0-3 and staring down the barrel


It’s the nature of sports reporting a common garden individual match , that could simply be enjoyed for its own sake, or by supporters of the competing teams, has to be built up to mean something more. This could be local bragging rights, revenge for a past defeat or some other real or imagined past incident, a (usually confected) rivalry, which can finally be resolved.

In the AFL, even better than any of these is to broaden the implications of the match – win today, or “stare down the barrel” (few other metaphors are authorised) of missing the finals. It’s particularly prevalent – in its irrational form – early in the season, when the loss of a few games has media vultures circling.

Will Brodie’s article is typical, and he even appears to have a decent statistical case. Articles so often tell you how seldom something happens, but rarely how often the opportunity has arisen for it to happen. Brodie does that, letting us know the rarity of teams making the finals when starting 1-3: 20 of out 70 times.

This post will concern itself with even worse starts – teams which lose the first three games. Over the last 50 years, only 8 of 108 teams have gone from there to the finals.

But, it could be argued (I’m about to) that many of these instances are trivial. When Gold Coast and GWS started 0-3 in 2012 and didn’t make the finals, no-one said, “See? It’s so important to make a good start.” And no-one writes the “staring down the barrel” articles about them, either. Barrel articles are only interesting when they’re about “good” teams – and “good” teams get to 0-3 much less often than “not good” teams. The real statement that writers are trying to make is that a good team experiencing a bad start is unlikely to make the finals.

This blog post isn’t going to get far without defining “good”. For simplicity, let’s call teams that made the finals the previous year, “good”. Once you separate the good teams from the not good, you end up with a different picture of the difficulty of making the finals.

But is that the end the discussion? Not quite. There might be quite some distance between the success percentages of the good and not good teams, but is it significant? Based on the sample sizes, each is 1.76 standard deviations away from a probability of 0.064 (6.4%). 1.76 standard deviations equates to around a 96% confidence that each is as high as, or as low as, the point at which they met in the middle: 0.064. The probability of both of these unlikely events occurring simultaneously is 0.04 x 0.04 = 0.0016, so we can be pretty confident that good teams with slow starts have a better chance of making the finals than not good teams – and should really be considered separately.

So, next time you hear a commentator say, “a loss would leave them at 0-3… and we all know how difficult it is to make the finals from there”, you’ll have a wide range of smarty-pants replies available – including something about the unusefully small sample size of 22 for “good teams”. Not to mention the nuances that await teasing out – over the years, there have been so many different combinations of numbers of teams in the competition, and numbers of finalists.

The fun thing about doing this research is remembering – and discovering – all those teams that have overcome the barrel-staring experience. But the historical asterix in Brodie’s article, North Melbourne, is neither a great story nor drawn from far enough back in history. What about the 1975 Kangaroos who lost their first four, fought back to third at 14-8, and won the grand final – their first – by 9 goals over Hawthorn? Thus “proving” that the premiership can be won from 0-4.

Actually, it can be won from 0-16. But that’s a discussion for the next post.


Click here to like, or tell me whether this is genius or gibberish

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s