Review What's wrong with Champion's Data?

[Mr Reliable]

It's not news that we won by 53 points after leading for most of the night. We had 54 inside 50s to 48 and we kicked 12 -1 to 3 - 5 from stoppages. A good number of our players must have been doing OK.

Despite this, the team totals for Champion Data were 1664 Carlton to 1634 us. WTF? I like to look at the stats but something is seriously wrong with them.

By way of example, LT scored 78 for 7 disposals, 6 contested, 100% efficiency, 5 tackles, 3 goals, 4 score involvements and 17 pressure acts.

Eddie Betts scored 82 for 11 disposals, 7 contested, 64% efficiency, 3 tackles, 2 goals, 7 score involvements and 18 pressure acts (and disappearing for most of the game).

Garlett and Yarran both got more Champion Data points than LT (and Anthony, Narni, Wright, Adams, Harvey, Harper, Atley and Scotty Thompson). WTF again?

You'll all be pleased to note that Narni was a one point better defender than Bryce Gibbs. WTF again?

It seems to me that there is something missing from the stats. I suspect it has something to do with the QUALITY of the play. What sort of organisation operates without proper quality control in the 21st century? C'mon Champion - lift your game.

 

[Thewlis Dish]

Kruezer somehow got 30 hit outs. I'm not sure where.

 

[Mr Reliable]

Kruezer somehow got 30 hit outs. I'm not sure where.

And Robinson got 2. But the one in the last qtr where Wellsy (third man up) taps to Campbell who handballs to Harper for a goal is not counted. Maybe they gave that one to Kreuzer.

 

[franjon]

It's not news that we won by 53 points after leading for most of the night. We had 54 inside 50s to 48 and we kicked 12 -1 to 3 - 5 from stoppages. A good number of our players must have been doing OK.

Despite this, the team totals for Champion Data were 1664 Carlton to 1634 us. WTF? I like to look at the stats but something is seriously wrong with them.

By way of example, LT scored 78 for 7 disposals, 6 contested, 100% efficiency, 5 tackles, 3 goals, 4 score involvements and 17 pressure acts.

Eddie Betts scored 82 for 11 disposals, 7 contested, 64% efficiency, 3 tackles, 2 goals, 7 score involvements and 18 pressure acts (and disappearing for most of the game).

Garlett and Yarran both got more Champion Data points than LT (and Anthony, Narni, Wright, Adams, Harvey, Harper, Atley and Scotty Thompson). WTF again?

You'll all be pleased to note that Narni was a one point better defender than Bryce Gibbs. WTF again?

It seems to me that there is something missing from the stats. I suspect it has something to do with the QUALITY of the play. What sort of organisation operates without proper quality control in the 21st century? C'mon Champion - lift your game.

Champion data ,Dream Team, Supercoach , all a fair amount of bullshit when it comes to North Melbourne,
Champion Data. NMFC 1634 to Carl 1664
Dream Team. NMFC 1432 to Carl 1637.
Personally I think all this crap is popularity driven and as such is media driven.
To all the Carlton people and all the media scum- Check the scoreboard and suffer!

 

[Benno_900]

I'm convinced Champion Data points is just subjective bullshit.

 

[Husk]

They out possessed us with meaningless uncontested ball chipping it around the wings etc.

Doing this not only gives them extra points, but also inflates their DE% which also increases their overall score.

Super Coach is generally the most accurate of the two IMO, but its still misleading, especially when considering DE%.

 

[Artichoke]

I'm convinced Champion Data points is just subjective bullshit.

Yes sir. I used to be a big wrap for their stats but this year, I'm convinced some of the stat guys are drinking their own tap water. Gary Ablett farts and gets a point for that. LT would have have to kick 10 goals to crack 100 points!!!! I don't get it. I'm glad those suckers aren't in control of my email inbox - I get 1000 emails/day so you can imagine how many emails I get per year. Sometimes my inbox gets 'clogged up' and I can't help but think Champion Data play a role in this. Put it this way, if Champion Data were in charge of time travel, I wouldn't be celebrating 20 years of it in Perth mind you!

 

[Ghostwriter]

We kicked a goal every second time we went inside 50, according to the radio today, maybe we were more efficient with ball use, just sayin'.

 

[GreatBradScott]

We were that good going forward, everyone is complaining. Love this shit.

