How analytics lost 2 of the last 3 World Series
Misapplication of statistics is gutting the heart of baseball.
(Original Nov. 5, 2022. Published on DeFranco Post on Sept. 5, 2023.)
On a Saturday night in Houston, Nov. 3, 2022, live on national television and in front of millions of people around the globe, Philadelphia Phillies manager Rob Thomson, casually as sipping a warm cup of coffee, committed highway robbery.
It was Game 6 of the World Series, his club led the Houston Astros 1-0 in the bottom of the sixth, trying to force a Game 7. His right-handed ace, Zack Wheeler, was dominating, having thrown only 70 pitches while surrendering just 3 hits.
The Astros, with a fluky hit-by-pitch, a fielder’s choice, and a bleeding, ground-ball single up the middle, had put runners on the corners with one out.
That’s when Thomson, with everything on the line, decided to stride out to the mound and remove Wheeler from the game, bringing in hard-throwing lefty Jose Alvarado to face left-handed Houston slugger Yordan Alvarez — whom Wheeler had retired twice, on three pitches.
“I thought Wheels still had good stuff, it wasn’t about that,” Thomson said. “I thought the matchup was better with Alvarado.”
The move almost immediately blew up in Thomson’s face, and blew up the Phillies’ season with it. Alvarez launched the fourth pitch from Alvarado — a flat, 98-mph sinker above the knees, in the middle of the plate — deep into another stratosphere, high over the centerfield wall and onto a towering balcony of filled seats, a 450-foot, mammoth shot that will live in legend, and Houston went on to a 4-1 victory and the World Series crown.
“It was the sixth inning, and I felt like the normal back of the bullpen guys could get through it,” Thomson said, after robbing Wheeler and Phillies fans everywhere of a shot at history.
The operative words were sixth inning, because Wheeler was going through Houston’s lineup for the third time. According to Daddy Analytics, that means you pull your starting pitcher. Because, statistically, the hitter often wins the third time through against a starter.
Fox play-by-play broadcaster Joe Davis oddly claimed, “This is what Thomson has done all year.”
Replacing a dealing Wheeler after 70 pitches, at the slightest sign of trouble, to play the matchup game with a struggling Alvarado, who had had difficulty locating his pitches all series, was what Thomson had done all year? Huh?
Nope.
Vanilla Joe Davis, with extra rosy TV makeup on his cheeks, does a solid job and is benign enough (anyone is more tolerable than Joe Buck, although that’s an issue for another day), but he was glossing over reality by not wanting to second-guess Thomson.
The reality is, Alvarado was sent down to Triple-A for a few weeks during the summer to work on his command. He returned a different pitcher, much improved, with an unwavering mindset and a badass cutter that hadn't been so effective prior to his stint in the minors.
But in this World Series, he had been (and ended up being) the old Alvarado, jittery and unreliable in big moments. His pitches hadn't done what he wanted them to do, and he had allowed every inherited Astros runner to score — and then some.
Thomson hadn't made such a pitching move all year. It was a colossal blunder, historically bad, one of the worst in World Series history.
“Honestly, he (Thomson) just caught me off guard,” Wheeler said. “That’s a tough pill to swallow, but it’s ultimately his call, and that’s the call he made.”
Thomson, despite his calm demeanor and pressing all the right buttons to guide the Phillies here, panicked. He yanked his best starter, the most exquisite horse in his stable, because of analytics.
Thompson said the move was about righty/lefty. But Alvarez is similarly dangerous against left-handed and right-handed pitchers. So why the hell would he put in a power pitcher who hadn't been able to locate all series against a power hitter who, as anyone who paid attention this postseason knew, could have erupted at any moment?
What is more hazardous than a power pitcher who can't locate? All the hitter has to do is put the bat on the ball, and it's over. And it was.
That was it.
The worst part is, we had seen this movie before — just two years earlier, in fact. And it, too, happened because of analytics.
‘Those who don’t learn history are doomed to repeat it’
In the 2020 World Series, it was Blake Snell on the mound for the Tampa Bay Rays, who were trailing 3-games-to-2 against the Los Angeles Dodgers.
