B
Bill James
Guest
The table below is an crude analysis of the stats provided by Pro-Stats
http://www.pro-stats.com.au/psw/web/team_rankings?yr=2006&ts=DA&sr=0
Prostats present "Team vs Opponent Averages" across a range of commonly used and quoted football statistics. Rather than look at each teams season totals compared to other clubs, the Prostats analysis looks at each team compared to the teams they played. In other words it is a way of looking at what the top teams do better/more often/least often than their opponents when they play them.
What I have done below is translate each of those stats into a 1-16 ranking. In order to see which stat rankings was the best predictor of the ladder the I have added up the difference between ladder and stat ranking for each club for each category and that number is at the bottom of the table. Simplistically the lower the correlation number the better the stat difference correlates with the ladder.
It is a crude statistical analysis but arguably draws some interesting conclusion.
Surprisingly emotive stat differences like 1%, clearances, first posessions, errors and tackles are the poorest predictor of ladder position. (Could explain why TW is not so worried about errors, 1% and tackles where we finished 13-14 and 15th respectively)
Somewhat surprisingly inside 50's, uncontested posessions, disposals, and long kicks are the stat-differences that are, in that order, the best predictors of ladder position. (RFC was ranked between 7-12 on these key stat differences)
Kicks, marks, contested marks, handballs, contested posessions, hitouts are stat differences that are average predictors of ladder position.
Free kicks for and against don't seem to have a major bearing on ladder position however its interesting to not that frees against are a better indicator than frees for.
Anyway make of it what you will for next years game plan.
http://www.pro-stats.com.au/psw/web/team_rankings?yr=2006&ts=DA&sr=0
Prostats present "Team vs Opponent Averages" across a range of commonly used and quoted football statistics. Rather than look at each teams season totals compared to other clubs, the Prostats analysis looks at each team compared to the teams they played. In other words it is a way of looking at what the top teams do better/more often/least often than their opponents when they play them.
What I have done below is translate each of those stats into a 1-16 ranking. In order to see which stat rankings was the best predictor of the ladder the I have added up the difference between ladder and stat ranking for each club for each category and that number is at the bottom of the table. Simplistically the lower the correlation number the better the stat difference correlates with the ladder.
It is a crude statistical analysis but arguably draws some interesting conclusion.
Surprisingly emotive stat differences like 1%, clearances, first posessions, errors and tackles are the poorest predictor of ladder position. (Could explain why TW is not so worried about errors, 1% and tackles where we finished 13-14 and 15th respectively)
Somewhat surprisingly inside 50's, uncontested posessions, disposals, and long kicks are the stat-differences that are, in that order, the best predictors of ladder position. (RFC was ranked between 7-12 on these key stat differences)
Kicks, marks, contested marks, handballs, contested posessions, hitouts are stat differences that are average predictors of ladder position.
Free kicks for and against don't seem to have a major bearing on ladder position however its interesting to not that frees against are a better indicator than frees for.
Anyway make of it what you will for next years game plan.
| Team | Disp | Kck | KLg | Mrk | CMk | Hbl | Tkl | F/F | F/A | H/O | 1st | Cle | In50 | Rb50 | CPs | UCPs | 1`% | Err | Bnc |
| West Coast | 1 | 8 | 2 | 16 | 3 | 1 | 13 | 3 | 14 | 1 | 6 | 2 | 4 | 13 | 3 | 1 | 11 | 9 | 8 |
| Adelaide | 8 | 7 | 4 | 11 | 5 | 8 | 6 | 11 | 5 | 4 | 11 | 8 | 1 | 16 | 8 | 6 | 9 | 11 | 4 |
| Fremantle | 4 | 4 | 12 | 1 | 2 | 5 | 12 | 14 | 3 | 11 | 12 | 15 | 5 | 10 | 4 | 3 | 13 | 10 | 10 |
| Sydney | 2 | 2 | 3 | 3 | 7 | 7 | 11 | 16 | 1 | 7 | 2 | 3 | 2 | 15 | 12 | 2 | 15 | 1 | 5 |
| Collingwood | 3 | 1 | 10 | 2 | 11 | 12 | 7 | 5 | 13 | 14 | 8 | 12 | 6 | 8 | 11 | 5 | 5 | 8 | 6 |
| St Kilda | 13 | 9 | 1 | 7 | 10 | 14 | 10 | 10 | 4 | 15 | 16 | 13 | 3 | 14 | 9 | 9 | 7 | 2 | 2 |
| Melbourne | 6 | 3 | 6 | 4 | 8 | 11 | 2 | 9 | 10 | 10 | 14 | 6 | 9 | 2 | 16 | 7 | 3 | 12 | 9 |
| Western Bulldogs | 9 | 13 | 13 | 10 | 14 | 4 | 1 | 4 | 12 | 6 | 15 | 16 | 7 | 6 | 13 | 13 | 2 | 15 | 1 |
| Richmond | 12 | 11 | 7 | 8 | 4 | 10 | 15 | 13 | 7 | 8 | 9 | 11 | 10 | 5 | 2 | 11 | 14 | 13 | 13 |
| Geelong | 5 | 6 | 5 | 9 | 1 | 3 | 14 | 7 | 11 | 9 | 10 | 10 | 8 | 11 | 1 | 8 | 12 | 3 | 16 |
| Hawthorn | 7 | 15 | 9 | 12 | 16 | 2 | 16 | 15 | 2 | 2 | 1 | 5 | 15 | 3 | 10 | 4 | 16 | 5 | 14 |
| Port Adelaide | 14 | 12 | 11 | 13 | 15 | 9 | 9 | 12 | 6 | 3 | 5 | 1 | 12 | 4 | 15 | 10 | 8 | 4 | 11 |
| Brisbane | 16 | 16 | 16 | 14 | 12 | 13 | 3 | 8 | 9 | 5 | 3 | 9 | 14 | 12 | 14 | 16 | 1 | 14 | 3 |
| Kangaroos | 11 | 5 | 8 | 5 | 6 | 15 | 4 | 1 | 16 | 13 | 4 | 4 | 11 | 7 | 5 | 14 | 4 | 16 | 12 |
| Essendon | 10 | 14 | 15 | 15 | 13 | 6 | 8 | 6 | 8 | 16 | 7 | 7 | 13 | 9 | 6 | 12 | 10 | 7 | 15 |
| Carlton | 15 | 10 | 14 | 6 | 9 | 16 | 5 | 2 | 15 | 12 | 13 | 14 | 16 | 1 | 7 | 15 | 6 | 6 | 7 |
| Correlation to Ladder | 46 | 60 | 50 | 60 | 66 | 64 | 106 | 96 | 68 | 70 | 100 | 86 | 28 | 112 | 88 | 34 | 102 | 92 | 66 |
