Half of the 2017 AFL season is in the book, so with just 99 games of home-and-away football ahead, who is set to rise or fall as we march towards September?

This is a follow-up column to my annual look at the AFL by the numbers ahead of the season.

As a reminder, what we’re doing here is throwing narratives out the window and sticking to the cold comfort of points, margins and other salient facts. The numbers presented in these tables below are the unspoken truths of professional Australian rules football; you can only outperform the maths for so long.

Before we begin, some narrative. We are in the midst of a phenomenal season, that can not be denied. While we crossed off four teams on Monday, as we sit here heading into Round 10 the ladder is congested.

Outside of the 8-2 Adelaide Crows and hot-fire spitting Port Adelaide Power, no team has a percentage of more than 120. The ladder changes shape by the week, and with six games a week for the next three rounds that volatility will continue.

Port Adelaide, Richmond and West Coast could nab a top-four spot this weekend. Essendon could temporarily find itself in the top eight. Geelong could rise to second with a win over Adelaide. Plenty is on the line week to week. But, this time of year makes sorting signal from noise mighty difficult. So let’s get to the numbers.

Pythagorean expected wins
The original and the best, Pythagorean expected wins seek to cut through the noise of 120-odd minute parcels of action that decide winners and losers to see which teams perform best if the game were hypothetically all about points for and against.

If a team has more wins than their Pythagorean win total, they’re outperforming their underlying ability and are a candidate for second half regression, and vice versa.

And folks, one team is outperforming like no team has outperformed in the period that I have this data.

The Fremantle Dockers have outperformed their underlying points totals by a remarkable three wins through ten games of football. It’s not driven by their first two rounds either; the Dockers have been outperforming all season.

They sit somewhat precariously inside the eight – if any of Melbourne, St Kilda or Essendon win one more game than the Dockers they will fall out of the finals.

It’s a remarkable statistic. Very few teams outperform their Pythagorean win total by three wins over a full year. Fremantle have done it in two and a half months. Their 3.1 win outperformance is the largest through ten rounds in 17 seasons of football that I have this statistic. Indeed, only four teams have outperformed by more than two wins through ten rounds.

Normally when teams outperform we should expect them to come crashing back to earth. But funnily enough, three other teams who’ve smashed through the +2 Pythagorean barrier after ten rounds have pretty much maintained their level of performance.

Indeed, North Melbourne have been the kings of beating the Pythag maths over the past ten years. Their 2010 side improved their winning percentage in the second half of the season from 40 per cent to 58 per cent – and almost made the finals in the process. That wasn’t an issue for North Melbourne’s 2005 side, who rode their luck to 13 wins and a fifth placed finish.

Is that the future for Fremantle? It’s tough to say. Carlton’s early season outperformance last year gave way to an end of season fadeout. The Blues were an old, transitioning team laden with some veterans – that’s where Fremantle finds itself in 2017.

The other interesting Pythagorean stories are the underperformance of Port Adelaide and Sydney – both of whom have two or more fewer wins than their underlying potential. Unlike Fremantle’s case, undershooting a Pythagorean win total is more common.

Port Adelaide have performed at the level of a seven to eight win team through ten rounds. It’s worth remembering they’ve only played nine games in those ten rounds, rather than ten like the rest of the competition. If the Power maintain their points for and against this weekend they will break above eight Pythagorean wins – joining the Crows at the top of the table.

The Power’s rise to the upper echelons of the premiership table – it’s worth remembering that during their 2013 and 2014 run Port Adelaide finished seventh and fifth, respectively – is as surprising as Sydney’s fall. And the two teams share something else through the first ten rounds of the season.

Close wins and losses
Port Adelaide and Sydney are two of the five teams that have played in close games (games with a final margin of 12 points or less) but are yet to taste victory. That helps explain part of their underperformance against Pythagorean wins – if we credit the Power and Swans with wins from one of their two close games, their win totals look more even.

They aren’t the only luckless teams thus far. Gold Coast (0-1), Melbourne (0-2) and North Melbourne (0-3) are also winless in games decided by less than 12 points after ten rounds. Richmond find themselves one game behind the ledger in tight finishes, their wet-weather victory over the Eagles in Round 3 their lone close win.

The Tigers lost three on the trot by a combined ten points in Rounds 7, 8 and 9. In Round 7, an insufficient intent out-of-bounds free kick paid against Jayden Short was the death blow. In Round 8, they got out of jail against the Dockers in the last minute, only to lock themselves back up with an atrocious centre bounce set up in the last 30 seconds.

In Round 9, a last-gasp goal was overruled on review, and a favourable bounce of the ball on a kick down the guts by the Giants created an opportunity GWS were good enough to take.

With all these luckless teams, it holds that many other teams have had a better time of it in the close ones. The biggest beneficiary of football vagaries to date has been the GWS Giants, who are 4-0 in games decided by less than 12 points through ten rounds.

The Giants have won four of their last five games as their injury toll mounts; all four victories have been by single-digit margins. The early candidate for game of the year was first, with Toby McLean’s tough miss the last act of a tense Friday night encounter in Round 6.

After their touching up at the hands of the Saints, the Giants have won three straight by margins of three, three and eight points against the Pies, Tigers and Eagles respectively. They were due some luck: last year, the Giants won just one of five close games.

Fremantle are also yet to lose a close game in 2017, winning three from three after losing their only close game in 2016. Other teams with a clean slate include Hawthorn (2-0), Geelong (2-0, meaning the Cats have won 11 and drawn another of their past 13 close games) and Brisbane (1-0, in Round 1. It’s been a little rocky since).

