• OtherStuff
• Footy

Interstate Advantage?

I crunched a few numbers on AFL footy from 2000 onwards looking for ways to visualise differences in performance between Victorian and Interstate clubs. Given that 2000 was the last year a Victorian club (namely Essendon) won the AFL premiership, just about everyone acknowledges that Victorian clubs are not enjoying the same success as their interstate counterparts.

This data would be even more one sided if 2000 was excluded since Essendon dominated that season so decisively.

Home and Away Ranking

This plot shows how many times in the seven years studied a position has been filled by Victorian (blue) and non-Victorian (red) clubs. Showing these side by side makes the bias pretty obvious.

The two sample t-test results below back this up

> t.test(rank ~ isVic, data=afl)

        Welch Two Sample t-test

data:  rank by isVic 
t = -3, df = 85, p-value = 0.003146
alternative hypothesis: true difference in means is not equal to 0 
 95 percent confidence interval:
 -4 -1 
sample estimates:
mean in group inter   mean in group vic 
                  7                  10 

Which means that for the period, the bias was in the range of 1-4 positions in favour of interstate teams.

This breaks out the data into individual states. Not only has Victoria faired badly against the combined interstate clubs, it has done worse than every other state individually as far as ladder rankings go.

Goal Scoring

This plot shows the frequency of goal scoring ratios (for/against) for each season. Apart from the outlier of Essendon in 2000, interstate teams consistently did better. Interestingly, the "goals for" in both groups was not all that disimilar, compared with the "goals against".

Data Source

The data was manually copied from the html at Match Scores 1897-2007 into Open Office.org Calc, saved as CSV, massaged with a short python script then analysed and plotted with R.