AFL Power Rankings 2014

[Spirit Level]

Here's a suggestion; if you don't like the rankings or robys methods or anything about this thread, don't ****ing read it

 

[mHmmm]

Question Five:
You make a lot of calls about umpiring and its importance on a game. Have you ever umpired a game of Australian Rules?

I umpire nine AFL games every round.

 

[Bunk Moreland]

Eh, I've made of money following Nate Silver's predictions. He knows he's accurate, and he knows that if you follow his odds, you'll win a fair bit.

He doesn't give his away for free.

And I doubt he even bets on them - that's not how they make him money.

It's a basic economic reality - if there's money to be made with something, people will exploit it and protect it.

Roby clearly does not make money with this. He wouldn't give it away if he did.

 

[TheMcGrathMan]

Chiz Interested (if you want) to see your rankings

At least Chiz had Fremantle/Hawthorn as his top 2 before finals last year (only dropping Freo out when they lost to Saints), unlike Roby who from a quick look at the thread had Geelong as the number 1 and tipped them to beat Dockers by 6 goals

 

[Roby]

Roby clearly does not make money with this. He wouldn't give it away if he did.

Won't be for ever, eventually it won't be for free, enjoy it while it lasts.

Had you followed the tips this week you would have made a profit as you would have most weeks with the rankings, with the game tomorrow as a bonus to surplus.

Chiz Interested (if you want) to see your rankings

At least Chiz had Fremantle/Hawthorn as his top 2 before finals last year (only dropping Freo out when they lost to Saints), unlike Roby who from a quick look at the thread had Geelong as the number 1 and tipped them to beat Dockers by 6 goals

Yawn.

Do you guys have nothing better to do than come to a forum around midnight and fabricate lies?

Are Freo the new Sydney?

It appears so. Fremantle (2nd) move to their highest position in their rankings history...

Power rankings round 10, 2013

1 Hawthorn (-)
2 Fremantle (+1)
3 Geelong (-1)
...

Power Rankings Round 12, 2013

1 Fremantle (+1)
2 Hawthorn (-1)
...

Fremantle vs Hawthorn now appears as the most likely Grand Final. Despite the highly intelligent Doppledanger illustrating that the model is no longer function, the rankings are still gong to put $1 on Hawks-Freo GF quinellawith Sportsbet who are offering best odds at $9 still.

These were posted back in June, 2013.

 

[Roby]

Hey Roby, I'm always interested when people put together predictive systems. I have one major issue with your ranking system, an issue that others have touched on in this thread.

Win or lose, these rankings make me want to drink more booze The rankings are about expected performance. So in other words, each team is expected to win or lose by a specific margin depending on the strength of their opponent. Once the predicted margin is calculated, if a team does better than the expected margin they gain ranking points, if they do worse than the expected margin, they lose ranking points. Simple? ~ Roby

I think you need to rework the above.

If a team wins, their position should ALWAYS improve (and vice versa for the losing team). Perhaps they shouldn't gain AS MUCH as they would have if JUST beat a poor team that they should have thrashed, but the fact that they got the win says something, even if they didn't perform as well as expected. Sometimes a team has to win ugly, but at the end of the day, a win is a win.

Say for instance, Hawthorn take Brisbane lightly and Brisbane get off to a flyer and kick the first 4 goals of the game, Hawthorn then have the weight of possession but poor kicking means they've wasted ample opportunities to control the game. Brisbane go into half time 30 points up. Clarko give Hawthorn an almighty spray and they click into gear and manage to win by 20 odd points. Based on your rankings, Hawthorn were probably meant to win by 40, underperformed and only won by 20, so they lose some points. In reality, they gave away a 30 point start and managed to outscore Brisbane by 50 points in a half, showing their true dominance.

Say Hawthorn = 1400 points, and Brisbane = 700 points. And say your rankings involve granting the winning team a bonus of 10% of their opponents worth. So if Hawthorn had performed as expected and won by 40, they would get 70 points for the win. Perhaps instead of getting the full 70 points, they should only get 35 points because they only beat them by half of what they were expected to win by. A system like that would be more accurate imo, than Hawthorn "losing" points for not winning by the expected margin.

Of course, I say the above having no idea of how you calculate your rankings, but it's just my attempt at showing you why I believe your reasoning is a little off. I also think, this may be the reason why teams don't often slide down the ladder as they should, because your system rewards honourable losses. So if you have 2 teams ranked 14 & 15, both losing, however 14 loses by a larger margin than 15, then 15 has the chance of leap frogging 14 even though they're not a better side. 15 should not leap frog 14 just because they lost a little better than expected.

I'd like to hear your thoughts

Since 1996, the team which has finished on top and with the most wins has only won the flag one-third of the time. You could almost make an argument that it is better not to win too many.

At the moment this power ranking system works better as there are more teams, more teams means more low-confident poor teams languishing at the bottoms of the ladder who can't make finals. This equates to a higher number of win percentage for the top teams, in what I call a league expansion win/loss inflation.

Also there is the other fact that in modern times the top teams are analyzed and hunted even more due to the increase in professionalism, media, and technology, therefore their gameplan is studied for breakdown more acutely and extensively by the top contenders. If Hawthorn do not continue to improve, then they are losing ground and by the time they play Fremantle in a final they will be outclassed and outmaneuvered. This is why it is becoming increasingly harder to win back-to-back, especially with equalisation.

