Saturday, 14 November 2015


With the 2016 fixture released last week. the season has come for pundits to assess which team got a good deal or not. Others already have, so its also fitting the FMI also join in the (alleged) fun of draw analysis.

Its all part of the FMI experience, with draw assessments now into their fifth fabulous season.
(See also assessments for seasons 2015, 2014, 2013, and 2012).


As per last year, the AFL Fixture [PDF] was release under their 6 / 6 / 6 rule of having (as best as possible*) the top 6 play each other home and away, with the bottom six and middle six also playing amongst themselves as well.

And while they choose to call the draw fairer, and in a way they are right**, its handicapped. As per last year, if Stevo gets it, then 99% of the footy world should too.

Last year, under the weighted ruling, and referencing the FMI rankings of teams, the AFL basically got the balance right according to its system.
Doesn't make it 'fair' of course, unless you consider handicapping teams fair.

* Where 'as best as possible' means the AFL can still wriggle out of locking themselves into positions where derbys and showdowns could only happen once.

** Handicapping makes things fairer to a point, by drawing better athletes (etc) back to a level of competition with others less skilled. But that is not what is in play here. Hawthorn dont get to play for only three quarters while Carlton get four. This is handicapping by setting harder bars for some, and not others.
The other logic behind the 'fair' argument is teams from each group with similar abilities playing against each other will make for better games on the day. Which is right, but also code for 'a better product on game day / on TV' but it could be argued not 'fair' across the whole competition.

What is fair competition wide is for each team to have the opportunity at playing other teams at similar levels to others, as proposed in this blog post.
Fair is giving the little guy a crack at the best, an opportunity to promote themselves in prime time, on the big stages.
This fixture remains everything but fair.


As always, there are other assessments of the draw out there, with other approaches to determining balance or otherwise.

As per last year, this years analysis by Troy Wheatley and based on his Power Rankings warrants discussion. Firstly as his principles are grounded to the same logic as the FMI analysis, with the measures based on his own independent rankings system

And also because of the outcomes. In a table that lists easiest draws at the top, he has only three of the AFL's bottom six having the easiest draws. Further curiosity comes from finding West Coast deemed the 7th best draw, just behind the Western Bulldogs. The Grand Finalists would be delighted with that outcome!
At the other end of the scale, Melbourne and Brisbane both are determined to have had the top five hardest draws.

With those oddities bobbing up, those results are counter to the effect the AFL was after.

Also weighing in with a fixture assessment for 2016 are the team over at "Hurling People Now" with various methods to determine the so called 'strength of schedule'.
Over several iterations of measures, they determine the AFL has, on the whole, achieved the objective of splitting the fixture into its three constituent levels by the 6 / 6 / 6 weighting rule.

What they do find of not though is some consistent anomalies, such as a mid ranked GWS being allocated one of the tougher draws over various measures. Also there is a few instances of the defeated Grand finalist faring well from the draw as well.

All in all though, a degree of confirmation that the AFL has achieved its objective.


As regular as night divides the day, Rohan Connolly runs his slide rule* over the fixture and lists who got what from Santa Gillon.

The Connolly method is an interesting one, given the detail below. Weightings are applied either by points allocated by ladder position, or 'arbitrarily', such as the points around 6 day breaks etc. It is odd to use reverse ladder points so simply as has been done, given that the 2015 was clearly fractured. The nett difference between Collingwood (12th) and Melbourne (13th) is a single point, yet the Pies won three more games than Melbourne and the trailing other 6 clubs.
This makes the pure ladder position for points a bit of a skewed starting position.

The detail around the points system is below, and Rohan's final analysis lies in this link, because this year the punkhawallahs down at Age Central decided to make the results table soo big that it couldnt be screen-capped effectively.

The fall-out from Rohan's analysis is that there are 7 of last years top 8 teams listed with the worst draws, headed by North Melbourne just ahead of Hawthorn.

The bottom six fare worse though, as some of them get pushed up the scale of hard draws by the encroachment of Geelong and Port Adelaide, who nudge in ahead of Brisbane and Melbourne.

No excuses though for Carlton and Essendon though by the Connolly Method, with both former 'powerhouses' handed a nice bunk-up. A quick return to finals action may not be the right thing to help these relics of the old VFL way reform themselves for the ruthless business of this AFL system.

