Bringing in Nick Anderson was a worse decision than pulling Blake Snell

On Tuesday night, the Dodgers wrapped up their first World Series victory in 32 years, but, for a good portion of the game, it looked as if we were on pace for a Game 7.

In the first inning, the Rays opened up the scoring with a solo home run from (who else?) Randy Arozarena, and things were pretty much deadlocked from there. Tampa clung to its 1-0 lead on the back of one man, Blake Snell, who was absolutely masterful in the team’s most important game of the year. The Dodgers had no answer, and Snell finished the evening after throwing 5 ¹⁄₃ innings, allowing just two hits and one run, striking out nine and walking none.

The story isn’t that simple, however. Snell only threw 73 pitches, cruising along without much trouble. A deep start appeared to be in order. . .until it wasn’t. It was only after a one-out single in the sixth inning that Rays manager Kevin Cash emerged from the dugout to lift him. In came Nick Anderson, and Tampa’s slim lead dissipated. Mookie Betts doubled, a wild pitch scored the tying run, and a Corey Seager fielder’s choice gave the Dodgers a 2-1 lead. That half-inning was the turning point. Instead of forcing a Game 7, the Rays went quietly the rest of the way, and the Dodgers secured a 3-1 win to take the series.

Cash’s quick hook was the talk of baseball. (That is, until we learned about Justin Turner’s positive test for COVID-19.) How could he pull a former Cy Young award winner, one who silenced the Dodgers’ potent offense, at just 73 pitches? Plenty of blame was thrown on “analytics,” including from noted-bunting-lover Alex Rodriguez, who summed up this argument by saying in the postgame show that the “front offices are really ruining our game.”

Cash explained his reasoning behind the move after the game, telling reporters that he “didn’t want Mookie or Seager seeing Blake a third time.”

“There was no set plan,” Cash said. “As much as people think, there’s no set plan.”

Of course, the third-time-through-the-order penalty is one of the staples of analytical theory. On average, pitchers are less effective when facing batters the third time through the order, and the Rays utilize this information in their in-game strategy more than any other team. It’s why they’ve adopted The Opener strategy. Getting starters — or, rather, “bulk” guys — to start facing the middle of the order can potentially increase their length.

These are averages, of course, and people were quick to jump on analytics as being the reason why Snell was pulled. Just because, on average, pitchers tend to do worse the third time through the order doesn’t mean they will in any particular start. Plus, if Snell would’ve done “worse” the third time through on Tuesday, there’s still a decent chance that worse for Snell would have still been much better than any other pitcher brought in.

On the other hand, though, there’s little predictive power in a pitcher’s start to his finish, even when he’s pitching well. A starting pitcher is only dealing until he’s not. As Connor Kurcon noted on Wednesday, even in games where a starter has jumped out of the gate with dominant stuff, that doesn’t mean he will maintain it. From 2017 to 2019, in the 202 games where the starting pitcher went five-plus innings with nine or more strikeouts, one or fewer walks and one or fewer runs allowed (like Snell), these pitchers posted a 0.62 ERA through five. 181 of the 202 pitchers came out for the sixth inning, and they allowed 72 runs in 167 ²⁄₃ innings the rest of the way, good for a 3.86 ERA. This, too, excludes the runners they left on base.

Cash’s move to replace Snell, from a mathematical standpoint, doesn’t look as bad through this lens, though it does remain a tough look to take out a franchise cornerstone at just 73 pitches in an elimination game. That’s why, even with this considered, I don’t necessarily think I would’ve pulled Snell at this same spot.

However, the worse decision, at least to me, was to bring in Nick Anderson out of the bullpen. If Cash brings in an arm that doesn’t allow any runs in this spot, the decision might be criticized, but it’s probably not chastised. Confirmation bias is an incredibly powerful tool. Seeing Snell come out of the game “early” before the Rays promptly give up their lead is what makes people believe that this is a bad strategy. A world in which the Rays win this game with their shutdown bullpen — as they had all year, I might add — would make people think vastly different things about this decision.

But why Anderson?

Of course, Anderson was the Rays’ best reliever during the regular season, but that clause alone gets to the crux of the problem: during the regular season. Into last night, Anderson’s playoff performance was anything but Anderson-like. As I wrote on Oct. 17, the warning signs were already there: the combination of a low strikeout rate and more hard contact allowed seemed like more than just a small sample aberration. In the time since I wrote this piece, Anderson’s playoff sample got larger, and the results were no better. He finished the postseason with a 14.8 percent strikeout rate, 6.6 percent walk rate, and a 5.52 ERA. It was a 10-appearance sample, sure, but that strikeout rate was by far his worst 10-appearance sample of his career, far lower than anything “better competition” alone could explain:

Over Anderson’s worst 10-game regular season stretch, he struck out 21.1 percent of hitters. In the postseason, represented by the black vertical line, he struck out just 14.8 percent. Clearly, something was off, and if there was any team that I think should have been able to recognize this, it was the Rays.

And it’s quite possible that they did recognize this (I mean, they are the Rays) and still felt that he was the best option considering the situation. He’s your relief horse, and you’re going to ride him through both thick and thin. But, you also have to figure that this should have worked its way into the calculus with regard to the yanking of Snell. If you know that Anderson isn’t as good as he normally is, but still want him to pitch in the high leverage postseason spots, wouldn’t that be the very reason to let Snell go deeper into the game? The Rays pulled Snell as if they had a regular season version of Nick Anderson coming in behind him, and alas, they did not. We all knew it, too.

The Rays had an impressive arsenal of high octane relief arms at their disposal. It didn’t have to be Anderson. But it was. And that’s what cost them Game 6.

Devan Fink is a sophomore at Dartmouth College and a Contributor at Beyond The Box Score. Previous work of his can be found at FanGraphs and his own personal blog, Cover Those Bases. You can follow him on Twitter @DevanFink.