Matter of Stats

View Original

MatterOfStats 2014 : Team Dashboard for Round 9

For this week's Team Dashboard commentary I'm going to focus on the rank correlations between the teams' ranking on various of the metrics on the Team Dashboard and their ranking on the competition ladder (using my approach to ordering teams that have played different number of games, which is to calculate competition points secured per game played).

These rank correlations are as follows:

  • Ladder Position and Scoring Shots Created per Game : +0.88
  • Ladder Position and Scoring Shots Conceded per Game : +0.71
  • Ladder Position and Own Scoring Shots Conversion Rate : +0.09
  • Ladder Position and Opponent Scoring Shots Conversion Rate : +0.36
  • Ladder Position and Q1 Performances : +0.84
  • Ladder Position and Q2 Performances : +0.59
  • Ladder Position and Q3 Performances : +0.59
  • Ladder Position and Q4 Performances : +0.68

In an historical context these correlations are generally quite low, most startlingly so for the correlation between Ladder Position and Own Scoring Shot Conversion, which at +0.09 is about as close to perfectly uncorrelated as you can get.

Six teams in particular are driving this low correlation, three from the bottom half of the Ladder:

  • Western Bulldogs who are 12th on the Ladder and 2nd on the Conversion metric
  • GWS who are 17th on the Ladder and 6th on the metric
  • Brisbane Lions who are 18th on the Ladder and 5th on the metric

... and three from the top half of the Ladder:

  • Geelong who are 3rd on the Ladder and 16th on the metric
  • Kangaroos who are 6th on the Ladder and 17th on the metric
  • Collingwood who are 7th on the Ladder and 18th on the metric

Scoring Shot creation has been the most important activity for teams so far this season (hence the +0.88 correlation with Ladder position). Converting those scoring shots has been of far less importance because winning teams have, on average, generated almost 8 more scoring shots per game than their opponents (27.8 shots versus 19.9). Using these averages we can calculate that a winning team registering 27.8 scoring shots and converting at only 50% will prevail over a losing team registering 19.9 scoring shots and converting at a much higher 77%.