Blow-out Me Down With A Feather.

The 2011 AFL season was the most lopsided in the history of the competition. The number of blowout games (on most measures by margins), and also by ladder percentage disparity (since 1970, when a 22 week home and away season became the norm). I don't put this down to the start of the Gold Coast Suns, but to that commodity every team, and everyone else, wants more of.

I had tweeted about this during the season, with charts etc, but I felt it also warranted further study and analysis. It may be that I have also found a possible root cause... more on that in the next post though.


The Blow-outs Of The Season.
There was a lot of media chatter about the 2011 season, particularly the blow-out games that have been posted in the back-half of the season. It all really fired up after this result:
        30/7/2011:   Kardinia Park  - Geelong  37. 11. 233  def  Melbourne  7. 5. 47

The aftermath of course brought the downfall of Dean Bailey, and an apparent push by the leading clubs to 'fly the flag' and post huge scores against lowly clubs as well. Which may have been the impetus for these results:

       06/8/2011:   Football Park  - Port Adelaide  3. 3. 21  def by  Collingwood  23. 21. 159

       06/8/2011:   Kardinia Park  - Geelong  29. 14. 188  def  Gold Coast  6. 2. 38 

All of this brought out talk (and tweets) about how blowouts were a big problem for the AFL, which His Andrew-ness said was an over-reaction (ABC News).

         "I just think they are a very rare occurrence," he said .... "We always knew that the

         Gold Coast would have a couple of largish results ... I don't think it's one that requires

         a crisis meeting and hysteria."

          - Andrew Demetriou: Aug 10 2011.

The following week, despite the head-honcho's assurances that massive blow-outs were rare, brought another belting (or big-club-showing-it-could-also-beat-up-a-lowly-club... read that as you will) 
       13/8/2011:   M. C. G.  - Hawthorn  31. 11. 197  def  Port Adelaide  5. 2. 32

So in the space of 3 weeks, the AFL served up the incredibly rare event of 4 blowout matches, for a combined scoreline of    120. 57. 767  -v-  21. 12. 138    Margin:  629 pts

And of these four, only one was a Gold Coast game... so much for His Andrew-ness's logic. Adding to the increasing rarity of these blowouts was the two earlier +120 point games in 2011 as well:

       01/5/2011:   Docklands  - Essendon  31. 11. 197  def  Gold Coast  8. 10. 58

       22/5/2011:   Subiaco  - West Coast  26. 19. 175  def  Western Bulldogs  8. 4. 52

Which benchmarks this season as the worst season EVER for
   - Margins 60 points and over           46 games
   - Margins 100 points and over         10 games (equal with 1991 season)
   - Margins 120 points and over         6 games (equal with 1991 season)
   - Margins 140 points and over         3 games

To put that in perspective, the chart below shows all results with margins of +60 points since 1897 (broken down into 20 point increments)

Click to expand.

And while this year has set a new high-watermark in un-evenness of the competition, the scoring rate (blue line) remains relatively stable at just around 180 pts scored per game.
Also interesting that the 2011 season has run against the trend of the last 10 years, which had been reasonably consistent.

You could also look to the disparity between teams ladder percentages over the years. Since 1970, when the now standard 22 round season was first introduced, 2011 is again the most unbalanced.

Click to expand.

The chart above shows the Standard Deviation of all teams scoring percentages in 2011 to be 31.8, And the net difference between the maximum and minimum teams scoring percentages in 2011 to be 111.4%.

     * NOTE: Removing results of games involving the Gold Coast, adjusts this to a Standard Deviation of 30.4, and
     total difference between the maximum and minimum teams scoring percentages in 2011 to be 104.2%. Both of
     these adjusted figures are still record levels.

The Reaction So Far
So has this sent alarm bells ringing down at AFLHQ?
Well, it possibly triggered an internal season review (read as "crisis meeting" if you like), because His Andrew-ness came out at seasons end on Melbourne radio (MTR) to advise that...

         “What will happen you will see more one side affairs, you will see more one sided
          affairs, and western Sydney and Gold Coast suffering big losses as they build their
          team ... It will probably take three years for a correction to happen into the game
          where we get back to much more even games, get back to the pure draft after next
          year ... But for the next couple of years you will see these things that are skewed
          to blowouts"
          - Andrew Demetriou: Sep 28 2011.

...none of which is any surprise to football followers. Thanks, Andrew, for the concession speech, and lets assume this was written up in the Grand Final Edition "Football Record", on page 31, small print, bottom inside corner under the title "We Were Wrong"...?

Or, maybe not.

