Champion Data’s List & Player Ratings - An Accurate Tool or Total Fabrication ??

hk89

Norm Smith Medallist
Jul 12, 2009
8,503
10,459
Melbourne
AFL Club
Hawthorn
More weirdness...


Yes, shows why teams are more than just a sum of their parts. I imagine champion data is summing the player ratings of the players in each unit and then ranking based on that. However cohesiveness of the unit and game plan are huge factors this type of analysis seems to ignore, so you get results like Geelong and Hawthorn being 1st and 3rd for least points conceded, but 16th and 10th ranked for defence. While bulldogs that only had 5 teams concede more points than them ranked 2nd for defence.
 
Apr 29, 2012
24,113
34,585
Perth
AFL Club
Carlton
Gets e even better when you see the players CD rate "Elite"
ELOUUjrU4AAdjsr

ELOUUjuUcAAIqO7

ELOUUjuUYAA0GHn


Sorry Hawk fans Tex Walker, Connor McKenna, Huge Greenfield and Majak Daw are Elite talent and Tom Mitchell isn't
Based off of that there are 7 "elite" ruckmen in the game...

How the * does thatclassification work?

That would be like saying there a 50 "elite" midfielders. It's categorically a useless metric.
 
Giants midfield 12th, really? North midfield 2nd, really? Other than the ever delusional North fans, who doesn't think those rankings couldn't be easily swapped?

IMO, Giants midfield is one of the best, Cogs, Kelly, Ward, Taranto, Hopper, plus Whitfield, Greene, Williams who have swung through there and high draftees that can't break in! Plus DeBoer who does a highly valuable job every single week. This is a top flight mid field.

Says all you need to know about Champion data, great at measuring stuff but produce a lot of rubbish for click bait purposes.
 

Virgin Dog

Cancelled
Oct 29, 2017
9,734
20,645
AFL Club
Western Bulldogs
Giants midfield 12th, really? North midfield 2nd, really? Other than the ever delusional North fans, who doesn't think those rankings couldn't be easily swapped?

IMO, Giants midfield is one of the best, Cogs, Kelly, Ward, Taranto, Hopper, plus Whitfield, Greene, Williams who have swung through there and high draftees that can't break in! Plus DeBoer who does a highly valuable job every single week. This is a top flight mid field.

Says all you need to know about Champion data, great at measuring stuff but produce a lot of rubbish for click bait purposes.
North at 2nd is obviously way too high, but 12th would be shockingly low for North if you swapped theirs and GWS' rankings. Goldstein is still a top ruck when fit, while Cunnington + Higgins both had outstanding seasons in 2019. Their midfield was great in terms of contested footy, while Polec is a decent enough winger. I'd consider them top 8 for midfield at the very least.

GWS is definitely let down by their ruck division. The 12th place surprises me, but it's surely to be blamed on the complete lack of competent rucks. Would've thought the mids would bring it up to within the top 8, but it's possible that injury affects rankings (so Ward, Kelly not considered maybe?)

Of course, CD rankings really should be taken with a grain of salt. Trying to take a statistical approach to it is always going to lead to odd results. Eagles were shithouse for contested ball in 2019 (hence their lower ranking), but their game plan revolves around uncontested footy (i.e. kick and mark style of play). There's really not much way to account for this when CD so heavily favours contested football for player rankings
 
Jul 26, 2007
31,920
33,102
Darwin
AFL Club
West Coast

I was looking through these 'rankings' and pondered whether they are an accurate reflection of each club's list or just a heap of meaningless stats cobbled together to come up with a skewed set of conclusions ??

I'm not against these sorts of things but I do question the accuracy and how they arrived at their findings.

Also who is Champion Data and what else do they do to give them some credibility in this area ??

By the way, I'm not going to nit-pick which club has what on their list, it's all rather subjective.

Last year they ranked Melbourne's list as number 1.

Twelve months later pretty much the same list is ranked 13th.

That basically sums up the accuracy of CD rankings IMO.
 
Jun 27, 2010
1,779
4,183
AFL Club
Sydney
Other Teams
Sturt
As people have pointed out, CD are backward-looking. Plus they don't account for the nature of footy as a team game -' the champion team will beat a team of champions.'

I did a quick analysis on CD performance at end of last year (though must have deleted it). From memory they pretty much nailed it for seven or eight teams - either exactly predicted their ladder position or were just one position out. For about five clubs, eg, Melbourne, they were at least half a dozen places wrong. So not terrible, but not much better than asking a random supporter.

