Why every top 4 prediction is wrong.

[Mead]

Obviously predicting the ladder is in vogue at this time of the year, most people either
a.) Pull a ladder out of their rear end on the basis of their own prejudices or likes, dislikes and gut feelings
b.) Go through and tip game by game using a ladder predictor program.

Virtually every one of those ladders will have Port, Brisbane, St Kilda and Geelong in the top 4. If you're feeling really daring, you might have pushed up a team from around 6th-8th. into the top 4. If you did either of those things, history suggests that you are totally wrong.

Here are some stats since the advent of the top 8 system in 1994.

-no top 4 has ever contained the same teams two years running.

- Only on one occasion have 3 top 4 teams stayed the same. On average, only 1-2 out of the top 4 stay there.

- Every year, at least one team who didn't make the finals the previous year made the top 4.

- There is a 70% chance that at least one team who finished 12th or worse the previous year will make the top 4.

So given that the bouncing about the ladder position shows no signs of altering, what predictions can we make? Unless there is a dramatic departure from the way things have been for the last decade;

-At least one of Port, Geelong ,Brisbane and St Kilda will finish outside the top 4. Most likely 2 or 3 of those teams will drop out of the top 4.

- At least one of Freo, Kangaroos, Carlton, Adelaide, Collingwood, Bulldogs, Hawthorn and Richmond will make the top 4.

So depressing as it sounds, when the wild eyed Collingwood or Richmond fan does their 'totaly unbiersed' ladder prediction which has their side finishing top 4 next year, there's actually more chance that they'll be correct than the reasonable sounding person who tips conservatively and says the top 4 will stay with Port/Bris/St Kilda/Geelong. Weird huh?

 

[liz]

-At least one of Port, Geelong ,Brisbane and St Kilda will finish outside the top 4. Most likely 2 or 3 of those teams will drop out of the top 4.

You could argue a case for any of these teams dropping down, but I suspect Brisbane and Geelong are the most vulnerable. It is feasible that the Lions will have a slow start to the year as they'll be without Brown (on top of losing Lynch) and Black. They'll be there or thereabouts later on but may find it hard to claw back lost games.

Geelong only crept into the top 4 last year and have a long way to go, IMO. The comparative lack of a forward line doesn't make their job any easier. Injury to King or Ottens will make life particularly hard for them, one would think.

- At least one of Freo, Kangaroos, Carlton, Adelaide, Collingwood, Bulldogs, Hawthorn and Richmond will make the top 4.

I find it hard to see any of these teams other than Freo or - if everything goes their way - the Pies making the top 4 in 2005.

 

[Mead]

I find it hard to see any of these teams other than Freo or - if everything goes their way - the Pies making the top 4 in 2005.

Yep as do I, but it will happen.

I daresay most people found it pretty hard to see Melbourne and Brisbane both going from wooden spoons to 4th and 3rd in one year, or the Bulldogs from 15th to 3rd.

I guess what that series of stats shows is just how dependent the ladder is on the luck of the draw and a good injury run. As bizarre as it sounds, a totally unfancied team like Adelaide, Carlton or even Hawthorn or Richmond could get lucky with those things and suddenly be a serious contender.

 

[Stakerz]

Good point!

 

[Slander]

Obviously predicting the ladder is in vogue at this time of the year, most people either
a.) Pull a ladder out of their rear end on the basis of their own prejudices or likes, dislikes and gut feelings
b.) Go through and tip game by game using a ladder predictor program.

Virtually every one of those ladders will have Port, Brisbane, St Kilda and Geelong in the top 4. If you're feeling really daring, you might have pushed up a team from around 6th-8th. into the top 4. If you did either of those things, history suggests that you are totally wrong.

Here are some stats since the advent of the top 8 system in 1994.

-no top 4 has ever contained the same teams two years running.

- Only on one occasion have 3 top 4 teams stayed the same. On average, only 1-2 out of the top 4 stay there.

- Every year, at least one team who didn't make the finals the previous year made the top 4.

- There is a 70% chance that at least one team who finished 12th or worse the previous year will make the top 4.