 

[Hank93]

We were that good going forward, everyone is complaining. Love this shit.

You would think the "most comprehensive ranking system in Aussie Rules Football" would award more points to a team that won by 54.

Carlton's cheap stats should mean nothing to their supercoach totals. I'm still miffed as to why AA defenders Glass, Rutten, Fletcher and Scarlett have never been good supercoach players.

 

[Mr Reliable]

You would think the "most comprehensive ranking system in Aussie Rules Football" would award more points to a team that won by 54.

Carlton's cheap stats should mean nothing to their supercoach totals. I'm still miffed as to why AA defenders Glass, Rutten, Fletcher and Scarlett have never been good supercoach players.

Absolutely correct. Cheap marks, kicks and handballs should not count for as much as quality possessions and defenders should be rewarded for good spoils.

In the second qtr Drew Petrie spoiled the ball to LT in the goalsquare for a goal. That probably didn't get counted either.

 

[Event Horizon]

During the 1990s, when we were seriously dominant, we'd still score very poorly if these stats existed back then. Our game style was all about efficiency and playing to our strengths.

Scoreboard is the only true indicator.

 

[TennisPlayerAndy]

Scoreboard is the only true indicator.

The ONLY thing that matters.

And boy, we are peppering it big time.

 

[Jacob Wilson]

It's not news that we won by 53 points after leading for most of the night. We had 54 inside 50s to 48 and we kicked 12 -1 to 3 - 5 from stoppages. A good number of our players must have been doing OK.

Despite this, the team totals for Champion Data were 1664 Carlton to 1634 us. WTF? I like to look at the stats but something is seriously wrong with them.

By way of example, LT scored 78 for 7 disposals, 6 contested, 100% efficiency, 5 tackles, 3 goals, 4 score involvements and 17 pressure acts.

Eddie Betts scored 82 for 11 disposals, 7 contested, 64% efficiency, 3 tackles, 2 goals, 7 score involvements and 18 pressure acts (and disappearing for most of the game).

Garlett and Yarran both got more Champion Data points than LT (and Anthony, Narni, Wright, Adams, Harvey, Harper, Atley and Scotty Thompson). WTF again?

You'll all be pleased to note that Narni was a one point better defender than Bryce Gibbs. WTF again?

It seems to me that there is something missing from the stats. I suspect it has something to do with the QUALITY of the play. What sort of organisation operates without proper quality control in the 21st century? C'mon Champion - lift your game.

Hello Mr Reliable,

I thought I would answer a few of your questions:

The reasons why Betts out scored Thomas are:

Betts had 2 goal assists, 1 smother, 3 hardball gets, 1 clearance
Thomas had 0 goal assists, 0 smothers, 1 hardball get, 0 clearances

Its also worth noting that Betts had 7 effective disposals, 4 ineffective disposals and 0 disposal clangers. Thomas had 7 effective disposals, 0 ineffective disposals, 0 disposal clangers.

The reason SC scores can be seem unfair at times is that they include so many stats like smothers, spoils, hitouts to advantage, possession efficiency, scoring assists and many others, so it is incredibly hard to estimate a players score.

Garlett's SC was high because he had a phenomenal 6 score assists (four more than anyone else on the field). Yarran's score was high because he had 18 touches (15 effective), a goal, a score assist, 5 running bounces (not sure if these are counted) and a clearance.

The team scores were close because all the key indicators of team performance were close and Carlton actually won a few.
For example:

Contested possessions were Kangaroos 135 - Carlton 132
Tackles - Kangaroos 46 - Carlton 61
Disposal Count - Kangaroos 323 - Carlton 378
Disposal Efficiency - Kangaroos 77% - Carlton 74%.
Scoring shots - Kangaroos 29 - Carlton 26.

So despite the final scoreline is deceving the match was actually not that one sided. Im not saying North didn't deserve to win because they did and they were clearly the better side. However, a margin of 1-4 goals would not have been unfair given how the match was played.

Another reason why the scores look a little off is that stats won after the match was over (in the 4th quarter) are discounted as they didnt really impact the match as much as the people who played well when the match was in the balance in the first 2 and a half quarters. And people who played well early's points are inflated. Thats why Petrie's score of 165 was so dam high. To Further explain this point Karmichael Hunt's goal was worth roughly 35 SC points where as Hansen's goal at the end of the match on friday night was worth about 10 points. I think we can all agree thats fairer and rewards key plays and prevents people getting cheap touches and high scores.

