Tuesday, 31 January 2017

NWL: 1- LAUNCH

This weekend sees the start of official play in the first season of the AFL's national women's league, and it is a very exciting time.

No matter what your views on the competition, we are seeing the birth of a new league. Just like our predecessors who oversaw the break away of 8 clubs from the VFA to form the VFL back in the 1890's.
Or those of us who went through the evolution of state-based football into a national competition.

We are all seeing this happen in roughly a single year.

It is a defining moment in football.




So, we have eight new teams to follow, track, support and tip.
Eight new entities spread over the nation. A whole array of new faces and fans come to football, as too venues.
And a promise of more to come.


Season 2017 of the AFL Women's League for tipsters and predictors offers challenges as well. Who are these new faces, and how will they gel and coalesce into a team to influence the game? What advantage is applied to playing at a venue bring?
There is no base line of form or venue benefit, other than what has been played out over the 2016 series of preliminary matches. The player draft has mixed the pot further.

This first season will be hard for most tipsters. And perhaps too the other seasons to follow given the short Feb/Mar season, and state league season in between (the 'AFL offseason').
There is limited data to establish a base line, but you do need to start somewhere, with a first tentative step.




Before launching into the first round tips for AFLW, a few other tables for assessment.


First - the team rankings:

Under the FMI system, each new team, from day one of the competition or introduced later, starts with a base value of 1000 points.
Launching a predictive model where all teams are equal makes tipping very difficult as there are only a few factors that influence the outcome per predictive iteration. The draw and the degree of home ground advantage are two key points in this. Again, with no established baseline, a nominal 7 point HGA has been used for all the season 2017 data.

Another approach is to drive in some subjectivity to the rankings by looking at assessments of players per team and squad depth. Subjectivity is best avoided in the FMI system, so this season will see something not undertaken here before - parallel rankings systems are being deployed.

So through assessing information on marquee players and drafted talent etc, the secondary, subjectively pre-weighted initial rankings are also listed below.


FMI RANKINGSsubFMI RANKINGS
TEAMPTSTEAMPTS
1Adelaide10001Collingwood1096
2Brisbane Lions10002W Bulldogs/FFC1084
3Carlton10003Fremantle1039
4Collingwood10004Brisbane Lions1015
5Fremantle10005Melbourne998
6GWS10006Carlton942
7Melbourne10007Adelaide923
8W Bulldogs/FFC10008GWS902

Interestingly, the FMI assessment of depth is in broad agreement with that done by the HPN team.
[Edit: See also the pre-play power rankings uploaded by friend of the FMI, Troy of the WFD.


Based on the above rankings and the supplied fixture, an assessment of each teams draw can be compiled.

Applying the above to a 'tough draw index' analysis provides a further pair of tables as below.
Of course, with the uniform ranking points, the only real dividing influence is the determined home ground advantage.

FMI RANKINGS
RNKTEAMLAD POSFMI RANK PTSSUM OPP'N PTSHOME / AWAY BONUSTOUGH DRAW INDEX
1Brisbane Lions-100070001000.39
2GWS-100070001000.39
3Carlton-1000700000.00
4Collingwood-1000700000.00
5Melbourne-1000700000.00
6W Bulldogs/FFC-1000700000.00
7Adelaide-10007000-100-0.39
8Fremantle-10007000-100-0.39
subFMI RANKINGS
RNKTEAMLAD POSFMI RANK PTSSUM OPP'N PTSHOME / AWAY BONUSTOUGH DRAW INDEX
1GWS-90271981000.66
2Brisbane Lions-101570851000.36
3Carlton-942705800.19
4Melbourne-998700200.00
5Collingwood-109669040-0.20
6Adelaide-9236977-100-0.26
7W Bulldogs/FFC-108469160-0.26
8Fremantle-10396861-100-0.48


When applying the subjective expectations, a wider spread through the index is found, though basically maintaining a similar outcome of toughness of draw.


From this, we can also extend into predicting the season for each team

Attempting to predict a number of wins is a challenge when all teams are evenly ranked and expected to perform at the same level. Hence the table below which has a certain degree of uniformity of percentage expectation to finish in a given spot on the ladder.

PS - The below are ranked in order, so yes Brisbane to finish top in the table immediately below. And the even-ness with everyone on 1000 points basically places the table below in a rough approximation to the tough draw index above.

FMI RANKINGS
LADDER POSITION12345678
1Brisbane Lions1413121213121111
2GWS1514141312121110
3Carlton1313121312121211
4Collingwood1413131312121211
5Melbourne1312131212121213
6Fremantle1112121213131313
7Adelaide1112121313131314
8W Bulldogs/FFC910111212131517
subFMI RANKINGS
LADDER POSITION12345678
1Collingwood462614841
2W Bulldogs/FFC29292012731
3Fremantle1219232015831
4Brisbane Lions8142022191052
5Melbourne4101620221584
6Carlton24916262418
7Adelaide12510213031
8GWS148162844

When the pre-weighted rankings are applied over a 7 game season, the expectation is as above, and fairly much in alignment with the weighting.
There just isn't enough games in the season to affect deep change.

Hopefully though, when the games are played and teams evolve, we will see something approaching a form line emerge.




So finally the tips for Round 1 of the AFLW.
[Edit: venue change for game one change the tips a tiny bit]

Under the non-subjective system:

HOMEAWAYVENUEPREDICTED RESULT
RND 1
Carlton0v0CollingwoodPrinCarltonby1pts
Adelaide0v0GWSThebAdelaideby7pts
W Bulldogs/FFC0v0FremantleWhitW Bulldogs/FFCby7pts
Melbourne0v0Brisbane LionsCaseMelbourneby7pts


Allowing for influence of playing strength:

HOMEAWAYVENUEPREDICTED RESULT
RND 1
Carlton0v0CollingwoodPrinCollingwoodby16pts
Adelaide0v0GWSThebAdelaideby10pts
W Bulldogs/FFC0v0FremantleWhitW Bulldogs/FFCby12pts
Melbourne0v0Brisbane LionsCaseMelbourneby6pts


Finally, the most important thing...

LETS HAVE US SOME FOOTY!

No comments:

Post a Comment