A group of 50 scientists sit around a table deciding how far an athlete will throw a ball.
They are experts in ballistics and aerodynamics.
Each is asked for a prediction and then the group is asked to make a group prediction that they all agree on.
The individual scientists come up with predictions ranging from 51 metres to 100 metres, in 1m intervals, no scientist predicts the same distance, so the predictions are 51m, 52m, 53m…to 100m.
What would the group prediction be? 75m?
This is a theoretical situation, but the answer is very unlikely to be 75m.
In order for the this to be the agreed group estimate, the scientist who predicted 51m, must sign on to the estimate. Are they going to do this, if they think the estimate is wrong by 50%?
The scientist who predicts 100m is much more likely to sign on to the lower figure as they will agree that the ball will travel at least 75m.
When studying deviation from a norm, it is easier to sign on to a lower deviation.
This is something to bear in mind when looking at scientific consensus on rising global temperatures and sea levels.