Game 6 of the 2020 World Series: worst managing decision ever?
Here’s something to read while we wait to find out who’s president…
Quick baseball quiz: recently, a team lost the World Series when the manager pulled his starting pitcher who was pitching well and replaced him with a reliever who wasn’t.
Name the team, manager and starting pitcher.
If you said the Tampa Bay Rays, Kevin Cash and Blake Snell, you’re right, but you also could have said the Houston Astros, A.J. Hinch and Zack Greinke.
The last two World Series were lost when managers ignored what they were seeing on the field and based their pitching decisions on analytics. This year Kevin Cash pulled Blake Snell because he didn’t want the top of the Dodgers’ lineup to see Snell a third time; last year A.J. Hinch pulled Zack Greinke because Greinke’s pitch count was in the eighties.
You can blame the managers – and plenty of people did – or the way baseball is being played these days. And for more on that, let’s go back to 2003 and Moneyball.
The role of the manager
The analytics movement – analyzing baseball through the study of statistics – was well underway before Moneyball was published, but the book made the general public aware of analytics, became shorthand for playing the game a certain way and is still a handy reference point because it records some of the principles that changed baseball.
Let’s start with the role of the field manager and a Moneyball quote from then Oakland A’s front office executive Sandy Alderson on the hiring of field manager Art Howe:
“Art Howe was hired to implement the ideas of the front office, not his own. And that was new.”
Might have been new back then, but it’s not anymore.
These days, if a guy wants to manage in the big leagues his odds are greatly improved if he’s willing to do whatever the front office tells him to do. Today’s managers are often seen as mid-level executives whose job is to talk to the media, implement the ideas of the front office and take the heat when those front office ideas don’t work out.
When a team loses, it’s the manager who gets to explain what went wrong in the post-game press conference and a manager who says, “Don’t look at me…the analytics department told me to do it that way” probably won’t be a manager for long.
Kevin Cash has said the decision to take Snell out of the game was not mandated or pre-scripted, but Kevin Cash also said, “I didn’t want Mookie and Seager seeing Blake a third time” which sounds pretty pre-scripted; he was pulling Snell after two trips through the order no matter what. And if the Rays analytics department had no input into Cash’s decision, it would be highly unusual.
If Cash fell on a front office sword, he wouldn’t be the first manager to do so; for a lot of teams, taking blame is part of a manager’s job description.
A.J. Hinch made a bad decision in Game 7 of the 2019 World Series, got caught cheating, was suspended for a year and none of that kept him from being named manager of the Detroit Tigers. If you’re inclined to be cynical – and if you’re not, you probably ought to read more – you might suspect Hinch got the Detroit job because he’s willing to do whatever the Detroit front office tells him to do.
Before you say Cash made the worst managing decision ever, it’s worth asking whether Cash made the decision on his own.
Don’t trust your eyes
OK, so all of America saw Blake Snell pitching well and Kevin Cash – who knows way more baseball than most of us – was seeing the same thing; why pull him?
A couple more quotes from Moneyball provide clues:
“The power of statistical analysis depends on sample size: the larger the pile of data the analyst has to work with, the more confidently he can draw specific conclusions about it.”
Stick with that logic and the fact that Blake Snell already punched out the first three batters in the Dodgers’ lineup six times doesn’t matter because six at-bats is a small sample size; what Snell had done over the course of the 2020 season when facing a lineup three times was more important.
Which would be an OK plan if the game was decided on what Blake Snell did on average, but averages are just that. Some nights pitchers are better than average, some nights they’re worse than average and the manager’s job should be figuring out what a pitcher has that night and manage accordingly.
But if you’re an analytics fanatic, what you see doesn’t matter and here’s another Moneyball quote about that:
“The naked eye was an inadequate tool for learning what you needed to know to evaluate baseball players and baseball games.”
The theory – and there’s some truth to it – is the eye can be fooled because emotions or confirmation bias or any number of other factors might get involved, so you ignore what you see and concentrate on the numbers.