So given that the bouncing about the ladder position shows no signs of altering, what predictions can we make? Unless there is a dramatic departure from the way things have been for the last decade;

-At least one of Port, Geelong ,Brisbane and St Kilda will finish outside the top 4. Most likely 2 or 3 of those teams will drop out of the top 4.

- At least one of Freo, Kangaroos, Carlton, Adelaide, Collingwood, Bulldogs, Hawthorn and Richmond will make the top 4.

So depressing as it sounds, when the wild eyed Collingwood or Richmond fan does their 'totaly unbiersed' ladder prediction which has their side finishing top 4 next year, there's actually more chance that they'll be correct than the reasonable sounding person who tips conservatively and says the top 4 will stay with Port/Bris/St Kilda/Geelong. Weird huh?

No you've just used a lot of historical/statistical rhetoric which means nothing in terms of current teams' ability no-one is game enough to say their top 4 predictions are gospel because they are just taking an educated guess.
Oh and BTW the Eagles are no chance of finishing top 4, I know it cuts deep and it's probably why you started this thread

 

[Mead]

No you've just used a lot of historical/statistical rhetoric which means nothing in terms of current teams' ability no-one is game enough to say their top 4 predictions are gospel because they are just taking an educated guess.
Oh and BTW the Eagles are no chance of finishing top 4, I know it cuts deep and it's probably why you started this thread

Yes and one day a series of lotto numbers might come up the same two days running, but it isn't likely.

If something has happened for 10 years running, I'd say there's a fair chance it will happen again. A load of historical rhetoric and stats would be the only thing making me predict that no player will kick 200 goals in season 2005, buit I guess its possible that could happen also.

I have this odd sense of deja vu, because iirc I made a similar post to this one at the same time last year, and got a very similar response from people who thought it was ridiculous to suggest that several of Port, Brisbane, Collingwood and Sydney wouldn't finish top 4. How much more absurd that St Kilda and Geelong (11th and 12th) would make the top 4 and go to prelim finals? I guess the trend meant nothing in terms of those teams abilities at that time?
It still happened, and barring a pretty amazing change in history, it will happen again.

Actually as a WC supporter I find the trend a little depressing, because it suggests that a team in the 8 really doesn't have much better chance of making the top 4 than any side no matter what their ladder position. Rest assured, I'm not inclined to go out of my way to find a reason why Freo has a chance to make the top 4!

 

[Altum Volantes]

Freo will make the top 4, lock it in.

 

[m0nty]

I daresay Melbourne's yo-yo form has contributed a lot to this trend. Nevertheless, it's a good point that Mead makes. Brisbane would be my first candidate to drop out of the top 4, followed by Geelong or Port Adelaide. In all three cases, it's due to personnel losses/changes.

Being a Hawthorn supporter, I'd have to say if any bottom four team from 2004 would scrape into the eight, it would be the Hawks. Seriously though, I think Collingwood has a better chance than us in '05.

 

[Two-Bobby]

Mead

I think your post is well-researched, well thought-out and sensible.

If nothing else, it highlights very succinctly how difficult it is basing an upcoming season's performances using LAST YEAR'S data.

2004 is gone and buried - for people to believe that the form of teams going in to 2005 is going to be "similar" I venture to say that the probability is probably stacked against them and highly unlikely. Of course, this doesn't mean that it CAN'T happen that way...but history shows that it is not likely to.

TB

 

[Jabso]

Ladder predictions threads are stupid as everyone gives pretty much last years ladder with small changes and one or two big falls/big rises.
They are not as stupid as these who will win the next 5 premierships threads as anyone can.
Now here is an example of how teams can fluctuate:
Collingwood:
1999: 16th
2000: 15th
2001: 9th
2002: 4th/runners up
2003: 2nd/runners up
2004: 13th

A better example might be Brisbanes last 5/6 years.

 

[Kildonan]

Clubs with a high turnover make assessment of their chances difficult. Their core players are capable of performing well, but depth and or key significant positions are sometimes makeshift changes till youth "grow" into the position.

