Tampa Bay Rays lose at Bullpen Russian Roulette

Tampa Bay Rays manager Kevin Cash pulls Blake Snell because his numbers say so…and the Dodgers make him pay

The removal of Tampa Bay Rays starter Blake Snell Tuesday may become the most talked-about moment of Tuesday night’s World Series Game 6.

Snell was removed by Rays manager Kevin Cash with one out in the sixth inning of a game the Rays led 1-0. He had just allowed a base hit to Austin Barnes, only the Dodgers’ second baserunner of the night.

Within moments, the Dodgers scored twice off reliever Nick Anderson and those runs were the margin by which LA clinched its first World Series win since 1988 with a 3-1 victory.

Cash explained his decision to remove Snell in straightforward terms. The left-hander had already gotten the Rays 16 outs; that was more than in all but two of his appearances since going down with arm problems in July of 2019. In the dumbed-down world of modern pitching expectations, Cash said “I felt Blake had done his job and then some.”

He also noted that Snell had completed  two circuits of thee Dodger lineup. Managers – especially statistically minded ones such as Cash – have very much of a thing about letting an opponent see a pitcher three times in one night. “I did not want Mookie or Seager seeing Blake a third time,” Cash said.

Snell’s 2020 performance could be read as an explanation for Cash’s reasoning. Opponents hit only .140 against him the first time through the order, but .307 in their second looks and .304 after that. Of course opponents got few third looks at Snell in 2020; he faced only 23 batters more than twice all season.

As with many managers who rely principally on stats to make their decisions for them, the problem with Cash’s decision was that he was basing it on aggregate, season-long data. In other words, on the performance of the Average Blake Snell.

But the Average Blake Snell wasn’t on the mound Tuesday night; the extremely good one was. Of those 18 opponents he faced, Snell struck out half, matching his regular season best.  He did not walk a batter.

Managers who lean  too heavily on statistical analysis in their planning generally frame their plans on the expectation that every pitcher – especially every reliever – will produce his baseball card numbers every night. In other words, that his humans are robotic creatures.

But pitchers are not robotic creatures; they have good nights and bad nights. When a manager removes a pitcher because his pre-game planning tells him the time is right to do so, he sometimes sets off a bullpen chain reaction that turns a baseball game into something approaching Russian Roulette.

Sometimes, as it did Tuesday, the gun goes off on the first pull of the trigger. Sometimes it takes two, three or four pulls. And sometimes the manager gets it right; every reliever does perform exactly as the baseball card says he should. In theory, anyway, there can be winners at Russian Roulette.

On Tuesday, Cash replaced Snell with Nick Anderson, a very good pitcher in 2020. He made 19 appearances, worked 16 and one-third innings, struck out 26, permitted just eight baserunners and allowed just a single earned run.

But Nick Anderson is human and therefore liable to fall. On Tuesday, he fell, one more victim of bullpen Russian Roulette. A double to Mookie Betts preceded a wild pitch that tied the game.  When Ji-Man Choi relayed Corey Seager’s ground ball home too late, Betts raced home with what would become the winning run.

It is plausible, certainly, that had Snell been left in to face Betts and Seager, the same thing – or worse – would have happened. All we know for certain is that Cash played Bullpen Russian Roulette, pulled the trigger and a fatal bullet tore through the Rays’ hopes.

The one fair question concerns the status of Snell’s stuff at the time he was removed.  In two words, darned good. The velocity of the last five fastballs he threw averaged 95.6 mph, almost precisely the 95.7 mph average of his first five.

Nor were the Dodgers rocking him on those rare occasions when they made contact. Including Barnes’ base hit, the average exit velocities of the last three batters Snell faced who made contact was 73.5 mph. nearly 10 mph less than the 82.5 mph average of the first three to touch him.

It may be historically unusual for Snell to improve as a game proceeds, but that’s what he was doing.

The real question, of course, is not whether Snell was better than when he started but whether he was better when he left than his replacement. He certainly didn’t throw as hard as Anderson, whose 94.9 mph average four-seamer Tuesday was two mph faster than Snell’s average.

A more important number than velocity, however, is exit velocity. The average contact off Snell Tuesday left Dodger bats at 78.4 mph…about 14 mph slower than he threw it. The average exit velocity from an Anderson pitch was 87 mph, nearly 10 mph faster than against Snell.

Worse, the removal of Snell with11 outs remaining set off that inevitable bullpen chain reaction. Anderson was followed by Aaron Loup, who was followed by Pete Fairbanks, who was followed by Ryan Yarbrough.

Collectively, the average exit velocity off the bats of all Rays relievers was 90.9 mph, a full 12 mph faster than off Snell.

Dodger  batters produced four contacts Tuesday in excess of 95 mph. Four of those five came off Snell’s replacements.

It will never be known whether Snell would have continued to be as dominant had he been left in. What is known is that Cash played Bullpen Russian Roulette, and the gun went off with fatal consequences to the Tampa Bay Rays World Series hopes.