I Hate Analytics. Why Don’t You?
One of the great reliefs of my life after retirement is that I have not lost my capacity to hate stuff. I have always considered it a strength to feel a strong negative passion when it was required.
These days, it is required more than ever.
For example, I hate analytics.
I am not alone, I know, but few younger people harbor the powerful anger that I feel every time I experience the deep dive into statistics that is drowning sports today.
Are these fans conditioned to accept robotic decisions by managers and coaches just because, well, the computer world of today overwhelms pretty much everything, so why not basic sports logic, too?
At the risk of seeming obsessive (which I am anyway), I cannot shake the feeling I had in the first playoff game last fall, when Phillies manager Rob Thomson removed an unhittable Zack Wheeler after seven innings and a 1-0 lead against the Mets. The ace had committed an unpardonable offense. He had surpassed 100 pitches.
The bullpen blew the game, of course, and the only thing that shocked me afterwards was the fact that pretty much no one blamed the manager for the loss, or for the subsequent defeat in the series. (Except my emailers, an unofficial club of dinosaurs like me.)
So anyway, I was watching the Phillies slogging through their worst slump of the season last week, hoping (stupidly) that the manager had learned his lesson after blowing an entire season last year with his unshakable devotion to analytics.
This time it was just a regular season game against the Cubs, and this time Wheeler had given up one run, but the number in the right-hand corner of the screen ominously reported 103 pitches.
Sure enough, Wheeler was done for the day, as was the 2-1 lead shortly after the manager’s mindless, programmed move. The Phils ended up winning the game in extra innings, but not because of analytics. In fact, they won because Thomson — for one brief, shocking moment — cast aside his IPad and notebook and went old-school.
The record will show that the Phillies successfully bunted on two straight plate appearances, and the Cubs were unable to handle either play.
When was the last time the Phillies bunted twice in a row?
I can’t say for sure, but my best bet is that it was before analytics smothered an effective ploy that had been part of the game for 150 years.
You see, bunting is considered dumb by stat nerds because it often sacrifices an out for moving a runner up. Analytics zealots believe you should never sacrifice an out.
They must have recoiled at the sight of two bunts in a row.
And they must have short-circuited when the bunts led to a win.
Of course, it goes beyond one play in one game for me – well beyond it. I hate analytics because of the notion that somehow a computer can be programmed to think more effectively than a human.
Rob Thomson has been in organized baseball for more than 40 years. He was trained to know there are times when a bunt is the right move. But, with very few exceptions, he stifles his own instincts in favor of a computer program.
Shame on him for that. He should know better. He does know better.
The other analytics-inspired issue (among many lesser indignities) that drives me bonkers is the silly idea that injuries can be prevented by load management.
That is the term used mostly in the NBA, but it applies to all major sports now. The idea is that by playing less, a star can avoid long absences caused by all kinds of physical calamities.
This is exactly why Thomson continues to hold Wheeler to strict parameters, including the biggest games in the playoffs. (The magic number is 100 pitches.) It’s also why the Eagles use two primary running backs instead of one. (The magic number there is 25 touches.)
And, in basketball, it’s why the Sixers have been able to avoid injuries so impressively in recent years.
Oh, hold on. Time out. My bad.
Actually, the Sixers were more injured than any team in the NBA last season, losing their two highest-paid players, Joel Embiid and Paul George, for a combined 104 games – and this despite a maniacal commitment to load management.
Embiid no longer plays back-to-back games because the tactic has worked so brilliantly in his career. After all, he missed only 400 of the first 846 games spread over the past decade by sitting out games while still healthy.
In contrast, let’s look at another seven-footer with major Philly ties – the greatest NBA player ever, Wilt Chamberlain. Prior to the concept of load management, Wilt played 80 or more games in nine of his 14 seasons, and all 82 in four of them. One season (1961-62), he played in every minute of every game except for the final eight minutes of one contest after he was ejected.
Chamberlain averaged 75 games a season. Embiid averages 43. Wilt missed one third of the games in his entire career that Embiid has missed already in five fewer seasons. So much for modern training methods and all of the other amenities of the 21st century.
Granted, Wilt and Joel may be outliers, but still it’s hard to deny the obvious fact that no one can ever predict who’s getting hurt, or why. No computer program is going to anticipate when a player twists the wrong way or crashes into a defender.
Durability is not subject to the simplicity of numbers. If anything, there is evidence that the more you play, the more likely you are to stay healthy. (I believe there were fewer injuries a generation ago. If you were watching sports in the 1970s and 80s, you agree.)
One thing is certain. There is no evidence – none – that load management has helped preserved the health of anyone.
Can you name one study supporting that claim?
No. Because there are none.
You see, that’s the real hustle behind analytics. Statistical analysis is supposed to provide a blueprint for many of the major pitfalls in sports, but then, when confronted with the reality of the situation, the stat nerds have no actual numbers to support their philosophy.
Are there fewer injuries now than before analytics?
I would bet my right hamstring and left femur that the answer is no.
In fact, my own analytics say there’s at least a 74.3 percent chance that I’m right.