Preview QF1 Geelong v Collingwood Sat Sept 3 2022, 435pm @ MCG

[God Mode]

This is another example of numbers giving a good picture.
Collingwood just scraped into positive numbers for total quarters won for the season 45/88

Geelong has won 67/88 quarters. A high of 75% the most of any team, which suggests while we’ve had a few big quarters our wins are based on consistent effort. Collingwood numbers suggest relying on concentrated moments (quarters). I prefer our model (method). Bring our best or very near and a single quarter won’t be enough to stop us.

It would be interesting to see how many qtrs we won in other seasons, where we've performed poorly in finals.

 

[BotsMaster]

Is
It
Gameday
Yet

It's tomorrow. 4:10pm start.

 

[Badtemperedbadger]

It would be interesting to see how many qtrs we won in other seasons, where we've performed poorly in finals.

Very true.
However I highlighted this as I believe this is by far the best shape, form and fitness wise we have entered a finals series for a very long time.

While many followers think we have stuffed up incredibly over the last decade, I’m not here to debate that, I confidantly believe this is also our best team since the 2011 version, certainly the most even across the board.

We have stars but we rely on, respect, and work for each other. That is a great sign and I’m hoping will carry the day.

I loved our 2000’s era, but I also have been stung by 89,92, 94, 95, 2008, and 2020.
One thing I agree with Scott on, is history (in competition) has little to do with current proceedings.

Otherwise 2007 et al would not have happened.

Its not going to be easy, and we have a good shot at succeeding.

 

[Footy_Fan2007]

Anyone in favour of the pre-finals bye needs only to read this thread. I mean, we're seriously discussing the etymology of the phrase 's**t the bed'. Worse, it has been a bloody interesting discussion!

But FFS, what do the AFL think this is, bush week? We're not here to * spiders - get the finals under way! Interest is running dry as a dead dingo's donger right now.

This.
I need game day to hurry the **** up.
Or some finals footy at least.

 

[Pollywaffle]

I hate bye weeks.

Please refer to your nickname

 

[Footy_Fan2007]

Looks like an early prediction of rain next Saturday unfortunately.

 

[And_ROOS]

Looks like an early prediction of rain next Saturday unfortunately.

Won't matter, we will be taking Stanley in as our ruck and Blicavs will do his thing.
Cameron doesn't play tall when he gets on his bike up the ground, and while Hawkins is a "tall" he is incredible at ground level.

The only thing I am not aware of, how does SDK go in the wet? No chance he gets dropped, but can't remember the last game he played in shit conditions.

Our best 22 has the right mix this year but if its shitty, wet conditions we are going to potentially see Atkins have a field day in midfield, and Close/Miers/Stengle run amok in the forward half and Stewart probably MOTM off half back.

 

[God Mode]

Very true.
However I highlighted this as I believe this is by far the best shape, form and fitness wise we have entered a finals series for a very long time.

While many followers think we have stuffed up incredibly over the last decade, I’m not here to debate that, I confidantly believe this is also our best team since the 2011 version, certainly the most even across the board.

We have stars but we rely on, respect, and work for each other. That is a great sign and I’m hoping will carry the day.

I loved our 2000’s era, but I also have been stung by 89,92, 94, 95, 2008, and 2020.
One thing I agree with Scott on, is history (in competition) has little to do with current proceedings.

Otherwise 2007 et al would not have happened.

Its not going to be easy, and we have a good shot at succeeding.

I couldn't agree more. I would be hoping to find that our qtrs won in previous seasons had been worse in an attempt to find some more reassurance about this year's finals series. I think we are in good shape regardless, but the scars do run deep.

 

[Landgraft]

Just a bit of fun, isn't it? I thought there'd been some amusing contributions over the last few pages. I assumed no one was taking the last few pages too seriously...



It really depends how you look at it. Collingwood were bottom 2 last year (probably a better team than that) and top 4 this year on the strangest season record I've ever seen. 104% and 14 games decided by 13 points or less is insane enough. That they won 80% of those close games is incredible (total games won by 12 points or less and total games won by single digits both VFL / AFL records). That their record at the end of it was 16-6 is almost beyond comprehension.

