I remember hearing one story about when Clive Woodward introduced heart rate monitors to the England Rugby team. After the warm up on the first day of the players wearing these new devices Woodward called everyone in to explain the next drill. While Woodward is talking Will Greenwood won’t stop jogging on the spot, after a few minutes Woodward loses his patience and ask what the hell he is doing; “Just getting my numbers up boss”, Greenwood replied. Everyone laughed (presumably he stopped running on the spot) and life moved on. But this does raise an important point, I have seen numerous examples of players making completely unnecessary runs in training just to get their GPS figures up. All data needs to be put in context but if the players are fully aware of what we are measuring does that always change behaviour in a positive way?
Heisenberg’s Uncertainty Principle
So called after Werner Heisenberg, a German physicist, determined that our observations have an effect on the behavior of quanta. Although borrowed from particle physics this principle is saying that the mere fact of us observing something changes what we were attempting to observe. This bears out our necessity for blind and double blind experiments, eliminating as much of these unintentional consequences as possible. Update; (I have since been informed that this may be the Hawthorn affect).
While double blind experiments can be set up in other fields they don’t work very well in the real world of team sports. So does the Heisenberg principle apply to us as analysts and coaches? What I am particularly talking about is the intentional and unintentional feedback we give. Coaches have natural biases and will often reward behaviour unconsciously. When creating KPI reports or delivering feedback to players we are effectively showing them what we were measuring. While we hope that does change behaviour it may not always have the intended consequences.
Sharing KPI’s with Players?
To take one simple example from rugby; if one of our key metrics is tackles, by counting this do we always change, for better, the behaviour of our players. Instead of joining a ruck they stay out hoping to increase their tackle count. I have seen this in several cases where a management team has started measuring something and showing it to players only to find that player X stops doing something else so he can get ‘the’ number up!
How to avoid this? I’m not sure there is an easy answer. I understand the examples I have presented may be an oversimplification of events but the point still stands nonetheless. Is there a case of not sharing certain KPIs with players, perhaps they are some things you just keep in the management? We often talk about involving players in the process so they take ownership – but does that always get us the desired results?