Geelong also get a decent crack at 2016 by this method, which is a point well documented in the article.

Effectively the Connolly Method suggests the AFL went quite close to achieving its objective with the fixture... top 6 teams in the hard basket, and bottom 6 teams with an easier run, but not exactly organised properly to achieve that.

* wiki will tell you young people what a slide rule is... and no, there is no inference to Mr Connolly 'being old'.


Last years assessment of the 2015 draw using the FMI rankings produced a real parity between the AFL's objectives of lower teams and higher teams playing against each other respectively. LAst seasons draw had the top 6 of 2014 listed in the top 6 hardest draw slots, and of the bottom 6, only Brisbane did not slot into the 6 easiest draw positions.

The assessment was that the handicapping worked...
"...all in all, it looks like the AFL have achieved their outcome, so ... well done!"

So to look at the 2016 draw, its  a matter of applying the FMI rankings to the fixture and calcualting the Tough Draw Index based on looking at the following inputs.
- the opposition strength,
- any home/away bias, and
- expectations of a 'win' for each game,
- an expected end of year rank
- an expected end of season ranking for a perfectly even draw, and the
- difference between those two ranking weights.
Working through the calculations yields the table below;
1West Coast2141623401-1004.37
4North Melbourne41245245361003.00
8W Bulldogs/FFC81149229701001.17
12Port Adelaide9122122380-200-2.38
14Brisbane Lions1778323378-100-2.66
17St Kilda1484723854100-3.50
18Gold Coast16837236640-4.53

And what do you know... hey presto, objective achieved (bar one outlier).
Richmond, who finished 5th after the Home and Away season, but get shuffled down to 7th after off-season shenanigans, draw the short straw and land up with the 3rd toughest draw.

Otherwise its a fixture to the script, concocted well in a masterpiece of fitting the teams into a handicapped format according to the FMI rankings.
KPI met, go and get yourselves a milkshake.
(PS, thanks for following the blog and studying the numbers, guys... *nudge wink*).

Other than Richmond, who are ranked 3rd after finishing 7th (+4), and Adelaide who have the 2nd hardest draw after a 6th placed finish (also +4), a few teams should feel lucky to benefit from the Fixture. Such as Fremantle, (minor Premiers and 3rd place after 2015), record the 7th hardest fixture (-4).
Hawthorn also. the three-peaters are slotted into the 5th hardest fixture (-4). Set up for another tilt? Look out Collingwood... Hodgey and co are gunning for your record.

Now, also your attention should be drawn to 2 other clubs who should be happy as Larry with the outcome. There are two mid-6 teams with easier draws than the others in the same pool, and easier by a long way.
Port Adelaide and Geelong (as also identified by the Conolly Method) are more closely aligned to the bottom 6 teams' fixture than any in their cohort.
No excuses now Power, Cats.

As per last year, the KPI seems to have been met and met well this year. A few exceptions to the rule, but a decent synergy between the objectives and the outcomes.

Which is all well and good... but its not strictly 'fair'.

Wednesday, 28 October 2015


The AFL is due to release their 2016 schedule this week.
And while they claim it is a fair system, it is actually the opposite of that.

The fairness they talk of is in pairing similar strength teams with each other, such as top 6 against top 6, bottom 6 against bottom 6, etc.

This may create a more even result on the day of the match, but it is not a 'fair' when teams play harder (and easier) fixtures over the course of the season. It handicaps the better teams and provides a bunk-up for the lower teams.

What is 'fair' is that all teams play a mixture of ranked opponents over the course of the season. A few seasons back the FMI solution was tabled on the blog to that effect. The scheduled matches should be based on a distributed system where each team plays others from across the spectrum of the ladder. (The blog post went further advocating for conferences, but the basic scheduling principle can be extracted from that easily and applied simply).

It is also done in such a way so that determining the schedule is via a clear and transparent method, as listed here.
Take the the 18 teams and split them as per
Set 1:   1st, 6th, 7th, 12th, 13th and 18th.
Set 2:   2nd, 5th, 8th, 11th, 14th and 17th.
Set 3:   3rd, 4th, 9th, 10th, 15th and 16th.
   *  note: add the ladder positions together for each conference for totals of 57 in each.

