I was at a football match last night. At the end of the game I commented to my son how poorly I thought one player had performed.

A mix of bad passes, mistakes and poor decision making contributed to my assessment. A chap walking in front of us suggested the opposite Andy that the player had been OK and it was the players around him who hadn’t performed, making his showing seem worse than it was.

As part of my work into evaluation I’m looking at bias. I’ve found over a dozen biases which we can unconsciously fall into in review and assessment.

The language this guy used suggested he was biased in his response. In the cold light of day I looked back at the statistics of the player’s performance and it seems I was right.

Partly.

The data was unequivocal. The player concerned lost the ball more times than any other player on the pitch in the time he was involved.

However, the context he was playing in changed throughout the game and some mistakes weren’t as significant as others when looking at the game as a whole.

Take care when you’re looking at measurement. Data without context is just that, data.

Understand it to make it information.

Understand it in context to make it knowledge.