A point of interest: there have been 18 games decided by less than 12 points so far in season 2017 – or just about two per weekend. That’s bang on the average over the past decade. Why’s that interesting? Because numbers are interesting.

Blow-outs
The inverse of close games are blow-outs: games where the margin of victory is greater than the average plus one standard deviation. History shows teams that are able to blow out their opponents tend to be the better teams in the competition, and conversely teams who are consistently blown out tend to be the worse teams.

This year, the price of entry to Club Blow-out (that sounded a lot better in my head) is 59 points.

Where there has been a garden variety amount of close games, the number of blow-outs recorded (15 thus far) is the fewest as a percentage of the total games played through ten rounds since Gold Coast joined the competition. In keeping with this theme, the points required to constitute a blow-out is the lowest it has been since 2009

It shows in the distribution of blow-out losses across the competition. Hawthorn and Gold Coast are the only teams to have given up more than two very large margins in their games – the Suns doing it in nine compared to the Hawks ten outings to date. Half of the competition has given up a blow-out, including four teams (Adelaide, Fremantle, Richmond and West Coast) that find themselves in the top eight.

That’s not the most interesting feature of this table. As you can see, of the 15 blow-outs to be recorded so far in season 2017, South Australian teams have handed out nine of them. Adelaide (five) and Port Adelaide (four) are the only teams to have dished out more than one kerb stomping this season – a fascinating insight into how far ahead these two teams potentially sit from the remainder of the competition.

The Western Bulldogs have continued their streak of relatively close margins. Since going on a wild streak at the tail end of 2015 (the Dogs blew out three straight opponents between Rounds 18 and 20, before being blown out themselves in Round 21), the Dogs haven’t played in a game with a margin that constitutes a blow-out. In that time, they’ve played in 13 games with margins of less than 12 points – winning eight and losing five.

Schedule change
Footscray sit 6-4 after ten rounds, with just shy of six Pythagorean wins, a 50-50 record in close games, and zeros in both the blow-out and blown-out columns of the table above. They are a rare team that has won pretty much what the underlying metrics say they should have.

Where the Dogs differ from the competition is in their schedule. To date, the Dogs have played ten opponents with an average Pythagorean win percentage of 42.1 per cent – ranked 16th in the competition (on the ‘easy’ side). It gets more difficult from here: the Dogs have an average opposition Pythagorean win percentage of 55.9 per cent over their final 12 games, ranked the fourth most difficult.

Footscray have a tough run home, but three teams have a more difficult run: West Coast (55.9 per cent), Carlton (61.1 per cent) and Collingwood (66.2 per cent). In all three cases, the team’s schedule difficulty on this metric increases by a bit to a lot over the final 12 games.

This is the case for Collingwood in particular. The Pies, 4-6 through ten games, have played the second weakest slate of games in the season to date – a Pythagorean win percentage of their opposition of 34.2 per cent. Their schedule almost doubles in average strength. Yikes.

By contrast, the winds of the fixture look set to blow Brisbane’s way, with a very tough first ten rounds (average Pythagorean win percentage of 68.7 per cent, the toughest of the season) giving way to a more middling final 12 (46.6 per cent, ranked 12th).

Similarly, Richmond’s final dozen games are very kind in the scheme of things, shifting from a win percentage of 48.2 per cent (ranked 11th) to 38.3 per cent (ranked 18th). Port Adelaide also have a charmed run (38.9 per cent in the last 12, ranked 17th).

The takeaways
As usual, that’s a lot of numbers with only an itty bitty bit of analysis. So what?

Port Adelaide look set to come home with a wet sail; they’ve underperformed their Pythagorean wins by 2.5, have lost both of their close games, have blown out four opponents (and not been blown out themselves) and have the second weakest set of opponents in their fixture in the final dozen games.

There are less abstract reasons for the Power to feel confident heading into the second half of the year. Their playing stocks are fit, their tactics effective against most opponents, and their ability to score and stop opponents from scoring per minute of possession are both in the top four in the competition.

We all missed Port Adelaide’s surge up the ladder; it’s hard to miss them now. Ditto Richmond, although thinking fans knew 2017 looked like Tiger time.

Sydney could expect to perform better given their underperformance against Pythagorean wins and relatively meagre slate. The question is do they want to?

In an even season, the Swans are still in the hunt for a spot in the finals if they make a concerted effort and shake off their flaky form. North Melbourne are in a similar boat (three close losses, one blow-out and no times blown out, a fixture that eases up in the second half), albeit they have only underperformed their Pythagorean win total by half a win.

Carlton and Collingwood could be in for a tough ride based on their fixtures, which ramp up significantly in the second half of the year. The numbers say Collingwood have performed at the level of a 0.500 side in the face of the second weakest fixture in the competition. Will they hold up as their average opponent doubles in strength?

Fremantle will also likely come under pressure; outperforming the maths by as much as they have through ten games doesn’t look sustainable. In their favour is a relatively kind slate of games. Greater Western Sydney have also outperformed, but their run of close wins in the face of a cataclysmic injury list surely can’t continue.

Whatever happens, the next 12 weeks loom as one of the most intriguing races for spots in the finals this decade. Where we were done and dusted after Round 10 last season, with only the seedings to be decided.

The entire ladder is in a state of flux in 2017. The promise of an unpredictable season was writ large in January and February. It has been delivered.