And to cap it all off, we can expect more in 2012, as

          "The AFL called a press conference yesterday to announce it was making ''no major
          rule changes'' for next year."
          - Sydney Morning Herald: Oct 12 2011.

 It will be interesting to see what happens further both in the off-season and during 2012.
- Will any post-season analysis and review bring up any root cause, or mirror to the past?
- Will AFLHQ read and consider any external opinion and study?
- Is Andrew planning on buying new curtains for the office and hoping it all goes away?


I will post up some ideas on why I think all this has happened and the link to the dollars later.
And also why I don't think this is a result of the arrival of a new team.

Feel free to offer your theories now.

4b. How Could Conferences Work in Practice.

As before here ("18 Into 22 Wont Fit") and here, here ("4a. How Could Conferences Work in Practice.") is another installment of a season trial run under a three conference system.


Case 2:
Conferences and Results:

Working from the same conferences as you would produce from the 2011 ladder, and adding in random scores, another season and conference tables can be generated, as at right.

In this, more complicated example, conference championships go to Sydney, Carlton, and North Melbourne (above the solid red line).

This, in my proposal, also gives those teams Qualifying Finals berths and the double chance. 

The fourth Qualifying final spot would be allocated to Brisbane, having the best record of the next 9 teams.

The other finalists would then be Melbourne, Western Bulldogs, Richmond and Gold Coast, and missing out on finals in this instance are West Coast, Adelaide, Geelong and St. Kilda.

The interesting point with this scenario is that Carlton (50pts, 103%), with a worse record than Brisbane (52p 103%) and Melbourne (52p, 96%) are guaranteed a Qualifying Finals slot and double chance as conference winners even though there record would see them finish 6th on a traditional AFL table.

This might create a kerfuffle with those teams having better records missing out on the double chance spot, but here is how I see it...

  - Winning the conference and gaining an automatic double chance slot, gives teams more encouragement and reward to gain that title outright. Without the extra carrot to the guaranteed double chance slot, the conference title would carry little value.*1

  - If you are a conference winner with a poor result, or from a poorly performed conference, there is a good chance you will be 'out in straight sets' in the finals anyway.

  

Into The Finals:

As above, the In this proposal, all the conference winners are guaranteed a final, and I would also give them a double chance slot (along with the best second-placed team). I think it is important to reward conference champions in this way as it helps give that title meaning and a value to strive for.

So in the example, the top four finalists (with the double chance) would be;
QF 1: North Melbourne -v- Brisbane
[i.e. best performed conference winner -v- best 'wild-card' team (2nd place finisher in all conferences)].
QF 2: Sydney -v- Carlton
[i.e. second best performed conference winner -v- 3rd best conference winner].

For the other four finalists, ranking the  9 teams positioned 2nd to 4th in each conference (less the best 2nd place finisher) are then simply ranked by points and percentage, then allocated an Elimination Finals berth (or not finals ranked). This would then rank the remaining finalists from those 9 teams as below:
    Melb. (52points, 96%), W.Bulldogs (50p, 110%), Rich (48p, 104%), GCS (48p, 103%),

So using this system, the two Elimination Finalists would then be;
EF 1: W. Bulldogs -v- Richmond
[i.e. 3rd best 'wild-card' ranked team -v- 4th best 'wild-card' ranked team].
EF 2: Melbourne -v- Gold Coast
[i.e. 2nd best 'wild-card' ranked team -v- 5th best 'wild-card' ranked team].


Setting Up The Next Season:
To set up the next seasons conferences, the full teams list is merged back into a single (traditional) AFL ladder as at left, and then separate the teams based on that.


So from the left, the next seasons' conferences get built again as per the process;
- 1st, 4th, 7th, 10th, 13th and 16th.
- 2nd, 5th, 8th, 11th, 14th, and 17th.
- 3rd, 6th, 9th, 12th, 15th, and 18th.
*2


This creates the new seasons conferences as below.



In this, we still have the three conference winners in different conferences for the next year, but this is not by design, just the quirks of the process. There could be times where two conference winners are in the same group the following year.


*1 Possibly a very simplistic thought, as teams also want to win as many as possible to get a finals slot, but I feel there needs to be some sort of 'aura' about winning the conference. Manufactured or not.


*2 Happy to consider other proposals on allocating teams here... Obviously using this system, the Barassi group is slightly stronger than Robran (all teams theoretically 2 places better, etc). I haven't thought properly about balancing / seeding it too much, so if anyone has any ideas I will gladly consider.

Collingwood: You Still Have This...

