On Tuesday night, the Tampa Bay Rays leveled the ALDS with a 7-5 win over the Yankees. Given that it ended with the tying run on base and Aaron Judge at the plate and featured some spectacular relief pitching from Tampa Bay’s stable of guys, as well as two ludicrous home runs from Giancarlo Stanton, this game delivered from an entertainment perspective. But more than that, it offered an enticing educational opportunity.
So let’s talk about the opener.
Two years ago, the Rays started an experiment that changed the nature of the starting pitcher. Rather than opening their games with a pitcher expected to go five innings or more, Tampa Bay began using a right-handed one-inning reliever (usually Sergio Romo, Ryne Stanek, or Diego Castillo) who would pitch to the first few batters of the game before being replaced with a left-handed pitcher (Ryan Yarbrough or Jalen Beeks) who would come in and pitch the next several innings.
Though the two terms get used interchangeably these days, the opener is a distinct subset of the bullpen game, in which many relievers relay through stints of one or two innings. As long as there have been relief pitchers, there have been bullpen games. The opener system uses the same pitchers as a traditional start, for the same length, but in a different order. The advantage of the opener is that a right-handed reliever can pitch to tough right-handed batters at the top of the order, rather than the left-handed starter.
When Zach Kram wrote about the concept two years ago, he examined a game in which Tampa Bay played the Angels, who stacked Mike Trout, Justin Upton, and Albert Pujols near the top of the lineup. Romo was a much better matchup than Yarbrough or Beeks against that part of the lineup, and by sitting out the first inning, Yarbrough could pitch into the seventh while facing Trout only twice instead of three times.
It was King Solomon who said that there is nothing new under the sun, and the opener is no exception. A variation on the strategy involves naming one starting pitcher who could credibly pitch a starter’s workload, then removing him almost immediately for an opposite-handed starter to screw up the opponent’s platoon plan. In Game 7 of the 1924 World Series, Washington Senators manager Bucky Harris started right-handed starter Curly Ogden, so Giants skipper John McGraw put first baseman Bill Terry in the lineup. Terry, a Hall of Famer, was the last National League player to hit .400 in a season, but in 1924 he hit just .182/.217/.318 against left-handed pitching, and was therefore frequently platooned.
Ogden pitched to two batters before Harris replaced him with George Mogridge, a left-handed pitcher who retired Terry both times he faced him. Washington won in extra innings. Two years ago, Brewers manager Craig Counsell brought back the Curly Ogden maneuver in the 2018 NLCS when he lifted starter Wade Miley after one batter and brought in Brandon Woodruff as a bulk reliever. For five innings, the Dodgers’ heavily platooned lineup had to contend with a pitcher it was distinctly ill-suited to face. (This tactic is an object of fascination in the Ringer MLB family; during Tuesday night’s game a colleague asked whether the Curly Ogden maneuver is becoming more common, or if it just feels that way because Zach and I talk about it all the time.)
The actual Curly Ogden maneuver was banned this year with the advent of the three-batter minimum rule, but in Game 2 of the ALDS, Yankees manager Aaron Boone tried to put his own spin on the scheme. Boone’s scheduled starter for Game 2 was 21-year-old rookie Deivi García, the team’s top pitching prospect. If not for the presence of Gerrit Cole, Garcia would have the best stuff on the Yankees’ staff, but he is inexperienced and was inconsistent during the regular season. Nevertheless, he’s one of New York’s five best starters and was used as a traditional starter in all six of his appearances this season. But while García was working through the bottom of the first, J.A. Happ was already warming up in the bullpen. Happ, a left-handed pitcher, took over in the top of the second and pitched into the middle innings—a traditional-opener/bulk-reliever dynamic by usage, with a soupçon of the element of surprise that characterizes the Curly Ogden maneuver.
It was an elegant plan, both in the sense that the opener in general uses the opposing manager’s own lineup against him, and in the sense that the Yankees were trying to beat the Rays with something they themselves had popularized.
But it didn’t work. Like, it didn’t work to the point that it’d be a genuine surprise if the back-page headline of Wednesday’s New York Post wasn’t some variation on “BOONE-DOGGLE!” García allowed a first-inning home run to Randy Arozarena, while Happ lasted just 2 2/3 innings, in which he walked three and allowed five hits and four runs. The Yankees trailed or were tied from Arozarena’s home run through the end of the game, and only kept it close because of Stanton’s two dingers, the latter of which went 458 feet and sounded off the bat like someone had set off an M-80 inside a destrung piano.
The Rays, expecting to face the right-handed García for five innings or more, stacked the lineup with six lefties. On paper, that would favor an unexpected swap to a left-handed bulk reliever. But three of the Rays’ five lefties batted in the first four spots in the order, which meant that they all came up in the first inning against García, not Happ. Another flaw in Boone’s plan: Happ isn’t particularly good against lefties or particularly bad against righties. In nine starts this year, right-handed hitters posted a .685 OPS against him, compared to .577 from lefties. For his career, the difference in his platoon split is 73 points of OPS, which is noticeable but not egregious.
Most importantly, though, Happ just plain didn’t pitch well. He allowed three walks and two hits to lefties, committed a throwing error, and allowed two home runs to right-handed opposition. Not even Stanton’s one-man ballistics experiment could put the Yankees ahead while they were allowing runs at such a pace. Even dumping the blame at Happ’s feet is unsatisfying. He’s made one regular-season relief appearance since 2015, and the last time he pitched in any kind of bulk-relief capacity in the playoffs was his very first postseason appearance, which consisted of three innings of mop-up duty in the 2008 NLCS. After the game, Boone, Happ, and García all told reporters that Boone hadn’t told either player exactly how long they were supposed to pitch. Happ said he would rather have started, and demurred when asked whether he felt like he’d been put in a position to succeed.
Then there’s the fact Happ has become something of an avatar for the inadequacy of the Yankees’ rotation in the playoffs; he’s good enough for regular-season work but not the kind of guy who can be counted on to shut down an opponent in Game 7 of the World Series. That’s why the Yankees traded for James Paxton and signed Cole, so Happ wouldn’t have to suck up innings in this situation.
Boone’s plan might have worked if he’d articulated it to his players more clearly, and executed it in a different moment, against a different opponent, and with different pitchers. But it was not to be. This plan, in this moment, with these players generated the result it deserved.
