Predictions Based on Won/Lost Stats Alone Problematic of AFL 2017

  • Posted By:

As the disclaimer says, “Past performance is not necessarily predictive of future results.”

That disclaimer is not only apropos when it comes to financial portfolios. It is applicable to all manner of scenarios, including predictions of sports teams’ prospects, but since it is not all that long until the 2017 AFL season bounces, now would seem as good a time as any to have a look back, and use that historical data to make some assumptions regarding which clubs will rise, which will remain stable and which may fall.

Looking strictly at wins and losses often neglects an important factor, the winning margin. MLB stats expert Bill James understood this, and used it to predict which teams “should have” won based on scoring.

The same analogy could be applied to the AFL.

If the Adelaide Crows, one of the top teams from 2016, were to play against, say, the Essendon Bombers, and beat the Dons by a narrow margin only, that would imply that not all was shipshape and Bristol fashion with the Crows, whereas, if they were to play one of the elite teams, the Swans perhaps, and beat them by 40 or more, the implication would be that the Crows’ potential was far greater than understood.