Ultimate AFL Fixture Guide - Season 2016 Release

Dylan82

Club Legend
Aug 14, 2004
1,924
5,824
Brisbane
AFL Club
West Coast
Other Teams
Cardiff City
The 2016 Season release is finally here!

Better yet, due to the ongoing fire sale of the economy whilst our "honest" representatives fiddle playing pin the dildo on the ideological butt-pirate, I’ve had more time to put this year’s edition together whilst still getting paid – thanks guys, really couldn’t have done it without you. :thumbsu:

Available for download here:
https://www.dropbox.com/s/f14jg2ewyeuo4wb/Ultimate-AFL-Fixture-Guide_Season-2016.xlsx?dl=0



Full list of features:

  • Complete season fixture list and finals schedule.
  • Season fixture breakdown for each club, including:
    • venue information
    • distance travelled
    • opposition strength
    • number of short breaks
    • simple (1 run) model forecasting results for each match
    • visualisation of overall fixture
  • Live ladder that updates from match results instantly.
  • “Premiership Zone” chart based upon average scores for and against that has become popular in the media of late.
  • Top 8 Finals calculator.
  • Complex (55 run) model for forecasting the result of individual matches, taking into account:
    • Recent vs long-term form.
    • Number of times a club has played at the match venue in the last two seasons.
    • Travel for the current match as well as in matches prior.
    • Crowd support and the “atmosphere” or intensity of it.
    • Strength of opposition teams played prior to the current match.
    • Statistical averages of the competing teams.
    • Finals experience over the past three seasons.
And includes the following:​
  • Forecast match final score.
  • % of model runs providing winning results for each team.
  • visualisation of match forecast statistics (ie % of disposals).
  • visualisation of match forecast model outputs.
  • Forecast match attendance and % crowd support.
  • Backward compatibility with season 2015 fixtures.
2015-GF-Forecast.JPG
2015 Grand Final Match Forecast
  • End of Regular Season (Round 23) Ladder Forecast
  • Fixture difficulty analysis and rating for each club with accompanying charts.
  • Average statistics and rankings for each club
  • Complete fixture distance travel data and charts
  • Full match results and statistics for previous seasons 2014 & 2015.

ade-fixture-2016.JPG
Adelaide 2016 Fixture

2015-Premiership-Zone-R23end.JPG
Premiership Zone 2015

2016-TeamStats-Start-Chart1.jpg
Example Team Stats Rank Chart

2016-Fixture-Travel.jpg
2016 Fixture Travel



Caveats:

  • As modelling is currently based purely upon team statistics, individual player developments that may impact team performance (eg trades, injuries) are not taken into account.
Hence Essendon can be expected to be overvalued due to the players now missing from their list. Additionally Geelong may be undervalued due to the Dangerfield trade and conversely Hawthorn may potentially be overvalued due to the long-term injury to Roughead.​

This is an area where further development has already begun and it is currently expected that modelling based upon the current available players in the squad will be operating in the 2017 release.​

  • It is acknowledged that not all membership figures posted by clubs are equal. But until the AFL start publishing apples with apples figures the current media release figures are the next best official approximations.
  • An edition was created where the complex model was utilised rather than the simple model for the Round 23 Ladder forecast. However this ended up bloating the filesize to over 50Mb and caused it crash when you were not running Excel-64bit due to memory limitations – so the decision was made to revert to the simple model for this purpose.



Fixture Observations & Predictions:

Difficulty / Hardness:

2016-Fixture-Difficulty-Commencement.JPG
2016 Fixture Difficulty Rankings
  • Based on 2015 final positions, Adelaide appear to have the toughest draw for the 2016 season, surprisingly followed by GWS and then last year’s runner-up West Coast.
  • Adelaide play against 2015 top 4 opposition 7 times – the next highest is Sydney and the Bulldogs with 6.
  • GWS only play 10 matches in Sydney, with one of those being at the SCG.
  • Two Melbourne clubs (Collingwood & Essendon) will not need to travel 17 times out of 22.
  • Four Melbourne clubs (Bulldogs, Collingwood, Richmond & St Kilda) are scheduled to play at the same venue 14 times.
  • Fremantle and West Coast both journey to NSW & Qld three times this season. In just these three matches the two WA clubs will travel more distance than all Victorian clubs will for the entire season, with the exception of Melbourne (due to their two “home” games in the NT).
Obviously as the season progresses, the shifts in fortune of clubs from last season to this one will change the view of its difficulty.
For example this time last year it appeared that Geelong had the most difficult fixture. 2015-Fixture-Difficulty-Commencement.jpg

After completion of the season however, it is apparent that Port Adelaide had the most difficult fixture of the 2015 season, playing top 8 opposition no less than 13 times – including a horror start playing all the top 6 teams in the first six weeks. 2015-Fixture-Difficulty-Hindsight.jpg



Ladder:

  • Simple (1 run) model forecast for the 2016 season prior to its commencement is the following:
2016-Fixture-Forecast-Commencement.jpg
End of Regular Season Ladder Forecast

The two grand finalists from last season will be in contention once again whilst a resurgent Port Adelaide will recover from last season’s underachievement and Richmond continue to develop. Fremantle slipping out of the top 4 and the Bulldogs slipping out of the top 8.
Carlton and Brisbane Lions in for long seasons near the bottom.

***Note: Refer the caveats listed above regarding Essendon, Geelong etc.




Match Prediction:

Round 1, Match 1.

Richmond to win by nine goals.

2016-Rnd1-Forecast.JPG
Richmond vs Carlton Forecast: 113-59



FAQs:

How do I keep this up to date?
  • To keep this guide up to date, all that is required is to input the match final scores and associated team statistics upon the “Fixture” tab as they occur.

WTF there is no way my team will finish that low – you must be trolling?
  • The forecast modelling is based upon real statistics from past matches. All calibration is done through the recreation of past known results from prior data (many hundreds of times over). Should the team you support happen to be undervalued then it is through no means of malice nor bias; it is just the expression of the highest probably outcome according to a formula.

Why have you done this?
  • I got sick of constantly searching for a decent fixture in order to plan which games I would be going to in a season. Without an alternative, the first version was created. I get paid to work in Excel everyday – it actually wasn’t too hard to put this together. Over time, more features were added to assist my tipping endeavours.

How do you work out the scores in the forecasts?
  • The average scoring shots for and against for each team are taken and then modulated according to team statistical average outputs to provide a “neutral” result (ie if all matches in the season were played at one venue with no support). This is then further modified by a series of factors to account for the match location and crowd support in order to produce the forecast score output.

Why use scoring shots rather than the actual scores?
  • Scoring Shots have used in place of actual scores in order to reduce discrepancy in result over smaller sample sizes due to playing conditions. (For example, a team may score 20.10 at Docklands with the roof covered whilst on the same day another team with similar statistical output for the match may score 10.20 at the MCG due to the weather.) Over increasingly larger sample sizes, score becomes statistically irrelevant as it converges upon a 3+(2/π) ratio to the number of scoring shots.

Why only use the team statistics [Disposals, Effective Disposals, Clangers, Inside 50s, Contested Possessions, Tackles]?
  • These are used as verified sources are easily accessible and they each have significant correlation with final results over the last two seasons (72%, 74%, 68%[inverse], 72%, 69%, 54% wins respectively). Hitouts and clearances are irrelevant unless they can be deemed to advantage – the statistics of which are not readily available. Marks as a statistic are a less accurate expression of effective disposal in today’s football which is way effective disposal has been used in its place. Other statistical inputs that may have been useful – ie Metres Gained, Intercept Marks have been omitted due to the lack of an easily accessible verified source.


Unrelated Quiz Night Trivia:
  • Canberra is closer to Brisbane than Adelaide.
  • Darwin is closer to Perth than Brisbane.


Enjoy! May your tipping go well and your team bring you happiness this year (except for when they play the Eagles, of course ;))
 

(Log in to remove this ad.)


Top Bottom