Bottom 2 was probably not indicative of their true ability last season, and top 4 is probably a bit inflated here. But you have to win enough games to be top 4, and they did. So they are. And winning 3 finals is definitely not beyond them.

And if the 3 they had lost by 13 points or less were reversed (which includes our ridiculous comeback), they would be top at 19-3 and around 104%. Madness.

16-6 is an exceptional season record. And they managed to find a pathway to that many wins kicking 1.04 points for every point they conceded. In EPL, MLB, NFL, NBA, AFL, NRL... every sport where a higher score wins, you rarely see a team so high on the table with such a neutral for and against.

Anyway, I disagree on the "if we can't beat the Pies, we are no chance" thing. Anything can happen on the day, and we could easily lose the QF and then win the next 3. But I think I get the intent behind what you mean, and agree in that sense if I'm right on how I read it.

It's a game we should win, because this year we are a much better side than Collingwood on the comparison of our seasons.

But on any day the best team in the comp can lose to the worst team... should they get us next weekend, it may just be one of those days and we will still very much be in it this year.

There's a documentary that (I think) is still available on Netflix called AlphaGo - this is the trailer. Brilliant movie for a number of reasons, but one of the smaller points it raises is the way that people will intuitively use margin/score as a proxy for how likely win or loss is - and how that just isn't remotely the case. AlphaGo would happily concede points that a human player wouldn't, in order to enable sequences that it believed gave it a greater eventual chance of victory.

Now all of that is to note that I don't think that Go is a great thing to compare to AFL footy, which has so many moving parts and inherent asymmetry and is also a contact sport rather than an ancient abstract boardgame. But at the same time you can't just think of teams as trying to continuously attack, score and dominate from every field position and game state - it would be great if that were possible, but opportunity cost and the (vaguely) equalised state of the competition stops it from being a realistic proposition. Collingwood have talked a fair bit about training scenarios more than other teams, but I wouldn't be surprised if the overall plan is deeper than that and they're breaking the game down into sequences that they think can be run in combination to produce favourable outcomes, even when losing in various key statistics.

That or they haven't thought anywhere near as deeply about this and are just lucky on an unreasonable scale (million to one events happen more than you might think, because the world is big and constantly happening!). But I do think that anyone can do much worse in a bye week than to watch AlphaGo if they're interested in AI, or the more fundamental game theory side of competition, or if they just like watching a man be dunked on remorselessly by the thematic equivalent of the dog being allowed to play basketball.

 

[Reppie]

We will win easily.

 

[Nankervis brothers]

I believe if we win 3 finals without a loss, our win streak of 16 would exceed anything we put together in 2007 - 2011? I think 15 was our best in that period.

Yes, if we win the next three, it betters the streaks of 2007 and 2008. Both those ended at 15 wins, 2008 on grand final day. Like in 2008, we will enter the finals on 13 wins.

Actually, if we win the next three, it's the second-longest streak in the club's entire VFA/VFL/AFL history.

 

[Moley Russels Wart]

Yes, if we win the next three, it betters the streaks of 2007 and 2008. Both those ended at 15 wins, 2008 on grand final day. Like in 2008, we will enter the finals on 13 wins.

Actually, if we win the next three, it's the second-longest streak in the club's entire VFA/VFL/AFL history.

what's the longest?

 

[barmy44]

what's the longest?

23 is the Geelong and VFL/AFL record winning streak across 1952-53...

26 is the longest non-losing streak, also held by Geelong - same win streak, but with 2 wins and a draw before it.

 

[barmy44]

There's a documentary that (I think) is still available on Netflix called AlphaGo - this is the trailer. Brilliant movie for a number of reasons, but one of the smaller points it raises is the way that people will intuitively use margin/score as a proxy for how likely win or loss is - and how that just isn't remotely the case. AlphaGo would happily concede points that a human player wouldn't, in order to enable sequences that it believed gave it a greater eventual chance of victory.

