This past week, I was working with a pitcher. It was the end of our second week working together. We were finishing up his throwing session getting some mound work in. His throwing intensity was pretty low – considering this was the second week he had picked up a baseball in over four months. We had a stalker radar gun set up, but I wasn’t quite concerned with the number on the gun. It was simply set up to optimize feedback for how his arm was moving – something we identified in his assessment as a great opportunity for improvement. The number on the gun reflected how well – or not so well – he was able to do this.
At least, that’s how I designed it.
After a few dozen throws off the mound, you could tell something was off. He wasn’t moving really well. Instead of working on the things we had hammered throughout the week, he was starting to fall back into common compensation patterns. The arm wasn’t as clean. The body was getting sloppy. The numbers were far from great – especially considering the intensity he was exuding. You could sense his frustration.
He asked, “Why am I not throwing hard?” I paused. Once he asked this question, I realized the fault in my design. My intention for the gun was not to become an objective. It was simply feedback. When I realized my mistake, I asked, “Is your goal right now to throw hard, or is it to throw well?” He responded to throw well. So then I asked, “Why are you trying to throw hard?” He didn’t have a great answer.
I responded, “Instead of trying to chase a number, let’s chase things that are going to help you throw better.” Immediately understanding, he dropped his throwing intensity. We created a new focus and measure of success for each rep. The number on the gun was irrelevant. Our focus was going to be on the inefficiencies we identified in his evaluation. If he was throwing slower, but felt better, we were going to be happy about the rep.
Ten throws later, his velocity jumped 3 mph.
British economist Charles Goodhart became an adviser to the Bank of England in 1969, where he remained for the next 11 years. He was a renowned educator, speaker, and economic adviser. In 1975, he published a piece regarding monetary policy – the authority of a country to control short term interest rates and/or money supply to maintain stability of currency (we might need his expertise here soon).
In the piece, Goodhart said the following: “Any observed statistical regularity will tend to collapse once pressure is placed upon it for control purposes.” It has been since simplified, and is now known as Goodhart’s Law. It goes as follows:
“When a measure becomes a target, it ceases to be a good measure.”
I cannot emphasize how important his discovery is.
If we go back to my throwing situation above, we see the perfect example. A specific measure – in this case, throwing velocity – transformed from feedback into an objective. The second it did, it ceased to be effective – inspiring old compensation movement patterns. Instead of fueling the overall goal for the day (throwing well), it became a distraction to the goal (I need to throw hard right now). If you’re objective isn’t having a positive impact on performance, you don’t have a performance problem. You have misaligned incentives.
While Charles Goodhart did not come up with his theory in today’s data-fueled consumer culture, he understood the downside of data. If it’s objective, it’s tangible. It’s easier to track and grasp, making improvements – or lack of – less abstract and more concrete. This can be a really good thing. Objective feedback on velocity is a great way to grasp whether mechanical changes are having a positive or negative impact on performance. It’s especially powerful in situations where the goal isn’t to harder (wait, you’re telling me I throw harder at 75 percent than when I try to throw 100 percent?).
Here’s the problem: While velocity is a powerful predictor of pitching performance, it isn’t the sole predictor of performance. There are tons of layers – health, command, secondary stuff, pitch-ability – that ultimately determine success on the mound. The “test” pitchers must pass does not come down to a number on a radar gun. It comes down to whether they can send hitters back to the dugout. Once you aspire to a certain level, the only thing standing between you and the next is your ability to get outs. It’s no secret velocity helps. I’m not arguing it doesn’t. I simply want to bring awareness to the implications for training – which get a hair messier.
Data should never become a “target.” It’s a dangerous game. Optimizing for metrics does not optimize for performance. In reality, performance is multi-faceted. Data does not win or lose games. It simply provides feedback on the winner and the loser. What we do with it is more important. Numbers aren’t goals. They’re destinations. How we traverse towards that destination is where coaching comes into play. We need objective feedback to eliminate guess work in our interventions, but we must not become slavish to it. If we sell out to achieve metrics, we start playing a game not worth playing.
Naval Ravikant – entrepreneur and investor – said it best: “If you play stupid games, you win stupid prizes.”
Before you pull out the Stalker or Rapsodo, ask yourself: What is the goal for today’s session? Once you establish this, ask the following questions:
- What kind of feedback is tech going to provide throughout the course of the session?
- Will this feedback enhance, or subtract from the quality of the session?
Charles Goodhart warned us over 40 years ago what happens when a metric becomes the objective. If you don’t know what you want to achieve, don’t just default to what you can see on the gun.
It’s going to hurt more than it will help.