Wells was awarded 2 hiouts to advantage on Friday night so I think that hitout would have been given to him.

Kruezer did play the entire match in the ruck and attended roughly 100 ruck contests so his 30 hitouts aren't that surprising.

When Petrie spoiled the ball to the Thomas goal he probably would have been awarded an effective spoil and a goal assist.

I hope this helps clear up your misunderstanding.

 

[Roobound]

During the 1990s, when we were seriously dominant, we'd still score very poorly if these stats existed back then. Our game style was all about efficiency and playing to our strengths.

Scoreboard is the only true indicator.
North Melbourne Football Club est 1869
Event Horizon, Sunday at 11:05 PMReport


EH that shits me to tears nowadays when people compare players we had 25% less possessions a game and won by 60 points but to compare players its done on possessions and we would lose this everytime due to the game plan.

On the topic from the OP they would have more points because they had 50odd more possessions you dont really get points for not having the ball. There non marking forward line meant they were always trying to manipulate a way in with lots of short passes but no real system when going forward.

 

[Spitta09]

Scotland, Carazzo and co play chip to chip on the wing of course it is going to elevate stats and points.

 

[Hunters Host #3]

Thanks for clearing that up Jacob.

I know of someone who works at Champion and they would reiterate everything you've said.

BTW, I do believe however, that Wells should be awarded a 2x multiplier for being 'Silky Sexy' on the field and LT should have been awarded an extra 10 points for his celebratory ball dodge on the sidelines.

 

[Mr Reliable]

Hello Mr Reliable,

I thought I would answer a few of your questions:

The reasons why Betts out scored Thomas are:

Betts had 2 goal assists, 1 smother, 3 hardball gets, 1 clearance
Thomas had 0 goal assists, 0 smothers, 1 hardball get, 0 clearances

Its also worth noting that Betts had 7 effective disposals, 4 ineffective disposals and 0 disposal clangers. Thomas had 7 effective disposals, 0 ineffective disposals, 0 disposal clangers.

The reason SC scores can be seem unfair at times is that they include so many stats like smothers, spoils, hitouts to advantage, possession efficiency, scoring assists and many others, so it is incredibly hard to estimate a players score.

Garlett's SC was high because he had a phenomenal 6 score assists (four more than anyone else on the field). Yarran's score was high because he had 18 touches (15 effective), a goal, a score assist, 5 running bounces (not sure if these are counted) and a clearance.

The team scores were close because all the key indicators of team performance were close and Carlton actually won a few.
For example:

Contested possessions were Kangaroos 135 - Carlton 132
Tackles - Kangaroos 46 - Carlton 61
Disposal Count - Kangaroos 323 - Carlton 378
Disposal Efficiency - Kangaroos 77% - Carlton 74%.
Scoring shots - Kangaroos 29 - Carlton 26.

So despite the final scoreline is deceving the match was actually not that one sided. Im not saying North didn't deserve to win because they did and they were clearly the better side. However, a margin of 1-4 goals would not have been unfair given how the match was played.

Another reason why the scores look a little off is that stats won after the match was over (in the 4th quarter) are discounted as they didnt really impact the match as much as the people who played well when the match was in the balance in the first 2 and a half quarters. And people who played well early's points are inflated. Thats why Petrie's score of 165 was so dam high. To Further explain this point Karmichael Hunt's goal was worth roughly 35 SC points where as Hansen's goal at the end of the match on friday night was worth about 10 points. I think we can all agree thats fairer and rewards key plays and prevents people getting cheap touches and high scores.

Wells was awarded 2 hiouts to advantage on Friday night so I think that hitout would have been given to him.

Kruezer did play the entire match in the ruck and attended roughly 100 ruck contests so his 30 hitouts aren't that surprising.

When Petrie spoiled the ball to the Thomas goal he probably would have been awarded an effective spoil and a goal assist.

I hope this helps clear up your misunderstanding.

Thanks, Jacob. I appreciate you taking the time and I do understand so much better now.

 

[liam_13]

The team scores were close because all the key indicators of team performance were close and Carlton actually won a few.
For example:

Contested possessions were Kangaroos 135 - Carlton 132
Tackles - Kangaroos 46 - Carlton 61
Disposal Count - Kangaroos 323 - Carlton 378
Disposal Efficiency - Kangaroos 77% - Carlton 74%.
Scoring shots - Kangaroos 29 - Carlton 26.