Now all of that is to note that I don't think that Go is a great thing to compare to AFL footy, which has so many moving parts and inherent asymmetry and is also a contact sport rather than an ancient abstract boardgame. But at the same time you can't just think of teams as trying to continuously attack, score and dominate from every field position and game state - it would be great if that were possible, but opportunity cost and the (vaguely) equalised state of the competition stops it from being a realistic proposition. Collingwood have talked a fair bit about training scenarios more than other teams, but I wouldn't be surprised if the overall plan is deeper than that and they're breaking the game down into sequences that they think can be run in combination to produce favourable outcomes, even when losing in various key statistics.

That or they haven't thought anywhere near as deeply about this and are just lucky on an unreasonable scale (million to one events happen more than you might think, because the world is big and constantly happening!). But I do think that anyone can do much worse in a bye week than to watch AlphaGo if they're interested in AI, or the more fundamental game theory side of competition, or if they just like watching a man be dunked on remorselessly by the thematic equivalent of the dog being allowed to play basketball.

I have read about AlphaGo a long time ago... that doco sounds interesting, I will try remember to watch it soon.

I may have misunderstood some of your post, so sorry if my reply is a bit off the mark.

I do think the boardgame and AFL cannot be compared in this sense. Scenario specific training is all well and good, but I'd say the game itself is far too dynamic to even begin to comprehend how it could be broken down in that way, let alone then plan a way to utilise that to produce favourable outcomes on balance more often than not and then communicate and implement that plan. That would surely take some chaos theory wizadry I'm not sure is possible.

Collingwoods close game streak is impressive, but also largely a season reliant on simply a lot of results falling their way by chance (IMO, of course).

I have been reading a lot about close games, and how they are largely - in the final minute or two, with a kick in it - chance outcomes. I've posted a number of links recently - happy to again, if you're interested.

Training to have games close and then win most of them... I don't know, doesn't seem to be maximising your chances. Winning 80% of 14 close games is crazy. If they replicate that over the next 3 or 4 seasons, their ability to win close games at better than chance due to the execution of a well drilled plan would have some merit. They are just as likely to go at 20% in close ones next year, and in far less than 14.

But no team does - the 3-peat Hawthorn sides were great from 2012 - 2016. But their results in close games in each year yo-yoed from woeful to perfect. Often, good sides are good sides and bad sides are bad sides across many seasons. I don't think there has ever been a statistically significant "good in close games" team across multiple seasons. Ever.

So planning to keep them close and win them in the end seems like planning to get rich by getting 10 consecutive spins right on the Roulette wheel. Not so outrageous, but that kind of thing.

Too much can go wrong, and your whole season is cooked.

That's why I brought up % and scores. Good teams just win big.

If Collingwood had a neutral points for and against because they won 16 games by 40 points but also got absolutely belted in their 6 games, that would make a bit more sense.

But to be +60 in points for over 22 games with 14 being decided by 2 goals or less (and winning 80% of those), and win 16 overall - that's not something you can plan for. And if you do plan for it, even just actually executing and pulling it off is so insanely remote.

In contrast, we are about +650 points for with only 2 more wins.

That's what I meant by the narrow path. Collingwood had to do everything perfectly to end up 16-6. A few things go wrong (Carlton had 20 - 11 i50s in the last quarter against them, and kicked 6 points), and they are out of the 8.

Most other sides had so much more wiggle room in their seasons. If our 50/50 results flipped, we are still probably top 4. if theirs are flipped, they are bottom 6.

So I have no doubt teams have methods and plans to increase their likelihood of winning... but I can't see how Collingwood can square their season record with that in any logical way.

Next season they are just as likely to be 8-14 at about 100%. Or more likely 11-11.

So I get what you're saying, and I think it is really interesting and probably in play in some ways in most clubs. But if I understood, and it was kind of "only have to win them all by 1 point", then that is simply an unsustainable way to try and have successful seasons.

I think a number of the 80% close games they won (7 of their last 8 games were by 7 points or less, and they won them all - some where the opposition played themselves out of it as much as Collingwood won them), it was a unique season record that, and I've said this before, they'd struggle to replicate if they played the season 100 more times.

 

[The Unforgiven]

I will ask a very simple question?
How are people picking Atkins as a possible sub?

