Unfair fixtures for Top 4 contenders

[Its Time for Another 4,051]

22 hours ago, WheeloRatings said:

I guess the issue with that approach (i.e. top 6 playing each other, 7-12 and 13-18 likewise) is that 6th gets a really tough final five rounds and much tougher than the teams in the next group. If 6th is a game ahead of 9th after 17 rounds, they could easily miss finals due to the much tougher final five rounds. 

Good point but this would maybe more relevant for teams in 7th and particularly 8th. Teams around 6th by then are usually in the 8 at the end. The line has to be drawn somewhere. Arguably teams chances of finishing top 2 or top 4 being compromised by an unfair draw, like Geelong getting Nth & Eagles twice is more of a problem. 

[WheeloRatings 900]

3 hours ago, Its Time for Back to Back said:

Good point but this would maybe more relevant for teams in 7th and particularly 8th. Teams around 6th by then are usually in the 8 at the end. The line has to be drawn somewhere. Arguably teams chances of finishing top 2 or top 4 being compromised by an unfair draw, like Geelong getting Nth & Eagles twice is more of a problem. 

It potentially increases the chances of a team "tanking" in round 17. Lose this game and play the last five games against teams below you, or win this game and play the last five games against the best five teams. I'd almost want to see a conference system in place for the last five rounds if that approach was adopted - e.g. the last five rounds are used to rank teams within each group of six teams, but the top 6 teams are set after round 17 and 7-12 can play off for the last two spots.

Alternatively, each team's last five games could be played against teams around them on the ladder but without specifically grouping teams into blocks. E.g. Top plays teams 2-6, 2nd plays against five of teams 1-7, 3rd plays against five of teams 1-8, etc. It can be random but within certain constraints like teams 4-15 need to play at least two teams above them and two teams below them.

Regardless, I agree that the draw/fixture could definitely be improved and Geelong should not be playing four games against the likely 17 & 18!

[Rocknroll 912]

Imagine the team's that have to play us twice

 

Poor buggers

[Sydney_Demon 1,175]

On 5/10/2022 at 11:42 AM, WheeloRatings said:

For what it's worth, here are the average ladder positions for each team's opponents this year, based on (a) last year's ladder position and (b) this year's projected ladder positions from https://squiggle.com.au/ladder/

Of course, ladder position isn't an authoritative ranking of team strength as teams may finish higher because they have an easier fixture and lower because of a harder fixture.

Thanks for this. It’s slightly misleading IMV as the higher ranked the team is the lower the average of their opponents’ positions. i.e. if every team played every other team only once the average ranking of Melbourne’s opponents would be 10 (average of pos. 2 to 18) while North’s would be 9 (average of pos. 1 to 17) with the other teams within that range. Better to show the variance from the starting point. i.e. Melbourne 0.5 higher than average, North 0.4 lower than average, so they have a better draw rather than worse. I also think it should be in order of finishing position after the finals, rather than after H&A, which in Melbourne’s case would make the number slightly lower (2 games against pos. 2, 3, 5, 11 & 17 rather than 2 against pos. 2, 4, 5, 11 & 17).

Melbourne’s number comes down from 9.5 to 9.3 comparing 2021 actual positions with 2022 expected positions which I guess reflects Brisbane, Fremantle & Collingwoods’ improvements more than offsetting Western Bulldogs & Port Adelaide going in the opposite direction.  

In any case, I guess what it does show is that it’s not a huge variation between sides (not surprising given we’re only talking 5 out of 22 games). What it doesn't show is the effect of different travel requirements which for Melbourne is definitely tougher than for most teams. Also I'd argue that averages don't tell the whole picture, variance is equally important (e.g. playing against pos. 6 & 10 produces the same average as playing against 3 & 13 but IMV is a much tougher ask).  

[Sydney_Demon 1,175]

On 5/10/2022 at 10:14 AM, ElDiablo14 said:

Just heard on the radio that the Cats are playing both North and West Coast twice! 4 easy wins there for the geriatric cats.