So the three sets of teams to play each other twice (one home and once away) are easily determined by anyone in the AFL diaspora simply by looking at the distribution list above and checking off teams accordingly.

The other games of the season are filled by paying other teams once, with the option to play home or away on a case by case basis to ensure teams play 11 home, and 11 away.

Note also, that the ladder at the end of the home and away series from 2015 is used, as this is the most balanced outcome of the year.

Adjustments to ladder positions from the finals series are not included as (obviously) not all teams have the opportunity to adjust their ladder position.

So the 2016 schedule should be set to teams in the groups below playing each other twice..
1Fremantle2West Coast3Hawthorn
6W Bulldogs/FFC5Richmond4Sydney/SMFC
7Adelaide8North Melbourne9Port Adelaide
13Melbourne14St Kilda15Essendon
18Carlton17Brisbane Lions16Gold Coast

Observant readers will note there is no double up of the Western Derby, SA Showdown or NWS or Queensland derbies as well. It is just the luck of the draw that this system spits out that combination this year. Equally, the only double ups for Melbourne 'big club blockbuster' games sit with Collingwood and Carlton or Hawthorn and Geelong. Again, luck of the draw.

And it is fully known here that SA and WA derbies are important for "giving the local fans that extra game to attend".
If your lucky enough to get a ticket as an 'away' fan, of course. (Hint, the extra local game benefits maybe 1,000 to 3,000 people).

So to look at a few further hypotheticals, let's examine possible draws for a team from each group.
First a team from the lower end of the 2015 table.
1Fremantle2West Coast1Fremantle3Hawthorn
6W Bulldogs/FFC5Richmond6W Bulldogs/FFC4Sydney/SMFC
7Adelaide8North Melbourne7Adelaide9Port Adelaide
13Melbourne15Essendon13Melbourne14St Kilda
16Gold Coast17Brisbane Lions

Carlton would be drawn to play Freo, WB, Adel, Coll and Melb home and away. This makes an initial interstate journey count of two (italicised for your convenience). The other teams they are to play once are listed on the right side of the home and away blocks above.
In the example, there are three more journey for a total of five.

Note also that if you sum up the ladder positions for both right hand side lists, you get 57.

A mid rung team now... lets look at 8th placed
North Melbourne
2West Coast1Fremantle2West Coast4Sydney/SMFC
5Richmond3Hawthorn5Richmond6W Bulldogs/FFC
11GWS7Adelaide11GWS9Port Adelaide
14St Kilda12Collingwood14St Kilda10Geelong
17Brisbane Lions16Gold Coast17Brisbane Lions13Melbourne
Home and away to the Eagles, Rich, GWS St K and Bris. That's three out of state journeys already.
So rounding out the other 'one only' season match-ups and keeping both sides to a total of ladder positions to 57, North would collect two more interstate trips to play Sydney and Port.

And finally, lets look at two top 6 teams. First,
W Bulldogs/FFC
1Fremantle2West Coast1Fremantle3Hawthorn
12Collingwood8North Melbourne12Collingwood9Port Adelaide
18Carlton15Essendon18Carlton14St Kilda
16Gold Coast17Brisbane Lions
Again interstate travel from opponents inside their set gives them two journeys, so we balance it up with three away (Syd, Port, Bris), and again both pairs of 6 opponents makes a ladder position total of 57.

Lastly, Hawthorn show where occasionally a team gets unlucky. Keeping the 57 point balance rule for home-away opponent strength, the scenario sends Hawthorn out of Victoria 6 times.
4Sydney/SMFC1Fremantle4Sydney/SMFC2West Coast
9Port Adelaide5Richmond9Port Adelaide6W Bulldogs/FFC
10Geelong8North Melbourne10Geelong7Adelaide
16Gold Coast14St Kilda16Gold Coast13Melbourne
17Brisbane Lions18Carlton
It can and will happen, and as before its just the luck of the draw.

So when the AFL release their fixture that purports to be fair, you will know of an alternate approach that is not rigged to suit the broadcast 'partners', or to satisfy an additional few thousand fans.

That is transparent and has a sense of balance. Where all teams play opponents of all strengths.