All due respect I find this ludicrous, in finals heat I want this beast on the ground not cooling his heels on the bench waiting on a possible injury.

I can see it now 2 minutes left Geelong desperate for a clearance or at least split the difference and the Axe is not even playing.

Surely by now people cannot be underrating his importance?

Surely after the game versus Port Adelaide, he couldn't possibly be left out of the side? Surely...

 

[And_ROOS]

Surely after the game versus Port Adelaide, he couldn't possibly be left out of the side? Surely...

I think in a final when the heat will be massive and it will be a midfield battle, we would be absolutely daft NOT having tackling weapons in Parfitt and Atkins in the middle. Especially Atkins who this year loves the contest and lifts in all our big games.
Flip side, nothing scares me more than watching O'Connor indecisively bring the ball out of defense and then not do much with it.

 

[Tristan34]

It would be interesting to see how many qtrs we won in other seasons, where we've performed poorly in finals.

Previous seasons we had star players on one end and very average players on the other. Geelong are more rounded this year, so we are a great chance to win it IMO..

 

[you pick one]

KissStephanie where are you we mis you.

 

[winty]

KissStephanie where are you we mis you.

I can still remember them saying we only won a game in 2011 because we were kicking points instead of goals. There's some logic there somewhere I guess.

 

[Landgraft]

I have read about AlphaGo a long time ago... that doco sounds interesting, I will try remember to watch it soon.

I may have misunderstood some of your post, so sorry if my reply is a bit off the mark.

I do think the boardgame and AFL cannot be compared in this sense. Scenario specific training is all well and good, but I'd say the game itself is far too dynamic to even begin to comprehend how it could be broken down in that way, let alone then plan a way to utilise that to produce favourable outcomes on balance more often than not and then communicate and implement that plan. That would surely take some chaos theory wizadry I'm not sure is possible.

Collingwoods close game streak is impressive, but also largely a season reliant on simply a lot of results falling their way by chance (IMO, of course).

I have been reading a lot about close games, and how they are largely - in the final minute or two, with a kick in it - chance outcomes. I've posted a number of links recently - happy to again, if you're interested.

Training to have games close and then win most of them... I don't know, doesn't seem to be maximising your chances. Winning 80% of 14 close games is crazy. If they replicate that over the next 3 or 4 seasons, their ability to win close games at better than chance due to the execution of a well drilled plan would have some merit. They are just as likely to go at 20% in close ones next year, and in far less than 14.

But no team does - the 3-peat Hawthorn sides were great from 2012 - 2016. But their results in close games in each year yo-yoed from woeful to perfect. Often, good sides are good sides and bad sides are bad sides across many seasons. I don't think there has ever been a statistically significant "good in close games" team across multiple seasons. Ever.

So planning to keep them close and win them in the end seems like planning to get reach by getting 10 spins right on the Roulette wheel. Not so outrageous, but that kind of thing.

Too much can go wrong, and your whole season is cooked.

That's why I brought up % and scores. Good teams just win big.

If Collingwood had a neutral points for and against because they won 16 games by 40 points but also got absolutely belted in their 6 games, that would make a bit more sense.

But to be +60 in points for over 22 games with 14 being decided by 2 goals or less (and winning 80% of those), and win 16 overall - that's not something you can plan for. And if you do plan for it, even just actually executing and pulling it off is so insanely remote.

In contrast, we are about +650 points for with only 2 more wins.

That's what I meant by the narrow path. Collingwood had to do everything perfectly to end up 16-6. A few things go wrong (Carlton had 20 - 11 i50s in the last quarter against them, and kicked 6 points), and they are out of the 8.

Most other sides had so much more wiggle room in their seasons. If our 50/50 results flipped, we are still probably top 4. if theirs are flipped, they are bottom 6.

So I have no doubt teams have methods and plans to increase their likelihood of winning... but I can't see how Collingwood can square their season record with that in any logical way.

Next season they are just as likely to be 8-14 at about 100%. Or more likely 12-12.