Do we know if there is any other team from last year's top 4 with such and advantage?

I understand West Coast was competitive last year, but why do you give 2 games against the wooden spoon to any of the top 4?

Yes, Melbourne. We get to play the 2nd bottom team twice and we won the Premiership. Some would argue that Collingwood weren't a true 17th team and North were a true wooden spooner (with possibly some marginal improvement this year) and that the AFL should be predicting where teams should finish in 2022 in creating the draw. I disagree. Teams go in both directions (as can be seen with Melbourne's 5 we play twice: Bulldogs, Brisbane, Port, Fremantle, Collingwood). It's hard enough putting a draw together without including subjective adjustments to ladder positions.

Edited May 12, 2022 by Sydney_Demon

[Sydney_Demon 1,175]

On 5/11/2022 at 2:51 PM, Jaded No More said:

Completely agreed regarding the Dogs 

The worst possible outcome for them this season was playing us round 1, getting out to a lead and then losing comfortably. Psychology I think it ruined them as they realized this season is not going to be one of redemption but again being second best (or worse as it turned out). 

Disagree entirely. Yes, they were beaten comfortably in Round 1 after getting out to a lead. Hardly a surprising result, as we are a better team than the Bulldogs last year and this year. They lose in Round 2 to Carlton by 12 points and should have won that game comfortably (kicked 13.12 to 16.6), beat Sydney by 11 points which should have been 6 or 7 goals (9.17 to 9.6), a bad loss to  Richmond (38 points but they kicked 7.19 to 15.9), thrashed North (meaningless),  lost by a point to Adelaide, beat Essendon by 32 points, lost by 17 points to Port with key players out injured. Form mixed, lost games through bad kicking, injuries. They're not where they are because they are psychologically scarred from a not-unexpected loss to Melbourne in Round 1. They are 3-5 but are a lot better than that. Port are on the way back as well. Having said that, obviously losing 4 points from playing Melbourne hasn't helped the ladder position for a number of teams (8 so far 😄).

[BDA 19,978]

Soft draws may help you make finals but I don’t recall any recent premier that had an armchair ride to a flag. Brisbane had a golden opportunity in 2020 but fluffed it.

[Sydney_Demon 1,175]

On 5/10/2022 at 11:13 AM, WheeloRatings said:

Here are the number of matches each team plays against each other based on last year's final ladder position.

Geelong, Adelaide and Gold Coast are the three teams to play both West Coast and North Melbourne twice this year but I don't think the AFL can really base the fixture on how they think certain teams will perform.

Thanks Wheelo. Really appreciate you providing some meaningful data on this topic. My take on this is that Geelong haven't got a particularly soft draw. They play the teams that finished 2nd & 3rd last year twice, also St Kilda. As I posted previously IMV the AFL shouldn't be adjusting the fixture based on what they think will happen in 2022, but what actually happened in 2021. The AFL use the draw, and obviously the draft, to equalise the comp and I totally support that. Melbourne have been massive benificiaries of this approach. Since the Comp became truly National, we have seen nearly every team become competitive at some stage and premiership droughts have been broken by multiple teams. This is a good thing!
 

On 5/10/2022 at 12:17 PM, Jaded No More said:

There is no point playing easy teams and making finals. If you can't beat top sides, you won't go far in September.

Last year we beat everyone in the top 8. I rather play hard games and test ourselves and our game plan against the best sides heading into finals. What do Geelong get from playing North and West Coast other than a false sH&Aense of achievement? 

Totally disagree. The season is really 2 seasons: H&A and Finals. The object for the 22 rounds is to finish as high on the ladder as possible because that positions you better going into the Finals. So clearly a better draw helps in this regard (not that I think Geelong are being unfairly helped- see above).