Richmond, Hawthorn and to some extent Carlton and Bulldogs are examples here. They may be good, but likely to be found out if there is a run with injuries.

Adelaide and Brisbane have an aging squad - with good youth coming through which is as yet underdone. Even Essendon might be seen as similar. These clubs are capable of being the best, but everything has to go their way.

Freo and West Coast are seemingly on the verge and both have expectations of a much better 2005 than 2004. Melbourne are similar - had a great year then faded but showed what they are capable of.

Collingwood can only improve.

Saints fans expect an even better season than last.

Port fans expect back to back.

 

[skywalk750]

-no top 4 has ever contained the same teams two years running.

I think this statistic will be 100% accurate next year as there is no doubt the Eagles will finnish first on the ladder. Heck, theres more chances of getting hit by lightning while having sex with a farm animal than the Eagles not winning the minor premiership.

 

[Blues_Man]

. Heck, theres more chances of getting hit by lightning while having sex with a farm animal than the Eagles not winning the minor premiership.

wow i din't even realise that could happen ..i've gotta be more careful

 

[ok.crows]

As bizarre as it sounds, a totally unfancied team like Adelaide, Carlton or even Hawthorn or Richmond could get lucky with those things and suddenly be a serious contender.

My question would be, why does it even sound bizzarre?

In 1996, Adelaide finished 12th. In 1997, Adelaide finished in the top 4, and even went as far as winning the flag.

Not that this sort of thing happens often, but it does happen. All it takes is a bad year with draw, form & injury, a change of coach and tactics, some fresh blood, and then a luckier run with injury & a better draw. From that a few extra wins can flow, and from those wins confidence can be found, and momentum built.

 

[VANDA]

Ashtray on a motorbike thread.

Albeit an interesting point that is made.

BUT

is there similar statistical interest in the previous Top 2's, 3's, 5's, 6's, 7's, 8's?

Mead your initial post is a compilation of interesting 'stats' that may not equate to too much relevance when examined more closely.

The assumption that you base your future predictions on what ladder placings have unravelled before is inherently flawed.

 

[skipper kelly]

the assumption that you base your future predictions on what ladder placings have unravelled before is inherently flawed.

are you sure you went to the same pubs as me.

 

[Rodion]

having sex with a farm animal

Cue RodgerFox

 

[VANDA]

are you sure you went to the same pubs as me.

Time & tide buddy.

 

[Stiffy_18]

Yep as do I, but it will happen.

I daresay most people found it pretty hard to see Melbourne and Brisbane both going from wooden spoons to 4th and 3rd in one year, or the Bulldogs from 15th to 3rd.

I guess what that series of stats shows is just how dependent the ladder is on the luck of the draw and a good injury run. As bizarre as it sounds, a totally unfancied team like Adelaide, Carlton or even Hawthorn or Richmond could get lucky with those things and suddenly be a serious contender.

I agree and it just goes to show that apart from top 2 teams the reast are pretty even. A good run with injury can see a bottom 4 team of previous year go into top 4 at the end of this year. Simirarly a top 4 team from last year can very easily miss the 8 all together if it gets nasty run with injuries.

Top 4 will not stay the same and one of the teams you mentioned will make the 8 and possibly push for top 4. I for one think WC will finish top 4, possibly top 2 while Brisbane and maybe Geelong will slide from the top 4.

 

[fishbowl]

Hawthorn will make the top 4 guaranteed. Collingwood will finish near the bottom, if they couldn't win a premiership with the team thay had in 2002/3 they will not win one for a while. The loss of Bo Nixon will hurt them. I think the sad loss of Troy Broadbridge will gut Melbourne. Sydney, West Coast and Fremantle are sleepers as they could finish 1st or 16th. Essendon's luck has run out, their ordinary list will finally catch up with them. Richmond won't improve much, Bulldogs are nearly guaranteed the wooden spoon, but will get competition from Kangaroos. Geelong was very lucky to finish where they did, can their luck continue? Has St Kilda gone ahead of themselves? Only time will tell. Port Adelaide have lost their CEO, a tough player, their Melbourne and Adelaide recruiter, an assistant coach and their fitness adviser. Tough luck! Adelaide might fluke a few wins. Which leaves me with Brisbane, they are looking at a tough start to the year.