At the end of the year I might compare CD's ratings with half a dozen random BF year-end predictions.
 
North second best midfield in the comp, and Brisbane best... spare me hahaha. CD are useless.
A Top 4 ruckman
Statistically elite wingmen in Dumont and Polec help bump it
Good inside core of Higgins, Ziebell, Cunners, Simpkin

shouldn’t be top 2 but I get the high statistical ranking
 
Aug 4, 2003
22,984
23,443
WA
AFL Club
West Coast
A Top 4 ruckman
Statistically elite wingmen in Dumont and Polec help bump it
Good inside core of Higgins, Ziebell, Cunners, Simpkin

shouldn’t be top 2 but I get the high statistical ranking

CD have come out and confirmed North’s ranking was purely based on midfielders they have failed to lure to the club within the last 5 years.
 

Ambrosia

Brownlow Medallist
Mar 11, 2010
15,596
10,422
AFL Club
Essendon
The Champion Data raw stats are excellent but their statistical model for rating individual players doesn’t seem to work particularly well.

I’m fairly sure these list ratings are just for clicks anyway as rating each player individually based on a formula then then aggregating their scores to determine who has the best list seems backwards when you could just use team based stats.
 

Ambrosia

Brownlow Medallist
Mar 11, 2010
15,596
10,422
AFL Club
Essendon
A Top 4 ruckman
Statistically elite wingmen in Dumont and Polec help bump it
Good inside core of Higgins, Ziebell, Cunners, Simpkin

shouldn’t be top 2 but I get the high statistical ranking

The splitting of wing from midfield is such a weird thing to do especially given plenty of sides don’t play full time wingman.
 

Final Siren

Mr Squiggle
Aug 18, 2009
4,229
17,495
AFL Club
Richmond
As people have pointed out, CD are backward-looking. Plus they don't account for the nature of footy as a team game -' the champion team will beat a team of champions.'

I did a quick analysis on CD performance at end of last year (though must have deleted it). From memory they pretty much nailed it for seven or eight teams - either exactly predicted their ladder position or were just one position out. For about five clubs, eg, Melbourne, they were at least half a dozen places wrong. So not terrible, but not much better than asking a random supporter.

At the end of the year I might compare CD's ratings with half a dozen random BF year-end predictions.
It's literally labelled "2020 List Ratings," so it can't be too backward-looking. Obviously it's based on past results, but it's a rating for right now, not a retrospective look at 2019.

I recorded and compared Champion Data's 2019 List Rankings vs other people's predictions, and it was pretty poor - ranked 38th out of 45. Although, to be fair, a List Ranking is not quite the same thing as a ladder prediction. I did get to see their actual ladder prediction, which was similar but a bit better, but they don't publish this publicly.
 
Last edited:

Bald

Premium Gold
Apr 5, 2016
7,255
16,322
AFL Club
Western Bulldogs
Other Teams
Boston Celtics
Dogs list being second is very flattering. The fact that we achieved such a number despite CD considering our midfield our weakest area is even more flattering.

Grain of salt.
 

Yojimbo

Cancelled
10k Posts
Nov 14, 2012
10,914
9,834
The "Elephant" in the room.
AFL Club
Western Bulldogs
Dogs list being second is very flattering. The fact that we achieved such a number despite CD considering our midfield our weakest area is even more flattering.

Grain of salt.
They "Champion Data" obviously have not read my thread "The Bubble Of Excellence" which is a glorious
tribute to the manifest greatness of Macrae, Bontempelli and Dunkley, but maybe they have factored in
the not quite there yet impact of Tim English in the ruck and that effected our overall ranking.

CONTEXT is the key word when it comes to statistical manipulation, show me some CONTEXT.
 

Foucault1989

Norm Smith Medallist
Oct 16, 2016
8,470
16,750
AFL Club
Carlton
A Top 4 ruckman
Statistically elite wingmen in Dumont and Polec help bump it
Good inside core of Higgins, Ziebell, Cunners, Simpkin

shouldn’t be top 2 but I get the high statistical ranking
Haha Polec and Dumont "elite". Ziebull is a plodder hanging on by a thread. Simpkin? Seriously he is a bit away from even being considered a good player. Higgins and Cunington are very good footballers but would be on the fringes of the top 15 mids in the competition
 
Back