Thursday, 10 September 2015


Before heading into the 2015 AFL Finals, a quick look at the top 8 teams' performance amongst themselves, and something a bit more 'experimental'.

Firstly, extracting only the games played between the finalists provides a little window into some possible outcomes over the coming weeks.
The table of those games results is
2W Bulldogs/FFC85037107572093.80.63
4West Coast1050599288420112.20.50
8North Melbourne93066789121274.30.33

The immediate interest point is how well the Bulldogs have gone, ranking 2nd in wins from available opportunities. Their percentage is negatively impacted, and we can blame West Coast for that, but none the less it highlights their great year.
It remains to be seen if they can go deep into the finals.

The three key disappointments are Sydney, Adelaide and North Melbourne, all with only 3 wins. Sydney and North in particular are looking to struggle in September with average scores of only 75. North the worse of the two with average conceded scores of 101.

The For/Against data also limits the likely effectiveness of the Tigers, with an average score of 79 against top 8 teams. The saving grace for Richmond is the defence only yielding 80, so it could be possible that when both phases are functioning, the Tigers might make an impression.

Predicting finals football is harder than regular season football as the importance that each game carries acts differently on teams. The pressures and stresses are higher, and the ability of players to cope varies.

So with that in mind, and the fact the game simulations are yet to properly account for drawn matches, a table has been prepared of an expectation of teams to reach various phases of the finals series.
2West Coast100729771.4
7North Melbourne10032140.1
8W Bulldogs/FFC1002600
NOTES: These predictions are made through scenarios that are based on the FMI team rankings after round 23, and adjusted as the finals series games are played. For another perspective, read the MatterOfStats blog on predicting Finals outcomes.

So, to read the table lets look a a few examples.

Lets start with the top of the table Hawthorn. They are (obviously) 100% chance to play week 1, and also 100% to play a Prelim final. They are also a 28% chance to play in week 2, the Semi Finals.
What this means is there is a 28% likelyhood of losing in Week 1, and to return to a 100% likelyhood of playing in the Prelim they are certain to win in Week 2.
The scenarios then give return the Hawks to an 84.6% likelyhood of a Grand Final berth.

That is an interesting outcome for a team with a double chance, but if you look at teams without that double chance, the numbers taper off in a simpler and clearer way.
For example Adelaide are expected to progress to Week 2 in 74% of scenarios run (the other 26 going to the Bulldogs), but then only move to a prelim in 3% of scenarios. Getting past Hawthorn or West Coast is a tough ask for the Crows.

Fremantle is also interesting, as they also tail away quickly even for a team with a double chance. The FMI rankings are unfavourable to Fremantle, sitting on 1224 now, but down from 1304 from the season start. This slip from the Dockers is factored in and the scenarios show a very slim chance of a Grand Final appearance.

The low Dockers FMI Ranking also is the 'door ajar' that allows the Tigers to convert the 67% scenario return of a Week 2 appearance into a 41% chance of a Prelim.
But as before, the predictions are quite difficult as the Finals are basically a different football season. Anything can happen, Dockers fans!

And Bulldogs fans please don't be disheartened by the predicted end to your finals tilt in Week 2. Your form is great, your head-to-head numbers against the other 7 opponents is also good. Your a developing team and the FMI ranking value now of 1154 is more than likely undervaluing your team.

As before, these numbers are rough and quite experimental (and only developed over the last 2 days) so read into them what you will.

And remember... no wagering!

No comments:

Post a comment