Wednesday, 22 March 2017

2017 UNPREDICTIONS: EXPECTED WINS

Twice more unto the breach, dear reader, twice more.

Now in its 6th big season, its your FMI's trusted and ever present assessment of the season ahead.
And this year its a double shot of posts.
This post is on the expected number of wins per teams generated via simulations of seasons. And an alternative look at the season finishing ladder position has been blogged, with both generated by the same simulations.

Basis:
Under the FMI approach, the ratings of each team are used to calculate ranges of possible match results, adjust ratings after each and event, and continue iterations for the whole season, then repeat the process again and again.

The rankings used are as per the end of the 2017 season, and are listed below. The league fixture also has an impact, and the handicapping of it is examined as per this post.

A further note: There has been a small style change to the blog this year as below. Nothing earth-shattering but just a small adjustment to make the blog present better.



SEASON WINS

After running 10,000 season simulations, the data was collected and aggregated to produce the below spread of expectations, expressed as a percentage.

With such a diverse array of outcomes, the order in the left most column is speculative to say the least. What is visible is the clusters of teams with similar expectations.
For example, there is a block of teams from Sydney down to West Coast all in the 14 to 19 win window. Compare this to the Hawthorn to North Melbourne block that is centered on a spread of 11 to 15 win window.
EXPECTED WINS012345678910111213141516171819202122
1Sydney/SMFC1248121618181362
2GWS1248131618171172
3W Bulldogs/FFC1358141718161152
4Geelong11359131617151162
5Adelaide124691215151412731
6West Coast12581215161413741
7Hawthorn1236811141313128521
8Port Adelaide1471217201611721
9Collingwood124691215161411732
10N Melbourne1247912141413107421
11St Kilda24812151715128421
12Melbourne13610141717149631
13Richmond1481519191496311
14Gold Coast371418201711631
15Fremantle3817232214831
16Essendon271320201611631
17Carlton412181917128531
18Brisbane Lions112328201151

At the foot of the table, the groupings become tighter as the rankings points become more distant from their neighbour. Even with that in play, there is still a grouping of teams around the 4 to 8 win field.


Taking a more broader view of the win windows, the below table is slightly easier on the eye and makes the mess of data above more discernible.

EXPECTED WINS17-2212-166-110-5
1Sydney/SMFC7426
2GWS55441
3W Bulldogs/FFC52461
4Geelong51462
5Adelaide51453
6West Coast39574
7Hawthorn176320
8Port Adelaide37125
9Collingwood56133
10North Melbourne347471
11St Kilda14795
12Melbourne10809
13Richmond27028
14Gold Coast5841
15Fremantle4951
16Essendon3861
17Carlton2970
18Brisbane Lions198

TEAM CHARTS

Below are charts for each team that show the distribution of expected wins, grouped 2 per chart (and regionally where appropriate).


Once again in 2017, the two Sydney teams look to be playing in the same space. Will be interesting times come finals, and lets look forward to another derby final.


Port Adelaide's numbers are strong in the FMI analysis, and this varies from other pundits. Adelaide should be on track for another finals berth based on the spread below.


Another lean year for the Lions is on the cards, while for the Gold Coast it is hard to see where improvement will come from. A good crop of kids are developing, but so are those in others teams.

West Coast look set to outshine Fremantle again. The uneven-ness in the Eagles chart is interesting and hints at a struggle to maintain a finals spot this year.



Both the Saints and Blues did exceptionally well last year, with one stopping the rot, while the other nudged the finals. For St Kilda, doing that well pushed them into a tougher fixture bracket, so maintaining the status quo will be a KPI. Carlton needs to focus on a rebuild.


The well defined peak in the Collingwood chart is interesting. The companion blog post has them finishing 9th overall, but the peak on 14 wins suggests a very close race for 7th and 8th this year. North's curve is flatter, reflecting the assumption that they are destined to fall from the finals race.



Another sharp peak for Essendon, while a more curved line for the Tigers. Both teams have upside this year as shown by the extension of the curve higher up the expected wins scale.



Geelong's possible place in the finals is emphasized by the strength of results over 15 wins. The Hawthorn curve is low and slow, and along with North Melbourne, the only ones to not broach the 15% mark at peak. Could finish anywhere.



The almost perfect bell curves here are a thing of beauty. For the Bulldogs, it pits them about where they were in 2017, so ready for another Premiership tilt. Melbourne's curve, like Essendon and Richmond's above has a nice forward lean to it. Upside, for the usually down Demonistas.


As before, the sister post to this one on 'expected ladder position' might be worth your time too.

Good luck to all teams and tipsters in 2017.

No comments:

Post a Comment