AFL Power Rankings - Round 10

This is the week where the relative part of "relative adjusted net margin" comes into focus. West Coast getting belted by Collingwood is still a better relative performance than Sydney's narrow win over North Melbourne, so the Eagles move up to sixth. Similarly, the Bulldogs' loss to Hawthorn is not as bad in a relative sense as Fremantle's loss to St. Kilda and Adelaide's loss to Brisbane, so the Dogs move up to eleventh.

1 (Last week 1) Collingwood 40.2 (Last week 41.8)
2 (2) Geelong 31.5 (30.1)
3 (3) Hawthorn 19.7 (18.5)
4 (4) Carlton 15.9 (13.4)
5 (5) St. Kilda 6.7 (3.5)
6 (8) West Coast 2.7 (1.4)
7 (7) Essendon 2.2 (2.2)
8 (6) Sydney 2.0 (2.3)
9 (9) Melbourne -6.0 (-3.9)
10 (12) North Melbourne -8.1 (-8.3)
11 (13) Western Bulldogs -11.1 (-9.7)
12 (10) Adelaide -11.1 (-5.3)
13 (11) Fremantle -11.9 (-8.1)
14 (14) Richmond -17.5 (-15.6)
15 (15) Brisbane -19.6 (-24.8)
16 (16) Port Adelaide -27.9 (-29.6)
17 (17) Gold Coast -49.4 (-48.2)

LeBron vs Jordan

OK, right now I'd say Michael Jordan was better than LeBron James is, but there's one argument that I never want to hear again for why Jordan was better, which is that LeBron has teamed up with Dwyane Wade and Chris Bosh to try and win a championship. Jordan had great teammates in Scottie Pippen and Horace Grant, and then Pippen and Dennis Rodman. The difference is this: Jordan's front office in Chicago got him his great teammates, LeBron had to leave Cleveland to find his. Yes, the 'taking my talents to South Beach' quote does grate, but it doesn't impact how effective LeBron is on a basketball court. If anyone wants to say Jordan was better than LeBron is, find another argument (and there's plenty of other arguments that one can use).

Thus ends the rant.

So Where Are the Cuban Mavs On The List of Most Memorable Non-Title Winning NBA Teams Now?

Up to No. 4 I reckon, although of course if they win the title they fall off the list entirely. And if Miami don't win the title, they knock out the 2004 Lakers for the No. 10 spot.

More Wins = More Brownlow Votes?

Yes they do, but the effect hasn't been uniform across clubs. In particular, the newer clubs have considerably higher correlations between average votes per match and winning percentages across home-and-away seasons (excluding those seasons in which the 3-2-1 system of voting was not in place).

Brisbane Bears/Lions: 0.97
West Coast: 0.96
Adelaide: 0.95
Port Adelaide: 0.92
Fremantle: 0.91
St. Kilda: 0.81
Hawthorn: 0.78
Richmond: 0.75
Bulldogs: 0.74
Geelong: 0.73
Carlton: 0.73
Essendon: 0.73
Sydney: 0.72
North Melbourne: 0.67
Melbourne: 0.66
Collingwood: 0.61
Fitzroy: 0.59

The likely explanation for this is that votes have become more highly correlated with wins over time.

AFL Power Rankings - Round 9

Most interesting week this year in terms of movement in the rankings. West Coast and the Bulldogs essentially swap spots after the Eagles toasted the Dogs. The Eagles, remember, were dead last when the season began. The Saints jump back to fifth after their win, but are still a fair way off the top four teams, which have been settled for a while now.

1 (Last week 1) Collingwood 41.8 (Last week 43.8)
2 (2) Geelong 30.1 (31.8)
3 (3) Hawthorn 18.5 (16.5)
4 (4) Carlton 13.4 (12.3)
5 (7) St. Kilda 3.5 (2.6)
6 (5) Sydney 2.3 (6.2)
7 (6) Essendon 2.2 (4.0)
8 (12) West Coast 1.4 (-9.6)
9 (9) Melbourne -3.9 (-3.1)
10 (10) Adelaide -5.3 (-6.8)
11 (13) Fremantle -8.1 (-11.9)
12 (11) North Melbourne -8.3 (-7.5)
13 (8) Western Bulldogs -9.7 (2.0)
14 (14) Richmond -15.6 (-17.9)
15 (16) Brisbane -24.8 (-25.8)
16 (15) Port Adelaide -29.6 (-25.5)
17 (17) Gold Coast -48.2 (-47.4)

The Overvaluation of Rings in Basketball

Most basketball fans would have heard this one before: Player X has won more (championship) rings than Player Y, therefore Player X is better. Certainly the number of rings that a player has won gives some indication to how good they are - no-one wins a championship, let alone multiple championships, purely through luck. But as a method for evaluating players they are vastly over-valued.

