Scheer’s Law – An Observation Regarding Metrics
When I worked as an internal Scrum consultant for a San Diego biotech, I had the pleasure of working with one of the most talented and creative programmers I have ever met – Tim Scheer. Tim was part of our team Scrum transformation team and he was the person on our team who represented the hacker ethos – a bit irreverent, very pragmatic and just wanted to help people write great code.
One day, we were talking about metrics and what would be the best metrics to measure the health of the Scrum Teams. A very common concern when working in a big organization with many active Scrum Teams. We must have talked about this for an hour and then finally Tim said, “I don’t care what metric you use. Give me any metric and I will make it my mission to figure out how to game it so I look good. I don’t care how long it takes, I’ll figure it out.”
This raw insight from Tim helped me understand that metrics are just a game to programmers and no matter how clever of a metric you devise, they will find a way to subvert it. It is in their nature to do this not because programmers are bad people, but good programmers are very, very smart and want to take things apart to understand how they work. It is was makes them good engineers and what makes metrics ephemeral.
So here is Scheer’s Law.
“It is only a matter of time before a metric will be gamed. Once a metric has been gamed, you can no longer rely on the data it provides and it must be retired.”