So despite the final scoreline is deceving the match was actually not that one sided. Im not saying North didn't deserve to win because they did and they were clearly the better side. However, a margin of 1-4 goals would not have been unfair given how the match was played.

Good post and thanks for clearing some things up.

BUT I still think the stats are worthless, clearly key indicators of team performance are not actually key. Geelong are a great example of why this wank over contested possessions is merely a wank, it means nothing unless you win it by probably 40 or more I'm guessing, and even then it doesn't decide a game.

Does CD take into account game styles etc when weighting things like contested ball?

 

[ThePezDispenser]

A clear example of where stats lie. At the endof the day the work done by Carlton players didn't create as many goals and they should be scored to reflect this.

 

[Jacob Wilson]

Good post and thanks for clearing some things up.

BUT I still think the stats are worthless, clearly key indicators of team performance are not actually key. Geelong are a great example of why this wank over contested possessions is merely a wank, it means nothing unless you win it by probably 40 or more I'm guessing, and even then it doesn't decide a game.

Does CD take into account game styles etc when weighting things like contested ball?

Hello Liam,

I think you are misinformed about Geelong - if anything they are a great example of how important contested possessions are. In there previous five seasons (their dynasty) leading up to 2012 they ranked in the top three teams for contested possession differential.

2007 - 2nd
2008 - 1st
2009 - 1st
2010 - 2nd
2011 - 3rd
2012 - 13th

Geelong have been an amazing contested ball team and finished in the top three teams in all five of these seasons.

In 2012 they are currently sitting seventh on the ladder (a four spot drop from previous seasons) and their contested possession differential is ranked 13th.

Geelong are also currently the best team at overcoming a poor contested possession differential.

The team who wins the contested possession count wins about 80% of matches.

I beleive on average the winning team wins the contested possession count by about 10. However, a team like Adelaide typically wins the contested possession stat by 20 in its wins.

There is a good article in the 2011 AFL prospectus which you should read. It details other key statistical indicators of success like defensive pressure and kick rating and how they convert into team success.

I hope this helps.

 

[Husk]

We lost the CP count by 24 on the weekend. But dominated the midfield at various times. Champion data is good, but its nowhere near 100% accurate.

 

[Mr Reliable]

Hello Liam,

I think you are misinformed about Geelong - if anything they are a great example of how important contested possessions are. In there previous five seasons (their dynasty) leading up to 2012 they ranked in the top three teams for contested possession differential.

2007 - 2nd
2008 - 1st
2009 - 1st
2010 - 2nd
2011 - 3rd
2012 - 13th

Geelong have been an amazing contested ball team and finished in the top three teams in all five of these seasons.

In 2012 they are currently sitting seventh on the ladder (a four spot drop from previous seasons) and their contested possession differential is ranked 13th.

Geelong are also currently the best team at overcoming a poor contested possession differential.

The team who wins the contested possession count wins about 80% of matches.

I beleive on average the winning team wins the contested possession count by about 10. However, a team like Adelaide typically wins the contested possession stat by 20 in its wins.

There is a good article in the 2011 AFL prospectus which you should read. It details other key statistical indicators of success like defensive pressure and kick rating and how they convert into team success.

I hope this helps.

Jacob, in your opinion what are 4 or 5 of the most important KSIS and where do we rank on each of them?

 

[Deep We11s 1]

Jacob, in your opinion what are 4 or 5 of the most important KSIS and where do we rank on each of them?

While we're waiting for Jacob I'll have a stab. In order and ranking.
1. Clearances (6)
2. Contested possessions (13)
3. Inside 50s (2)
4. Disposal efficiency (12)
5. Goal / Inside 50 % (3)
6. One %ers & Pressure acts (13)
7. Clangers (10)
8. Uncontested possessions (5)
9. Goal accuracy (10)
10. Rebound 50 % (8)

Tells the story ... win more contested ball, improve disposal efficiency when we've got it and apply more pressure when we don't have it. Everything else is trending in the right direction. Be interesting to compare rankings & differentials over the last 6 weeks to the first part of the year.

 

[Western Taipan]

At the end of the Day, if our score Goals +Behinds = Total is Higher than the opposition GOALS+ BEHINDS= Total. .
WE WIN.
Simple really.