So I get what you're saying, and I think it is really interesting and probably in play in some ways in most clubs. But if I understood, and it was kind of "only have to win them all by 1 point", then that is simply an unsustainable way to try and have successful seasons.

I think a number of the 80% close games they won (7 of their last 8 games were by 7 points or less, and they won them all - some where the opposition played themselves out of it as much as Collingwood won them), it was a unique season record that, and I've said this before, they'd struggle to replicate if they played the season 100 more times.

To be honest with how many moving parts a single footy team has, let alone trying to accomodate for the opposition, it's incredible that there are any recognisable game plans/styles that are consistently implemented over the course of the season. Extremely complex systems processing a ridiculous array of information to make split second decisions, all for a game of football...

But I do think that if you just take the raw score view then you're at risk of missing something. Obviously some of what I said is taken to the most absurd length possible, but I think that if you are a Collingwood and you have this high degree of efficiency both in attack and defence then you might not be so concerned about volume of possession going to the opposition, and tight situations become more comfortable because you have a system that is better at seizing or denying critical opportunities. I definitely don't buy into the idea that all of those games were effectively 50/50 situations, even though they were more susceptible to small freak events than if they had built a bigger margin. And some of them would still have been, or near enough, of course! But as someone who has always enjoyed building 'skews' for particular match-ups or game states, I do like to speculate and theorise.

 

[Biscut]

Also, the inverse - statistically improbable / "impossible" events happen every day.

Given the age and complexity of our Universe, events that are one in a quadrillion will happen almost routinely "somewhere". Even just on Earth. One in a million events on an individual scale never happen... worldwide they happen everywhere, every day.

To tie it back (maybe loosely) to the topic at hand - Geelong's "finals record".

We all get a bit nervous, because Geelong amd Scott tend to "s**t the bed" in finals.

The large sample size (wins from all games played) gets ignored for the small sample size (wins from finals).

It is not THAT big a stretch to say that Geelong may have just hit a bad string of "luck" across multiple games. Human minds really want to see patterns in randomness.

So we assume Scott and Geelong forget the game of footy as soon as finals start.

Is it that "finals are different"? That's the romantic view. And maybe they are? There's definitely an apparent pattern to them.

Or have Geelong just experienced a run of unlikely / bad outcomes across multiple tiny samples removed once more across different years that we then conflate?

Given how many close / improbable wins we had in years like 2014 / 2017 (two wins as a result of opponents missing shots after the siren coukd have had us 7th, all 5 close games reversed we were maybe 12th)... our top 4 position possibly flattered us.

Collingwood somehow scrambled 16 wins by scoring 1.043 points for every point conceded. The path to that number of wins from so many close games is incredibly narrow. If you were given the % of 104 and asked to wager on how wins the team had, you'd never guess 16. 10-12 seems right. So similar to us in 2014 / 2017, Collingwood are probably a "lucky" top 4 team.

But you have to get there, and if they win 3 more games their "luck" or unique season / % becomes irrelevant.

We won 18 games scoring 1.44 points for every point conceded.

But our finals record... you know, if these two teams played 100 games, we will beat Collingwood in the majority. But next weekend could be one of those minority that condense into reality.

And then our "better team / season" becomes irrelevant.

It will confirm the Scott and Geelong finals theory in most minds.

Because once you have an outcome - however "improbable" or "unlucky" - that is the what ultimately happened.

And maybe it is true...

Plus those dang byes.

Here’s one for you. Maybe, just maybe the cats of the past were just not good enough to get it done in September. I think this year is different.

 

[thecatattack]

Here’s one for you. Maybe, just maybe the cats of the past were just not good enough to get it done in September. I think this year is different.

Agreed. Sure there are some seasons we were pretty damn close and maybe c(sh)ould've gone all the way. But there were reasons we didn't. Be it injuries, personal, gameplan etc. We're a lot better in all those aspects this year, IMO, and this is by far our best chance.

 

[barmy44]

To be honest with how many moving parts a single footy team has, let alone trying to accomodate for the opposition, it's incredible that there are any recognisable game plans/styles that are consistently implemented over the course of the season. Extremely complex systems processing a ridiculous array of information to make split second decisions, all for a game of football...

