BIG STICKS IDIOT (with stats!)

This morning, the Hawks are favorites, at $4, for next years flag. I’m sure they’ll find that thrilling news as they look back on The One That Got Away. It seems like a recurring theme in Grand Final history – a team that by rights – and by most stats – should have it in the bag yet somehow wind up with heads in hands come the final siren.

Yesterday’s game was one where Sydney earned every plaudit it gets this morning, grinding out a gripping win against a super-talented, highly-favoured opponent, all the while carrying injuries across what seemed like half the team. Goodes and Mumford probably wouldn’t have played after the first quarter in a normal game, and Ted Richards battled valiantly against Buddy Franklin on a dodgy ankle.

Yet, the real story is one of missed opportunity. Missed goals – gettable goals – seem to be what cost Hawthorn a premiership in 2012. Franklin kicked 3.4 with one out on the full, two of those shots at least were easily kickable. This is not to blame Franklin specifically. But. The effect of poor kicking surely was that Sydney hung on in patches where Hawthorn was on a tear. Ordinarily, scoreboard pressure in Grand Finals is enough to prevent too many comebacks. But for Sydney, seeing the tail end of a Hawks blitz from only two or three goals behind was probably enough to keep their tails up.

This got me to thinking about accuracy. I could recall, off the top of my head, at least two other Grand Finals in the past 15 years or so that saw a team kick themselves out of the game they should have won; Geelong in 2008, kicking 11.23 to Hawthorn’s 18.7 (memorable for Cam Mooney missing from 10 metres out), and Adelaide’s 15.15 to North Melbourne’s 8.22 in 1998, when North was well up at half time.

Basically, I thought, you wouldn’t win if you didn’t kick straight.

I decided a lazy Sunday morning would be a good time to crunch the numbers in greater degree, so I dug up the results of every VFL/AFL Grand Final, and tracked accuracy versus victory.

Before I dive in, a disclaimer; I’m no scientician. I can manipulate an IF function, and I know my way around a spreadsheet, but I can’t remember a thing from Econometrics 1001 in 1999. This is a rudimentary analysis, and if any number nerds out there want to crunch harder, please be my guest; here’s the raw data.

I tracked two simple pieces of data; accuracy of winning and losing team, as well as the number of shots on goal. The short story; taking more shots on goal is far more important to winning than kicking straight.

Of the 115 Grand Finals since 1900 (including draws), only 18 (16%) were won by the team with fewer scoring shots, while 44 (38%) were won by the team that kicked less accurately.

The results actually put the lie to my unscientific theory. Kicking straight doesn’t seem to matter. Also, these results are including those games where it’s certain that accuracy really had no bearing because it was so one-sided. What happens if we look only at games that were decided by three goals or less?

Even worse. Of the 44 games that had a tight result, nearly half (21, or 48%) were won by the team that doesn’t kick straight. Fifteen, or 34%, were won by the team with fewer shots.

Of course, this is not taking into account causality. There is clearly no great correlation between kicking straight and winning flags. Indeed, inaccurate kicking may simply mean that your opponent is applying more pressure in the forward line. But if you’re in the forward line twice as often as your opponent, it shouldn’t matter too much.

Even though goals are worth six times a behind, taking sufficient shots can easily override inaccuracy. The biggest difference in accuracy in a ‘close’ GF was the 1948 Melbourne vs Essendon game, in which the Dees scored 10.9 to the Dons’ 7.27 (!). Melbourne’s accuracy was more than 2.5 times greater than Essendon.

The result? A draw.

Essendon took enough shots that in the end they only scored three fewer goals than their opponents, despite shooting at about 20% accuracy.  This story pops up again and again.

It seems, in the end, that we have a bias to remember the close ones that teams lose from inaccuracy, but they aren’t in any way unusual. If you fail to kick straight, it basically doesn’t correlate at all with the margin you win or lose by. Shots on goal, however, correlates quite closely.

I guess, then, in the absence of any more detailed (or accurate) numerical analysis, the takeaway from this is not “kick straight” but “kick often”.

Again, if someone wants to take the numbers and offer a better analysis, please do.


So I graphed something. Below is the accuracy differential versus the score differential of the winners of GFs. That is, how much more accurate were you as a proportion of your opponent’s accuracy, versus what percentage of their score did you score?

As you can see, there is no real correlation to speak of.