game 5. Kershaw was awesome in Game 1, giving up one run over six innings. Yet, going into Game 5, Roberts said that Kershaw was only going to face around 21 hitters. For a veteran starter, let alone a Hall of Famer like Kershaw, the third time through the lineup isn't a problem. So why are you committing to two turns through the lineup plus three hitters? And why are you announcing it to the game?
Kershaw wasn't as good in Game 5 as he was in Game 1, but he was still effective. Yet, sure enough, as soon as he faced his 21st hitter, here comes Roberts out of the dugout to lift his starter (for the record, Kershaw also faced 21 hitters in Game 1). Why are you taking him out with two out and nobody on in a 4-2 game just because he hit his magical (and arbitrary) number of hitters limit (which is vastly different than a pitch count limit)? At the very least, have him face the next hitter and try to get out of the inning without having to use a reliever.
There were plenty of examples of sabermetric/analytic-based bullpen mismanagement throughout the playoffs. The Rays benefited from the Yankees' bullpen overuse and ill-fated attempt at using an opener during the Division Series, and the Dodgers almost blew it in the NLCS with their Game 7 opener strategy.
One of the biggest cliches about playoff baseball is that what works in the playoffs doesn't necessarily work in the regular season. You have more off days. Thus you can get away with using the same relievers more frequently. And you can use them for longer outings or in different situations than you otherwise would. (Or, as the 2018 Red Sox showed, sometimes you can use your starters out of the bullpen.)
However, the opposite is also true. Don't overthink it. Stick with what got you there. If you don't normally do bullpen days or use openers during the regular season, don't do it in the playoffs! And, even if it is the strategy that got you there, don't let it dictate everything you do. Because sometimes watching the game can tell you far more than analytics can.
No comments:
Post a Comment