But I do think that if you just take the raw score view then you're at risk of missing something. Obviously some of what I said is taken to the most absurd length possible, but I think that if you are a Collingwood and you have this high degree of efficiency both in attack and defence then you might not be so concerned about volume of possession going to the opposition, and tight situations become more comfortable because you have a system that is better at seizing or denying critical opportunities. I definitely don't buy into the idea that all of those games were effectively 50/50 situations, even though they were more susceptible to small freak events than if they had built a bigger margin. And some of them would still have been, or near enough, of course! But as someone who has always enjoyed building 'skews' for particular match-ups or game states, I do like to speculate and theorise.

Yeah, maybe.

Like I said, if that is their training / plan to keep it tight and deny / capitalise, then we should see a few seasons in a row of lots of close games and a high win % in them.

The problem is, of course, that Collingwood really shouldn't have won some of those games. A system that relies on such fine margins in a game involving 36 plays on a 360° field that covers a few hundred square metres with a ball that could bounce in any direction... it's not sustainable or reliable.

I'm sure they have trained for close game situations (most teams would), but no team in history has sustainably "been clutch in the close ones".

Their results over the next few seasons may assist in determining if its plan or luck.

I'm definitely not saying theybare a rubbish team this year or ONLY there from luck. I'm just saying they had a season that is entirely unique, and their win record is in part due to more wins falling their way than chance would suggest.

Raw score isn't my focus. I'm not saying footy teams should be trying to score in every situation.

But really good teams tend to score more than they concede regularly. Across all sports.

This Collingwood team didn't across 14 games. And yet they won 16. Never been done before like this.

If it's a system, the next few seasons should provide similar results.

If they don't - then this season involved a lot of luck falling their way, or their system treads such a fine line it's almost a bad one, or both.

Either way, % only matters to break ties and they won 16 games which is enough for top 4.

So they're top 4! And I remember what they did to us recently from 4th. So from here, it's about what happens on the day.

Also, I don't think I'm some guru with all the answers so I hope it doesn't seem like I'm talking down to you...

Just find this stuff super interesting (as well as your views on it!) and like thinking about it / discussing it.

 

[barmy44]

Here’s one for you. Maybe, just maybe the cats of the past were just not good enough to get it done in September. I think this year is different.

Totally! I've said similar on here myself recently.

I even mentioned 2014 and 2017 above - both years I think our position after H&A flattered us.

The complicating factor is it's hard to say we weren't good enough when we have been perenial finalists / often top 4, and beating other finalists throughout the seasons.

But yeah, I'm definitely not saying it's all just "luck", and we have in some recent seasons simply not been up to it in the finals.

 

[Heckler]

There's a documentary that (I think) is still available on Netflix called AlphaGo - this is the trailer. Brilliant movie for a number of reasons, but one of the smaller points it raises is the way that people will intuitively use margin/score as a proxy for how likely win or loss is - and how that just isn't remotely the case. AlphaGo would happily concede points that a human player wouldn't, in order to enable sequences that it believed gave it a greater eventual chance of victory.

Now all of that is to note that I don't think that Go is a great thing to compare to AFL footy, which has so many moving parts and inherent asymmetry and is also a contact sport rather than an ancient abstract boardgame. But at the same time you can't just think of teams as trying to continuously attack, score and dominate from every field position and game state - it would be great if that were possible, but opportunity cost and the (vaguely) equalised state of the competition stops it from being a realistic proposition. Collingwood have talked a fair bit about training scenarios more than other teams, but I wouldn't be surprised if the overall plan is deeper than that and they're breaking the game down into sequences that they think can be run in combination to produce favourable outcomes, even when losing in various key statistics.

That or they haven't thought anywhere near as deeply about this and are just lucky on an unreasonable scale (million to one events happen more than you might think, because the world is big and constantly happening!). But I do think that anyone can do much worse in a bye week than to watch AlphaGo if they're interested in AI, or the more fundamental game theory side of competition, or if they just like watching a man be dunked on remorselessly by the thematic equivalent of the dog being allowed to play basketball.

Check the Link... that one heads under Canberra.