Analytics: is the movement losing steam?
Now that their ideas have been road-tested, some flaws are starting to show...
Hey, just because the season is over, it’s 18 degrees and snowing outside, that doesn’t mean I’m done writing about baseball. When you’re in the middle of the action it’s all you can do to keep up; once the dust settles you have time to sift through the wreckage and try to figure out what just happened and why.
And since Jack Frost is nipping at my nose and every other appendage near and dear to my heart and belt buckle, I’ve decided to stay inside and look at some numbers.
Like these:
During the 2019 postseason the Houston Astros hit more home runs and walked more often than the Washington Nationals, but still scored fewer runs. Clearly, walks and home runs don’t hurt your cause, but just as clearly walks and home runs aren’t the only things that matter.
During the postseason the Nationals had a higher batting average than the Astros, a higher on-base percentage and a higher slugging percentage. They also had more saves and a better team ERA. In almost every category the margin was thin, but enough to make a difference.
Bottom line: the better-balanced team won the World Series and that gets us to today’s subject – analytics.
A Moneyball myth
One of the fallacies promoted by the book Moneyball is that certain parts of the game really don’t matter. Bunts, steals and hit and runs are stupid and as long as a guy walks and hits home runs you don’t care if he couldn’t catch a cold in an Alaskan nudist colony.
Author Michael Lewis quoted former aerospace engineer Eric Lewis who estimated that defense was, “at most five percent of the game” which tells you a couple things for sure:
Eric Lewis never spent much time pitching with a horseshit defense behind him and aerospace engineers might not be the best source of baseball wisdom. Just because a guy knows a lot about one thing, it doesn’t mean he knows a lot about everything and that includes baseball.
I’m guessing when that Juan Soto single skipped past right fielder Trent Grisham and knocked the Brewers out of the playoffs, defense – or the lack of it – felt pretty damn important to the people of Milwaukee.
How a large sample size can be misleading
If you come to the game with a background in statistics you probably believe a large sample size is always better than a small one. But what’s true over 162 games might not be true tonight and once you get to the playoffs, tonight is what you care about.
Always doing what a large-sample size suggests can lead you to some goofy conclusions.
A large sample size would indicate I don’t need to carry a jack in the trunk of my car because the vast majority of the time spent driving I don’t get a flat tire. But when I do get a flat tire I need that jack pretty damn bad.
The same goes for sacrifice bunts.
If you just had that don’t-you-know-a-successful-sacrifice-bunt-decreases-the-chances-of-scoring-a-run knee-jerk reaction, you’re wrong and in a future column I’ll explain why I think that. But for now, hold your sabermetric horses and let’s concentrate on the World Series.
During the 2019 regular season only four teams bunted less often than the Houston Astros and the only surprising thing about that is they actually bunted 10 times which seems like a lot for a team that seems to have a hardbound copy of Moneyball shoved up its collective ass.
Someone (wait…it was me) predicted that some team who had ignored the bunt all season would panic in the playoffs and ask a guy who didn’t have much practice bunting to get one down and that probably wouldn’t work so well.
In the second inning of Game 7 Robinson Chirinos attempted a bunt and popped it up. The fact that Chirinos had bunted successfully only twice in the previous two seasons and hadn’t put one down since April 26th probably didn’t work in his favor.
Ignore the bunt and it won’t be there when you need it.
Chrinos failed to move two runners up and may have cost his team a run when the next batter hit a ground ball. (I just watched the 132nd Terminator sequel and it turns out predicting the future and alternative realities is kinda tricky, but nevertheless, let’s press on.)
Had Chirinos got his bunt down and had the Astros scored another run, the score would have been 3-1 after Anthony Rendon hit his homer and maybe A.J. Hinch wouldn’t have made that disastrous decision to pull Zack Greinke – a really dumb decision based on a number; pitch count.
Analytics helped the Astros get to the World Series and analytics helped lose it once they got there.
Meanwhile, the Nationals – who bunted 48 times during the regular season and nine times in the playoffs – used that supposedly dumb tactic to help win Game 2 of the NLDS and Game 6 of the World Series.
The pendulum swings back and forth
Baseball is ever-evolving. Somebody steals 100 bases; somebody else develops a slide step. Somebody hits 70 home runs; somebody else decides to start testing for PEDs.
When analytics became a big thing every team jumped on board because nobody wanted to look like they were behind the curve of what was happening – that’s a good way to get fired.
Fans of analytics point out that every team now uses them and to a certain degree those fans are right – some teams more than others – but all teams use them to some degree.
But if the movement wants to take credit for the analytics teams that have won championships, they also need to take credit for the analytics teams that have finished in last place.
And in the past five years two teams that used tactics analytic advocates don’t like won the World Series and I’m not the only one who’s noticed.
According to a New York Daily News article written by Bill Madden, some owners are starting to turn on their analytics-heavy GMs and the front-office puppets those GMs hired to manage their teams.
Here’s that link:
Blend teams
For the 154th time I’m going to say that nobody wants to get rid of analytics or the guys who provide them and that includes me. The numbers guys have added and continue to add something to the game.
But the guys who stand on dirt have something to add as well. The smartest teams blend the two sources of information and that gives them an advantage over teams that let the shirt-and-tie guys dictate everything.
The Washington Nationals will get rings the size of gear shift knobs in part because they did some of the stuff analytics fans advocate, but didn’t ignore the old-school parts of the game while doing so.
Analytics are here to stay, but they’re no longer delivered on stone tablets.