Half-baked PyEx: Where are we at in season 2014?

We’re effectively at the ‘one third’ mark of AFL season 2014, though mathematically we’re 9/22 of the way through.

We’ve seen enough of every team now to make some reasonably sound conclusions on who’s got it and who hasn’t.

Before we get into the numbers, think about this:

Port Adelaide’s only loss came against North Melbourne;
North Melbourne were beaten by Sydney;
Sydney were beaten by Collingwood;
Collingwood were beaten by Fremantle;
Fremantle were beaten by Hawthorn
Hawthorn were beaten by Geelong; and,
Geelong were beaten by Port Adelaide.

Seven of the Round 10, 2014 top-eight sides are listed in that little passage above, and each side has only played nine games. Fair to say it’s been a pretty excellent season of footy so far, all things considered.

Now, on to the serious stuff. Earlier this year, I introduced a statistic called “Pythagorean Expected Wins”, or PyEx for short. You can read all about it here.

The stat attempts to cut through the various vagaries of footy and determine where a team ‘should’ be given their offensive and defensive output. It’s used a great deal – and effectively – in American sports, and I’ve showed that it was pretty good at sorting the best (and worst) from the rest in the AFL.

I made a commitment to take a look at this for 2014 throughout the year, and given we’re now nine games in the stat should start to be telling.

So what does PyEx have to say about Season 2014 thus far?

Once again we haven’t fully nailed the ladder, but in terms of overall wins and losses, PyEx has got it pretty well correct.

PyEx says that our top four, based purely on offensive and defensive output, should be Hawthorn, Port Adelaide, Geelong and Sydney.

Some of the other notable differences are:

• PyEx reckons West Coast are worthy of a top eight spot, and should be ahead of both North Melbourne and Collingwood. I disagree with this assessment.

From a more in-depth look at the figures, it’s clear that blowout results against GWS and St Kilda are biasing West Coast’s offensive output numbers up. Offsetting that to some degree is the fact that the Eagles have lost both of their close games.

• There are currently seven of 18 teams that are at or around the 4.5-win mark (+/-1 win), including Collingwood, Fremantle, North Melbourne, West Coast, Adelaide, Essendon and Richmond. The five teams above them all are all at or above the six-win mark to this point in the season.

This tells me that it’s likely five of the current top eight will still be there come September, and that these seven teams will be jostling it out for the final three places. At the moment, PyEx has Fremantle, North and West Coast in there.

• Hawthorn has ‘lost’ 2.3 wins so far this season, according to PyEx. Their output suggests they are a clear eight-win team.

Taking a bit of a deeper look yields some interesting numbers:
o Total Winning Margin: 420 points in six games (70 ppg)
o Total Losing Margin: 52 points in three games (17.3 ppg)
o Differential: +472 points (second, behind West Coast)

Another interesting point to raise about Hawthorn is who they’ve lost to – Port Adelaide, Sydney and Geelong, all teams that are around them on the ladder. Losing sucks, but losing to teams that are in your peer group sucks more.

• Richmond have cost themselves 1.2 wins so far this year. This is also unduly influenced by a blowout result (this round’s pumping of GWS), but the Tigers have lost by 2 points (Round 3 v the Bulldogs), 5 points (Round 7 v Geelong), 17 points (Round 8 v Melbourne) and 18 points (Round 1 v Gold Coast).

So as average as they’ve been, they are perhaps playing slightly better footy than the wins column indicates.

Do you think the PyEx Top Four (Hawthorn, Port, Geelong and Sydney) a better indication of where AFL Season 2014 is at? Or are the numbers presented above showing us a different picture again?

And what are your predictions for this second third of the season?