Once you're in the Finals, I agree form is crucial., no matter how that is ramped up. I would argue that the demolition we had against Gold Coast in Round 20 last year, and then wins against easy teams West Coast & Adelaide played us into form. Importantly it gave us confidence heading into the Finals and allowed us to be in a position to go top by beating Geelong in Round 23. 

[WheeloRatings 900]

11 hours ago, Sydney_Demon said:

Thanks Wheelo. Really appreciate you providing some meaningful data on this topic. My take on this is that Geelong haven't got a particularly soft draw. They play the teams that finished 2nd & 3rd last year twice, also St Kilda. As I posted previously IMV the AFL shouldn't be adjusting the fixture based on what they think will happen in 2022, but what actually happened in 2021. The AFL use the draw, and obviously the draft, to equalise the comp and I totally support that. Melbourne have been massive benificiaries of this approach. Since the Comp became truly National, we have seen nearly every team become competitive at some stage and premiership droughts have been broken by multiple teams. This is a good thing!

I completely agree with the points you made here and in your other responses. The AFL cannot base the fixture on expected performance.

In relation to the average ladder positions of opponents as a measure of fixture difficulty, I agree that it would be improved by looking at the comparison to the average if they played everyone. I also agree that playing 6 & 10 is not equivalent to playing 3 & 13. I didn't want to suggest that this was in any way an accurate measure of fixture difficulty. There are sophisticated models of "strength of schedule" which take into account the actual quality of opponent and, as importantly, where the match is played.

Given the uneven fixture, ladder positions are definitely not an accurate measure of team strength. Also, there isn't a linear relationship between team quality and ladder position. This year, the difference between 16th and 17th might be equivalent to the difference between 10th and 16th.

In relation to Melbourne benefitting in the past, we have played fewer away state games in our opponent's home state than any other team since 2008.

Matches from 2008 to 2022 by location, excluding 2020

Away state matches in opponent's home state from 2008 to 2022 by state, excluding 2020

[La Dee-vina Comedia 16,381]

59 minutes ago, WheeloRatings said:

I completely agree with the points you made here and in your other responses. The AFL cannot base the fixture on expected performance.

In relation to the average ladder positions of opponents as a measure of fixture difficulty, I agree that it would be improved by looking at the comparison to the average if they played everyone. I also agree that playing 6 & 10 is not equivalent to playing 3 & 13. I didn't want to suggest that this was in any way an accurate measure of fixture difficulty. There are sophisticated models of "strength of schedule" which take into account the actual quality of opponent and, as importantly, where the match is played.

Given the uneven fixture, ladder positions are definitely not an accurate measure of team strength. Also, there isn't a linear relationship between team quality and ladder position. This year, the difference between 16th and 17th might be equivalent to the difference between 10th and 16th.

In relation to Melbourne benefitting in the past, we have played fewer away state games in our opponent's home state than any other team since 2008.

Matches from 2008 to 2022 by location, excluding 2020

Away state matches in opponent's home state from 2008 to 2022 by state, excluding 2020

Fascinating data, but please don't tell the AFL.

Quick question - why have Geelong and Adelaide played one fewer H&A game than every other team (other than GWS and GCS)? I can't recall what might have happened there.

[WheeloRatings 900]

1 minute ago, La Dee-vina Comedia said:

Fascinating data, but please don't tell the AFL.

Quick question - why have Geelong and Adelaide played one fewer H&A game than every other team (other than GWS and GCS)? I can't recall what might have happened there.

Haha yes, I definitely don't want to tell the AFL!

Their game in 2015 was cancelled after Phil Walsh's death:

https://www.abc.net.au/news/2015-07-03/afl-phil-walsh-gillon-mclachlan-adelaide-crows-geelong/6593282

[La Dee-vina Comedia 16,381]

34 minutes ago, WheeloRatings said:

Haha yes, I definitely don't want to tell the AFL!