 

[arrowman]

Ashtray on a motorbike thread.

Not if it helps to hose down some of the wilder ladder predictions around the place

...Mead your initial post is a compilation of interesting 'stats' that may not equate to too much relevance when examined more closely.

The assumption that you base your future predictions on what ladder placings have unravelled before is inherently flawed.

Not flawed at all; it would be if Mead were attempting to predict WHICH teams will make the top 4, but the general view is quite valid.

When you look at the recent history of the AFL ladder, and you see statistical trends (eg changes in the top 4 or top 8) you can extrapolate those trends into the future, provided there is a valid "real-life" reason for those trends in the first place, and it's not just coincidence.

There are many valid reasons why there should be changes in the top 4 (and top 8) every year. Not every club that enters the top 4 does so on the back of a genuinely powerful team (ie can be expected to repeat the process next year); some do it on the back of a good player list, a good run with injuries etc. The latter teams will often drop out the following year, and then (maybe) bounce back up again. Etc, etc.

Using statistics to predict the AFL ladder is like using stats on the stock market - you can confidently predict the general trend, but you can't use the stats to predict individual "winners and losers".

If someone wanted to bet me, for example, that there will be only one change in the top 8 in 2005, I would take their money in a flash. And I would base this decision on history / statistics, and I would be right to do so.

 

[VANDA]

Not if it helps to hose down some of the wilder ladder predictions around the place
Not flawed at all; it would be if Mead were attempting to predict WHICH teams will make the top 4, but the general view is quite valid.

When you look at the recent history of the AFL ladder, and you see statistical trends (eg changes in the top 4 or top 8) you can extrapolate those trends into the future, provided there is a valid "real-life" reason for those trends in the first place, and it's not just coincidence.

There are many valid reasons why there should be changes in the top 4 (and top 8) every year. Not every club that enters the top 4 does so on the back of a genuinely powerful team (ie can be expected to repeat the process next year); some do it on the back of a good player list, a good run with injuries etc. The latter teams will often drop out the following year, and then (maybe) bounce back up again. Etc, etc.

Using statistics to predict the AFL ladder is like using stats on the stock market - you can confidently predict the general trend, but you can't use the stats to predict individual "winners and losers".

If someone wanted to bet me, for example, that there will be only one change in the top 8 in 2005, I would take their money in a flash. And I would base this decision on history / statistics, and I would be right to do so.

So go ahead.............tell us which team will go out of the final 8 & which team will come in.

If you cant do this then it is the same in effect as predicting that their will be 8 teams in the final 8.

Extrapolation from a statistical base IS a risky business. Granted you may be able to predict 'trends' but one that is guaranteed is that sometimes these 'trends' go a completely different way.

Am not seeking an argument here because it has no concrete explanation. In an AFL sense i agree it is highly unlikely that the bootom 4 of 2004 will not make the top 4 in 2005. BUT at this stage of 2005 it is possible.

As a betting man i would be reluctant to make a bet that only 1 team of the 2004 final 8 will slip out of the final 8 in 2005. Statistics or not.

 

[nonchalance]

Using statistics to predict the AFL ladder is like using stats on the stock market - you can confidently predict the general trend, but you can't use the stats to predict individual "winners and losers".

So go ahead.............tell us which team will go out of the final 8 & which team will come in.

I'd like to introduce you to a friend of mine.
His name is 'comprehension'.

If someone wanted to bet me, for example, that there will be only one change in the top 8 in 2005, I would take their money in a flash.

As a betting man i would be reluctant to make a bet that only 1 team of the 2004 final 8 will slip out of the final 8 in 2005. Statistics or not.

Say hi, comprehension.

 

[VANDA]

Two with sauce thanks Nonchalance.

 

[PeteV]

Looks like you've been ripped off, Mead.

http://www.footygoss.com.au/modules.php?op=modload&name=News&file=article&sid=17009