Now I think it's true that in basketball one or two or three players can have a much larger influence on the outcome than a lot of other sports. So the amount of games a player wins does say quite a bit about a player's ability. The fact that Jordan, Magic, Bird, Tim Duncan, Shaq and Kobe have all been major contributors to teams that won championships indicates they were among the best players than the NBA has ever produced - you would have a very hard time arguing otherwise. But the difference in ability between them and a few of the players who did not win championships is not as large as often thought, and in some cases may not even exist at all.

The first reason is that a lot of basketball playoff games are very close, and there was a substantial probability that they could have turned out differently to how they did. Consider Kobe for instance - three of his championship runs involved a Game 7 that was close to going the other way: 2000 (against the Blazers), 2002 (against the Kings), and 2009 (against the Celtics). Had the Lakers lost all three he would have won "only" two rings to date. Now I'm not saying we shouldn't think of him any differently at all because he has won five rings: his team was able to win those games after all. What I'm saying is that we shouldn't think that much differently of him whether he ended up winning two championships or five, because the margin between him winning two or five championships was very slim.

The second reason is that, as dominant as the best players are, whether you win a championship or not depends to a considerable degree on your teammates. If Kobe had been an LA Clipper instead of an LA Laker, would he have won a single ring? Unlikely. But the wide disparity in results would barely be any reflection on him, rather it would be a reflection of the respective teams around him. Conversely, put Tracy McGrady (at his peak) on those Laker teams and he would have had a great chance to pick up a few rings by now. It should never be argued that a player wasn't great simply because he never won a ring. What that means is his team wasn't great. If you want to argue a player was or wasn't great, you have to have more of an argument that the number of rings he has won.

Yes it is great to win and it sucks to lose. But don't let the huge difference in emotion fool you into thinking there is a huge difference in ability. The difference is often not as large as it seems.

AFL Power Rankings - Week 8

There were a few unexpected results this week, with Geelong beating Collingwood, North Melbourne beating Melbourne, and West Coast beating Fremantle. St. Kilda fall to 7th after another loss, while Fremantle fall to 13th.

1 (Last week 1) Collingwood 43.8 (Last week 45.3)
2 (2) Geelong 31.8 (30.2)
3 (3) Hawthorn 16.5 (15.5)
4 (4) Carlton 12.3 (12.1)
5 (6) Sydney 6.2 (4.1)
6 (7) Essendon 4.0 (1.2)
7 (5) St. Kilda 2.6 (4.7)
8 (8) Western Bulldogs 2.0 (0.8)
9 (9) Melbourne -3.1 (0.3)
10 (10) Adelaide -6.8 (-6.7)
11 (12) North Melbourne -7.5 (-11.3)
12 (13) West Coast -9.6 (-13.8)
13 (11) Fremantle -11.9 (-9.6)
14 (14) Richmond -17.9 (-16.8)
15 (16) Port Adelaide -25.5 (-23.8)
16 (15) Brisbane -25.8 (-22.7)
17 (17) Gold Coast -47.4 (-49.3)

AFL Power Rankings - Round 7

All the action is around the middle this week, as Melbourne rebound back into ninth, while the Crows and Dockers take sizable hits in terms of ranking points. In terms of next week, even if Geelong beat the Magpies it's highly, highly unlikely they'll take the top spot, but it will close the gap between them.

1 (Last week 1) Collingwood 45.3 (Last week 46.1)
2 (2) Geelong 30.2 (29.6)
3 (3) Hawthorn 15.5 (14.2)
4 (4) Carlton 12.1 (10.9)
5 (5) St. Kilda 4.7 (5.3)
6 (6) Sydney 4.1 (2.4)
7 (8) Essendon 1.2 (2.2)
8 (7) Western Bulldogs 0.8 (2.4)
9 (11) Melbourne 0.3 (-8.2)
10 (9) Adelaide -6.7 (0.8)
11 (10) Fremantle -8.8 (-3.9)
12 (12) North Melbourne -11.3 (-10.5)
13 (13) West Coast -13.8 (-14.2)
14 (14) Richmond -16.8 (-20.6)
15 (15) Brisbane -22.7 (-21.3)
16 (16) Port Adelaide -23.8 (-23.6)
17 (17) Gold Coast -47.4 (-49.3)