So in the end it doesn't really matter if you finish on top, or how many wins, it's how you perform at you're best with the top teams. The longer you can do this for, the better you peform in the finals, or at least more likely to.

 

[Air_Chester]

Since 1996, the team which has finished on top and with the most wins has only won the flag one-third of the time. You could almost make an argument that it is better not to win too many.

Yes, I can definitely see the argument for not wanting to win too many games of football in a football competition.

 

[PowerForGood]

That's it Roby. I think I'll have to have to publish my Power Rankings (the original and still the best) in competition to yours in this thread.

Because there's only one thing that can improve this, and that is turning it into a tipping equivalent of a "West Horsham s**tfight"

 

[raskolnikov]

After Friday night's game Essendon are now wooden spoon favourites and North have soared into premiership favouritism. Am I doing it right Roby?

 

[MrKK]

Yawn.

Do you guys have nothing better to do than come to a forum around midnight and fabricate lies?

Power rankings round 10, 2013

1 Hawthorn (-)
2 Fremantle (+1)
3 Geelong (-1)
...

Power Rankings Round 12, 2013

1 Fremantle (+1)
2 Hawthorn (-1)
...


These were posted back in June, 2013.

Power Rankings Round 23, 2013

1 Hawthorn (+1)
2 Geelong (-1)
3 Fremantle (-)

 

[Richo83]

Power Rankings Round 23, 2013

1 Hawthorn (+1)
2 Geelong (-1)
3 Fremantle (-)

I don't understand why that's a problem for roby. Most people thought that Geelong would beat Fremantle, and would go further. This smacks of revisionism, Fremantle made the grand final so anyone who predicted otherwise is bad at this. You need to rate the system as a whole as well.

 

[Underarm]

Since 1996, the team which has finished on top and with the most wins has only won the flag one-third of the time. You could almost make an argument that it is better not to win too many.

This is the stat that you should be pushing Roby. Teams on top of the Power Rankings win 2/3 times! thats like 33% better!

 

[WhyAlwaysMe]

Won't be for ever, eventually it won't be for free, enjoy it while it lasts.

So people are going to pay to lose money?

 

[Crankyhawk]

That's it Roby. I think I'll have to have to publish my Power Rankings (the original and still the best) in competition to yours in this thread.

Because there's only one thing that can improve this, and that is turning it into a tipping equivalent of a "West Horsham s**tfight"

I'm doing the betting half - to see if noob betting style H2H on home team, no exotics can outperform Robys method.

 

[rickyp]

So people are going to pay to lose money?

No, they'll pay so they can continue to read the best thread on bigfooty. The 'betting' is secondary

 

[PAFCsince1870]

Here's a suggestion; if you don't like the rankings or robys methods or anything about this thread, don't ****ing read it

LOL this thread would have zero replies if the only people who posted were those who agreed with (or even 'liked') Roby's rankings.

 

[Chiz]

Chiz Interested (if you want) to see your rankings

At least Chiz had Fremantle/Hawthorn as his top 2 before finals last year (only dropping Freo out when they lost to Saints), unlike Roby who from a quick look at the thread had Geelong as the number 1 and tipped them to beat Dockers by 6 goals

I'll hopefully get them up in a few weeks. Unfortunately I wasted way too much time in this thread last season trying to explain why my model was far superior to Roby's...

 

[Dirty Bird]

http://www.bigfooty.com/forum/threads/2013-afl-power-rankings.994497/page-89
http://www.bigfooty.com/forum/threads/2012-end-of-season-afl-power-rankings.976012/
http://www.bigfooty.com/forum/threads/2011-end-of-season-afl-power-rankings.879425/

Didn't you go into Debt in 2013 and have to throw more money at it mid season to try and recover?

 

[Dirty Bird]

Here's a suggestion; if you don't like the rankings or robys methods or anything about this thread, don't ****ing read it

The only things people like his is stupidity.
If his rankings were accurate and actually made money, people wouldn't give a shit and we'd just be reading "take it to the punting board"

 

[MWPP]

Are the eagles down to 17th or 18th after that win? GWS are probably 19th

 

[The Falcon Strike]

Are the eagles down to 17th or 18th after that win? GWS are probably 19th

I think we've been downgraded to junk status.

Terrible by W.C, I'd expect bullies into flag favourites and we'd be odds on for the spoon

 

[Crankyhawk]

OK, here goes the noob betting response.

Geelong 1.33
North 1.64
Hawks1.08
Saints 2.45
WCE 1.35

4 of. 5 for a return of $1.21 profit on $5 outlay

 

[Bunk Moreland]

Didn't you go into Debt in 2013 and have to throw more money at it mid season to try and recover?

Hang on what?

Profit turned by simply upping stake amount half way through?

Genius!

 

[Jade]

And these rankings are once again worth the paper they are printed on.

Whats that you say, not printed?

Like I said.

 

[Lensen]

Hang on what?

Profit turned by simply upping stake amount half way through?

Genius!

Yes, but you don't understand. It's a totally legitimate non-martingale strategy to do this because even though you've gone from $2 max bets to $4 bets:

The $4 bet could still lose.

Bad luck.