OK, but which numbers?
A plethora of numbers
Let’s say you buy into analytics 100 percent and decide to do what the numbers tell you to do; you’re still gonna have to decide which numbers matter.
For example: in 2020 opponents hit .304 and slugged .609 against Snell the third time they faced him, so you might want to pull Snell after two trips through the order. But over Snell’s career those numbers are .247 and .413, so you might want to leave Snell in the game.
Do you base your decision on Snell’s 2020 numbers or Snell’s career numbers?
And we’re just getting started.
Run Support suggests Snell should have stayed in the game.
Platoon Splits say Snell is better against lefties and Mookie Betts is right-handed.
Number of Outs in Inning, Bases Occupied and Clutch Stats favor leaving Snell in.
By Inning suggests taking Snell out.
Mookie Betts’ Platoon Splits says leave Snell in.
Mookie Betts’ Clutch Stats says take Snell out.
Mookie Betts’ Times Facing Opponent in Game vs. Starting Pitcher says leave Snell in.
If you want to make yourself crazy – assuming you’re not already – you could also factor in Pitch Count, Leverage, Ballpark, Game Conditions, Days of Rest, By Catcher, Batting Order Positions, compare 2020 stats with career stats and regular season stats versus postseason stats and if you don’t get the point by now, you’re not going to.
The analytics guys can swamp you with enough numbers to bury a Ford F-150 so you better decide which numbers matter because you’ll have a hard time absorbing all of them. Which is why some people believe you might want to go through the numbers and know what they are, but in the end you better make your decisions based on what you see that night.
Numbers tell you what happened in the past; your eyes tell you what’s happening right now.
And that night Blake Snell was dominating.
The refusal to acknowledge playoff baseball is different
Here’s another Moneyball quote and this one is from Oakland’s GM at the time, Billy Beane:
“My shit doesn’t work in the playoffs. My job is to get us to the playoffs. What happens after that is fucking luck.”
Beane was trying to explain why the Oakland A’s had never done all that well in the playoffs and blaming bad luck. But the playoffs are different because you no longer have to win 95 games out of 162, you have to win right now tonight and what happens in the long run or on average may not matter.
Pulling Snell and going to reliever Nick Anderson because it’s the same move you made all year – one of the excuses given for doing it – ignores the fact that Anderson was scuffling in the postseason and Snell was pitching well.
Ignoring what Anderson and Snell were doing in the present in favor of what they’d done in the past was one of the reasons the Rays lost the World Series.
Thinking outside the box
One last quote from Moneyball:
“The new outsider’s view of baseball was all about exposing the illusions created by the insiders on the field.”
The analytics guys got a lot of credit for questioning baseball dogma and how things were done; they were willing to think outside the box. But then a lot of the analytics guys built a brand new box, climbed inside and nailed the lid shut. Home runs are good, the stolen base is bad, defense doesn’t matter and if you question any of that, you probably shouldn’t work here.
I say this every time I write about analytics:
I’ve talked with all kinds of players and coaches from a bunch of different teams and I’ve never heard one of them say they don’t want the information their analytics department provides. They want that information, but then most of them then want the freedom to play the game as it unfolds.
If a runner on first base realizes he can see the catcher’s signs and knows when a curve is coming and can steal second base, there’s no time to call a meeting of the analytics department and get permission to run.
If a catcher looks down and realizes the batter has moved up in the box to catch the off-speed pitches early, there’s no time to call a meeting of the analytics department and find out if it’s OK to throw fastballs.
And if the manager sees his starting pitcher dominating, there’s no time to call a meeting of the analytics department and get permission for his pitcher to face the other team’s lineup three times.
Getting locked in to a game plan created before the game is played is dumb and smart guys – and anyone in analytics isn’t a dummy – shouldn’t do it. The guys who play the game are pretty smart as well and sometimes you just gotta let them play.