Rays Manager Kevin Cash pulled Snell — also in the sixth inning, with one out and a 1-0 lead, in Game 6. Snell (who later became well-known to Phillies fans when he broke Bryce Harper’s thumb) had thrown only 73 pitches and was absolutely dazzling. He had struck out the Dodgers' first three hitters six times. He had given up just two hits and fanned nine.
After Cash’s move, Los Angeles rallied for two runs against Tampa Bay reliever Nick Anderson, going on to a 3-1 win and the Dodgers’ first World Series title since 1988.
Several members of the Dodgers, including manager Dave Roberts, said afterward that they were totally relieved and thrilled when Snell was pulled, because he was in complete control of them.
“At that point I was like, ‘I got a chance,’” Mookie Betts said. “Snell was rolling.”
Cody Bellinger said he was “shocked” by the move because Snell “had his stuff .. he was gross.” He added that the move “definitely uplifted us.”
Cash, for his part, was unapologetic, citing analytics.
“The lineup the Dodgers feature is as potent as any team in the league,” he said. “Mookie coming around for the third time through, I value that. … I didn’t want Mookie or (Corey) Seager seeing Blake a third time.
“I’d do it again,” Cash later added.
The analytics of the moment
Analytics has been part of the game almost since its inception, but never to such an absurd extent. The Dogma of Analytics fails to account for feel.
How do you feel? I feel good, how about you? How does your arm feel? How do you feel about me? How do I feel about this situation?
Daddy Analytics fails to account for momentum, clutch performance, being “in the zone” — otherwise known as the human element. It fails to account for your own eyeballs and common sense.
Daddy Analytics also doesn’t include real-time analytics. For example, just because a guy performed a certain way on a Tuesday in late July doesn’t mean he’ll perform that way in the postseason (for better or worse). Some guys shrink, some guys excel. In basketball, it’s called “the Jimmy Butler effect,” because he seems to find new heights in the playoffs.
What made Derek Jeter great? Certainly, part of it was his consistency over two decades in the majors. But a big part of it was how he elevated his performance in the clutch, with memorable play after memorable play, in October and November.
Where was the analytics of the moment in the 2020 and 2022 World Series? Where was the real-time analytics?
Where was the analytics of, How does Snell perform the third time through in Game 6 of the World Series when he’s living his childhood dream and dominating the Dodgers? or How does Wheeler respond with the World Series on the line when he’s pitching the game of his life?
There’s usually no book on those situations. That’s when a manager has to rely on the limited stats he has available in real-time as well as on his eyes, his gut, and his personal knowledge of his players.
It is why analytics, while important, aren’t the only important factor in winning. It is why push-button managing doesn’t work.
Could the Astros and Dodgers still have ended up winning the World Series if Thomson and Cash hadn’t made those moves? Of course. But who knows. Those were pivotal points on the biggest stage.
Overuse, and misuse, of analytics robs us all.
What fan, outside of Houston and Los Angeles, didn’t want to see Wheeler and Snell stay in the game?
“I am definitely disappointed and upset,” Snell said after his early exit. “I just want the ball. I felt good. I did everything I could to prove my case to stay out there, and then for us to lose, it sucks.”
Most managers pray their ace starter has his electric stuff in a World Series game. What are the analytics on how rare it is for your starter and every guy out of your bullpen to have their best stuff on the same day under such a spotlight?
Billy Beane made a name for himself as general manager of the Oakland A’s by employing Bill James-style analytics to build a team on the cheap, to re-conceive the approach to the game and recreate the statistics of unaffordable star players in the aggregate.
Even in the years when Oakland has been good, when ownership hasn’t sold off its talent as soon as it came time for new contracts, Beane’s adherence to analytics has yielded a grand total of zero championships.
The A’s haven’t even come close.
“My shit doesn't work in the playoffs,” Beane himself admitted. “My job is to get us to the playoffs. What happens after that is fucking luck.”
And the human element.
Email the DeFranco Post at chris@defrancopost.com, or comment below.