Their game in 2015 was cancelled after Phil Walsh's death:

https://www.abc.net.au/news/2015-07-03/afl-phil-walsh-gillon-mclachlan-adelaide-crows-geelong/6593282

Thanks. I'd completely forgotten about that most unfortunate tragedy.

[58er 5,591]

6 hours ago, WheeloRatings said:

I completely agree with the points you made here and in your other responses. The AFL cannot base the fixture on expected performance.

In relation to the average ladder positions of opponents as a measure of fixture difficulty, I agree that it would be improved by looking at the comparison to the average if they played everyone. I also agree that playing 6 & 10 is not equivalent to playing 3 & 13. I didn't want to suggest that this was in any way an accurate measure of fixture difficulty. There are sophisticated models of "strength of schedule" which take into account the actual quality of opponent and, as importantly, where the match is played.

Given the uneven fixture, ladder positions are definitely not an accurate measure of team strength. Also, there isn't a linear relationship between team quality and ladder position. This year, the difference between 16th and 17th might be equivalent to the difference between 10th and 16th.

In relation to Melbourne benefitting in the past, we have played fewer away state games in our opponent's home state than any other team since 2008.

Matches from 2008 to 2022 by location, excluding 2020

Away state matches in opponent's home state from 2008 to 2022 by state, excluding 2020

You have omitted our NT games and if you added then we would be about even.  The AFL maybe taking that into account?

Also if ladder positions are not reliable maybe percentage could be used or do you have another solution. Lucky Dip maybe?

[WheeloRatings 900]

3 hours ago, 58er said:

You have omitted our NT games and if you added then we would be about even.  The AFL maybe taking that into account?

Also if ladder positions are not reliable maybe percentage could be used or do you have another solution. Lucky Dip maybe?

I haven't omitted the NT games, they're included in the neutral state games. Maybe the AFL has taken it into account. On the face of it, it would appear that Melbourne hasn't really been negatively impacted by selling home games. 

I wasn't proposing a solution to the fixture, I was merely stating that ladder positions are not necessarily a true reflection of a team's quality over a season, given the inequalities in the fixture, and there isn't a linear relationship either. I was just acknowledging @Sydney_Demon comment that you can't simply look at ladder positions to determine overall difficulty of team's fixture.

[Sydney_Demon 1,175]

On 5/12/2022 at 6:00 PM, Sydney_Demon said:

Also I'd argue that averages don't tell the whole picture, variance is equally important (e.g. playing against pos. 6 & 10 produces the same average as playing against 3 & 13 but IMV is a much tougher ask).  

Just wanted to correct the above. I meant to post the opposite. 6 & 10 I think is generally easier than 3 & 13. The point I was trying to make was that Melbourne's Draw playing 2, 3, 5, 11 & 17 twice is tougher than say 3, 5, 7, 9 & 14 although both produce an average across 22 games of 9.45. 

On 5/12/2022 at 8:22 PM, BDA said:

Soft draws may help you make finals but I don’t recall any recent premier that had an armchair ride to a flag. Brisbane had a golden opportunity in 2020 but fluffed it.

All I was saying was the first bit. I agree no recent premier has had an easy ride to a flag except perhaps West Coast in 2018 but even they had to overcome the requirement of 'hosting' the Grand Final at the MCG, an unfair imposition on all interstate clubs (especially when playing Collingwood, Hawthorn, Melbourne or Richmond who play home games there). 

11 hours ago, WheeloRatings said:

I completely agree with the points you made here and in your other responses. The AFL cannot base the fixture on expected performance.

In relation to the average ladder positions of opponents as a measure of fixture difficulty, I agree that it would be improved by looking at the comparison to the average if they played everyone. I also agree that playing 6 & 10 is not equivalent to playing 3 & 13. I didn't want to suggest that this was in any way an accurate measure of fixture difficulty. There are sophisticated models of "strength of schedule" which take into account the actual quality of opponent and, as importantly, where the match is played.