I have co-opted someone else's data (via the web) into excel, and now have a list of every VFL/AFL game ever played. Teams, scores and venues. I used this to analyse 'blowout games' as I tweeted about over the last few months ...and may add that analysis onto this blog in the future.

But just as a little look back over all 13959 games, a full ladder of all games ever played (including finals, excluding non-premiership games) gives us the below ladder.

Note that I have had to base this on a 'Match Ratio' basis, as the number of games is imbalanced.

So over the journey, Collingwood hold the title of the team with the best win loss record. Something to help them sleep well at night, I'm sure. 

Also in the above I have grouped teams results as per the below, as ...

      South Melbourne (1897 - 1981) and Sydney (1982 - current)

      Fitzroy (1897 - 1996), Brisbane Lions (1997 - current) and Brisbane Bears (1987 - 1996)

      Footscray (1925 - 1996) and Western Bulldogs (1997 - current)

      North Melbourne (1925 - 1999, 2007 - current) and Kangaroos (1999 - 2007)

The table also roughly mirrors what you could consider the ranking of Melbourne clubs, membership and supporter wise... (debatable point, as it is...) 

    The Big Six:     Collingwood, Carlton, Essendon, Geelong, Richmond, Hawthorn,

    The Others:     South Melbourne/Sydney, Melbourne, Fitzroy

                          Footscray/Western Bulldogs, North Melbourne, St. Kilda

Also an interesting disparity between West Coast and Fremantle in performance, that is not mirrored in Adelaide.

I will pick apart these stats further in future posts.

* For purely academic purposes, a split-out those entities above to make the below table is produced.

The Grand Final We Had To Have *1

When I hear the terms 'Collingwood' and 'Grand Final', I always think.... "Gran Sale, Gran Sale, Gran Sale"

... or "Megalo Megalo Megalo Xepoulima", or "Grande Svendita".... or whatever. Because when I was a lad, 'Collingwood supporter' was a euphemism for 'multicultural'.

All my Greek and Yugoslav*2 mates were Wobblers (with the Italians firmly of the Blue Filth faith, of course).

Anyway...

As I have said before, the 2011 Premiership is Collingwood's to lose. Best team all year, depth of talent, depth of pockets. Unbackable favorites during the season, at least until September, they have the building blocks to go back to back*3. 

The only problems for the Wobblers have been...

1) being beaten by Geelong twice this season, once in a crunch close game, and the other a blow-out... Value of that second game? Probably nil if your a Wobbler fan, probably lots as a Hand-bagger.

Me, as an independant... any win is good, and one for the psychological bank.

It's there in the back of everyone's minds - They. Got. Smashed!

2) An apparent drop-off in the late season... have they been worked out as The Age says? Maybe, and would be interesting too. Football and these 'structures' everyone talks about. At the end of the day, score more than them and win. Simple.

3) Last week, the Wobblers got beat-up by the Dawks, while the Hand-baggers had a reasonably easy game against the Weagles*4. If fitness levels and bums-on-the-ground numbers are weighted one way or the other in the 3rd quarter, that might be all she wrote.

Its not an easy one to pick.

- Well, we would love the Wobblers to get done. So would most of the football world.

- It would also be good for the philosopher Malthouse to go out on a win... a consistently good coach and former Tiger.

- If the Wobblers win, they will go down in history as 'those flogs that ditched a back-to-back premiership coach'... well worth writing to the papers about. Nice!

- Might be fun times if Mick gets a win, and Bucks brings nothing to the club for the next 5 years... just a bit of crystal ball schadenfraude-ing.

- Would also be good for the Scott-boy (I never know which one he is) and Geelong to prove one coach and one Browlow medalist do not a club make. You know... strength of team, and all that. 

So, time to put them on the block... lets call it.

Heart - Geelong ... and I don't really know why. 

Head - Wobblers or Hand-baggers... too close to call, but I still lean to Geelong.

Happy to hear your comments, either in the comments section here, or via twitter @TKYC.

... AND, no correspondence will be entered into with 'Monday's Experts'. Tell me now, or blow it out your ring-piece.




*1   With apologies to Precious Paul J Keating.

*2   Gen-Y'ers, ask your Mum and Dad about Yugoslavia.
*3   In some ways, Collingwood going back-to-back is a better thought than Collingwood supporters going at it in the-beast-with-two-backs mode... if you get my drift... Are you with me?
*4   So I'm told by my office non-serious football followers... never saw, heard or read anything about those games. The joys of a great week in FNQ!

4a. How Could Conferences Work in Practice.

I have written quite a bit over here ("18 Into 22 Wont Fit") about an AFL draw that tries to balance out the playing year to 22 rounds, with teams playing other teams at least once (and 5 other teams at least twice) inside a conference system. That post has grown so long that it might be easier to review how it works with results in a 'trial' on a separate blog post or two (or more).

Case 1:
Conferences and Results:
So now we have our conferences split and the draw organised, here is how the actual conferences work in practice.

I have used excels random number generator to determine each game and score line. The Micrsoft Excel function =RANDBETWEEN(40,150) was used to generate team scores between 40pts and 150pts*, which were plugged in to each game to generate wins, draws and losses.

The conferences, after a full (random) season, then produced the ladders as below, giving conference championships to West Coast, Geelong, and St. Kilda (above the solid red line).

It should be worth noting that in this first example, I have chosen a very basic set of results, where all the conference winners are also the best three teams of the season. This is done to highlight what 'should' happen using this system. I will follow-up later with posts on other possible scenarios as well.

Allocating teams from three conferences into eight finals slots creates issues. For example, the top 3 teams from each conference only gives 9 teams, which is one to many.

I would propose the conference champions get automatic entry into the finals as the first three seeds. The next three teams in each conference (below the solid red line and above the dotted red line) are 'wild card' teams for the finals, with the best 5 of those 9 teams gaining a finals spot.

I have run many iterations with results on this, and have found that on some rare occasions, a fourth placed team in a conference can have;
   a) a good enough record to be in a traditional finals 8, and
   b) a better record than 2nd or 3rd placed team in other conferences.
...I may even run more iterations to find one and then add it to the blog.

So working with the best 5 from the 3 groups of 2nd to 4th, there are finals berths for
- Melbourne and Collingwood;
- Adelaide, and;
- Hawthorn and Richmond.

Missing out would be Port Adelaide; Carlton and Essendon; and Brisbane, as they have the worst record.


Into The Finals:
In this proposal, all the conference winners are guaranteed a final, and I would also give them a double chance slot (along with the best second-placed team). I think it is important to reward conference champions in this way as it helps give that title meaning and a value to strive for.

So in the example, the top four finalists (with the double chance) would be;
  QF 1: West Coast -v- Melbourne
[i.e. best performed conference winner -v- best 'wild-card' team (2nd place finisher in all conferences)].
  QF 2: St. Kilda -v- Geelong
[i.e. second best performed conference winner -v- 3rd best conference winner].

For the other four finalists, I would propose that the 9 teams positioned 2nd to 4th in each conference (less the best 2nd place finisher) are then simply ranked by points and percentage, then allocated an Elimination Finals berth (or not finals ranked). This would then rank the remaining finalists from those 9 teams as below:
   Coll. (48points, 101%), Hawt. (48p, 100%), Rich (44p, 102%), Adel. (44p, 100%), 

So using this system, the two Elimination Finalists would then be;
  EF 1: Hawthorn -v- Richmond
[i.e. 3rd best 'wild-card' ranked team -v- 4th best 'wild-card' ranked team].
  EF 2: Collingwood -v- Adelaide
[i.e. 2nd best 'wild-card' ranked team -v- 5th best 'wild-card' ranked team].


Setting Up The Next Season:
To set up the next seasons conferences**, I would merge the three conferences back into a single (traditional) AFL ladder as at left, and then separate the teams based on that.


Using the combined ladder completed after the home and away season means the next seasons conferences are based on a balanced results card... i.e two games vs teams from the same conference, and one game vs all other teams. I would not advocate incorporating the results from finals, as they may create imbalance.

So from the left, the next seasons' conferences get built again as per the process;
 - 1st, 4th, 7th, 10th, 13th and 16th.
 - 2nd, 5th, 8th, 11th, 14th, and 17th.
 - 3rd, 6th, 9th, 12th, 15th, and 18th.





This creates the new seasons conferences as at right. In this example, there has been
 a) some shuffling of teams into new Conferences,
[i.e. Richmond, Carlton, move into the Barassi Conference from other conferences].

 b) stasis of teams in the same conferences.
[i.e. Melbourne Port Adelaide, Western Bulldogs and West Coast have stayed in the Barassi conference, but now has different teams to play against home and away].

Worth noting in this example, there is not as much mixing of teams as there could be. But that is just the nature of the system being reliant on previous seasons ladder position. Indeed there could even be a case where there was no change year-on-year.



* The end points of 40 and 150 are reached as an indication point from calculations on all the scores of AFL games from 2000 to 2011. The average score over that period is 95pts, and +/- 2 standard deviations of that in 38 to 151. I have not added any 'home ground advantage' as I have not adjusted the draw to balance for that. As before, it can be done, but I just haven't the time to go through and complete it. For the record, there is an average 10 point advantage as the home team over that same period.