Given the uneven fixture, ladder positions are definitely not an accurate measure of team strength. Also, there isn't a linear relationship between team quality and ladder position. This year, the difference between 16th and 17th might be equivalent to the difference between 10th and 16th.

In relation to Melbourne benefitting in the past, we have played fewer away state games in our opponent's home state than any other team since 2008.

Matches from 2008 to 2022 by location, excluding 2020

Away state matches in opponent's home state from 2008 to 2022 by state, excluding 2020

I hadn't really thought about ladder position not reflecting strength of teams because of the impact of easier draws occurring due to prior year's form. So if you get 2 games in 2022 against a side that's position is elevated in 2021 due to a soft draw that reflected poor performance in 2020 you're getting an advantage. It all gets horribly complicated!

I totally accept your point and supporting data showing Melbourne's advantageous total number of away games but point out that games against GWS in Canberra (I recall we've had a number of them) are not counted as away games (have we had more than other teams? I suspect looking at the number of games played in Sydney by other teams that we have). You obviously can't compare Victorian & Interstate teams because the latter play away more but have the corresponding advantage of teams from Victoria & other States having to travel to play them at home. You would also expect poorly performed teams (like Melbourne) in that period to have easier travel schedules against (generally) better performing Interstate teams.

I made the point when the Draw first came out that at least for 2022 we have a tough travel schedule. We play non-Melbourne teams 12 games of 22 (when they are 9 of 18 teams). We play Brisbane H&A, Fremantle H&A, but of the remaining 8 games only play 2 in Melbourne (Sydney & GWS) while playing 6 away from Melbourne (Port in Alice Springs,  and Geelong, Port, Adelaide, West Coast,  & Gold Coast Away). So 4 of 12 games at the MCG, and 1 at a neutral venue. I've pointed out before that if Victorian teams need to sell home games like St Kilda to Cairns or Melbourne to Alice Springs it would be fairer if other Victorian teams played them there and Interstate teams had to play their away games in Melbourne rather than at neutral venues. Why should Port who finished 3rd in 2021 get to play Away games against Melbourne & St Kilda in Alice Springs & Cairns rather than at Marvel and the MCG?

[biggestred 4,259]

Obviously which teams improve and fall away year to year cant be predicted with any sense of accuracy in November. 

Id do 17 games and then for the last 5 id have an FA Cup style pull the balls out of the hat to determine rounds 18-23. At least then people cant complain about it being fair or not... but also its a TV event id watch in November when theyre struggling to get some footy content. 

 

[58er 5,591]

9 hours ago, WheeloRatings said:

I haven't omitted the NT games, they're included in the neutral state games. Maybe the AFL has taken it into account. On the face of it, it would appear that Melbourne hasn't really been negatively impacted by selling home games. 

I wasn't proposing a solution to the fixture, I was merely stating that ladder positions are not necessarily a true reflection of a team's quality over a season, given the inequalities in the fixture, and there isn't a linear relationship either. I was just acknowledging @Sydney_Demon comment that you can't simply look at ladder positions to determine overall difficulty of team's fixture.

Sorry but the ladder mostly and generally about right each year. Whether 10th 15 th or 18 th you are graded for the fixture the following year on a designed program formula as much as can be incorporated with local Derby's  Q clashes and showdowns as well as marquee and blockbuster games. 

The Finsls then  sorts out the best as you can't hide any more behind you home advantage like The Cats tend to do. With the higher sides bring the home teams that rewards excellence as a rule. 

I don't  agree that the ladder is not a true reflection. It's W and L that count and we suffered in 2019 because we didn't  win  enough games. There are no asterisks attached to ladders or Flag  winners so the runners are grinnetd and Rups and also tans c an please themselves. 

[Demon17 4,531]

Final 8 now set.  Historically after Round 9 it doesn't change, just the order of the top 8.

 

Thus pies and Dogs are missing the finals this season.