** As before, in the "18 Into 22 Wont Fit" post, I advocate a season-on-season re-organising of conferences. This is both a way of distributing games among teams (keeping each season fresh and distinct from others before), as well as to dissipate any concerns of either 'strong conferences / weak conferences' or of developing a 'tiered, promotion/relegation' system. If strong conferences or promotion/relegation systems were introduced, I fear it would mark the beginning of the end for some clubs. 

Who's Got A Semi?

So I heard David Parkin talk up Carlton to win in Subiaco because they "have great form over there against West Coast". At least, I think that is a direct quote from SEN during the week.

So, do they have any form (in corresponding games ...i.e. at home)?


Semi Final 2 - West Coast -v- Carlton

Looking at the last 5 years, its 2-1 to Carlton. Can we take anything from this?

Well in 2010, Carlton won by 4 goals. In the middle of West Coast’s run of 13 losses from 14 games. 1 week after beating Essendon (did they get ahead of themselves?) and 1 week before another Derby (did they think ahead too much?)

The other two games (1W -1L) were in 2007 and ’08… take what you can from those games when a good deal of players, staff and the game itself would have changed.

Here are a few changes...
Carlton: 10 players were involved in the 2007 and 2011 games... Betts, Carrazzo, Gibbs, Houlihan, Murphy, Russell, Scotland, Simpson, Thornton, Walker.
West Coast: 9 players were involved in the 2007 and 2011 games... - Cox, Embley, Glass, Hurn, Kerr, Lynch, Priddis, Rosa, Selwood.
...ALSO: Judd and Kennedy also played twice, but for different teams!
...AND: Carlton were also under the guidance of D. Pagan in 2007 as well.

With the all that, the only good for is recent form, and Carlton lost the Etihad game this year. Parkin... I think there is little in it, other than inspiration for a blog post.
I think the Toasters are different proposition in 2011 than at any time in the last 5 years, so I think at home they should win. Then there is the Judd factor.

Semi Final 1 - Hawthorn -v- Sydney

Hawks v Sydney at the MCG, its 3-1 for the Dawks. Interestingly, Swannies can get past 80 points.

...and the Swans lost to Richmond at the 'G too. But some think that is the catalyst to their late season revival. Could be.
And Hawthorn got them in Sydney too, by 46 points (Sydney only scored 60pts).

Tough one... I think if Buddy plays, the Dawks get up. Without him, the Swans might get home but the Dawks still look likely.
Or maybe I just go with the heart, not the head...
Go Bloods! *

* Disclaimer: Author is a Richmond F.C. member, which has no bearing on this at all. Author's dad is a South man from way back... so now you know.

Sideshow Mick's Last Gleaming

Qualifying Final 1 - Collingwood -v- West Coast

Krusty old Eddie is now so close to bumping off his old mate, Sideshow Mick. But Mick wants to go out with one last bang, so looks to the big (air) show to do so... the finals. He has already been on the Tyrano-Vision, talking about how he doesn't feel the spark for 2012 at Collingwood, and that agreements may have been made under duress (sure to get any employer running). Now to play out the last act... Sideshow Mick v West Coast Eagles, with Bart Buckley waiting in the wings.

1. Collingwood
The 'Elgin Marbles' of the AFL. Slightly dusty. Slightly broken. Have been around forever. Removed form their former home to the 'Bank Centre' (formerly 'Japanese Car Centre') by those hanging on to the dreams of former empire, with 'new money'. If only they were back at their home.

The premiership is theirs to lose in 2011, and maybe 2012 as well. Strong form at the G, with the only chinks in the armor exposed by Geelong. Might be worried again by then, or the Swans (purely as they choke games). Can't see the Hawks or anyone else getting past them.


2. West Coast
And speaking of 'new money', the very AFL example of flaunting it. Back into the finals after being the best playing list to finish last since... for ever?
And what of the best ruck combo in the land... Can the Cox / Naitanui pair make any dent in the finals. No doubts on Cox as hes proven, but will NicNat continue to be the free-spirited player of week-to-week football, or does stage-fright beckon.

West Coast of course strong at home (barring excusing an early season Sydney loss), but the away form is simply rotten... though we are happy they bumped off the Blue Filth.
With the weather the way it is and the Eagles having a reputation (deserved?) of being dry-track bullies... what looked a good match-up last week may be a train wreck now.


Verdict:
Well... its the maggot-bags for mine. Has to be.
I also have a suspicion the Wet Toast might tank it, knowing a home final next week will give them a chance to re-load and go again.