Champion Data's Analysis of Collingwood so far in 2025

[Finnishpie]

He does - but he's a salesman selling their data models. Some of the commentators though... Bucks is the worst. He talks as though he doesn't get the basic thing that the the play produces the stats. He talks as though the stats produce the play.

The thing that Hoyne does that drives me nuts is say things like: When X plays in defence, team Y win 58% of the time compared to 45%.

But football doesn't work like that -- maybe that stat, for example, is because another player is playing.. Or because he plays defence on wet days rather than dry ones, or because of all sorts of things. Football is a complex and holistic game.

 

[sideswipe]

The thing that Hoyne does that drives me nuts is say things like: When X plays in defence, team Y win 58% of the time compared to 45%.

But football doesn't work like that -- maybe that stat, for example, is because another player is playing.. Or because he plays defence on wet days rather than dry ones, or because of all sorts of things. Football is a complex and holistic game.

Have you ever looked at golf's Strokes Gained system, devised by a guy named Mark Broadie? It's a pretty nifty system which identifies how golfers score against their competition. Now used across the professional golf landscape.

 

[doppleganger]

The thing that Hoyne does that drives me nuts is say things like: When X plays in defence, team Y win 58% of the time compared to 45%.

But football doesn't work like that -- maybe that stat, for example, is because another player is playing.. Or because he plays defence on wet days rather than dry ones, or because of all sorts of things. Football is a complex and holistic game.

The stat is just that, a stat.

They don't tell you why...the are good to back up what you see with the eye.

All sports are trying to develop more and more complex models to explain the game.

And the really hard thing is trying to attribute individual player impact in the team game...by there nature they award the same ratings for all acts on the ground in all games and teams, despite teams not all playing with the same game styles or strategies.

 

[Finnishpie]

The stat is just that, a stat.

They don't tell you why...the are good to back up what you see with the eye.

All sports are trying to develop more and more complex models to explain the game.

And the really hard thing is trying to attribute individual player impact in the team game...by there nature they award the same ratings for all acts on the ground in all games and teams, despite teams not all playing with the same game styles or strategies.

My point is this: it assumes that the stat is an important factor in our win/loss ratio - but it might not be at all. Eg: We lose 100% of games when Josh Daicos plays in the centre square. That’s true and also the implication is nonsense. NONSENSE.

 

[doppleganger]

My point is this: it assumes that the stat is an important factor in our win/loss ratio - but it might not be at all. Eg: We lose 100% of games when Josh Daicos plays in the centre square. That’s true and also the implication is nonsense. NONSENSE.

But that is the point with all stats, majority are not important and most are not even understood properly by the people using them.

Plenty of fans and commentators don't even appear to understand that a possession and disposal are two different things!!

It is good that people are delving more into the analysis side of things, but have a long way to go.

 

[manicpie]

But that is the point with all stats, majority are not important and most are not even understood properly by the people using them.

Plenty of fans and commentators don't even appear to understand that a possession and disposal are two different things!!

It is good that people are delving more into the analysis side of things, but have a long way to go.

I've been trying to work out how you get a possession without a disposal and I can't work it out. The only thing I can think of is demon possession

 

[doppleganger]

I've been trying to work out how you get a possession without a disposal and I can't work it out. The only thing I can think of is demon possession

D.Cameron wins a HO to Naicos, who gathers the ball but is immediately tackled for a second ball-up.

Naicos is accredited with a contested possession, but no disposal.

 

[manicpie]

D.Cameron wins a HO to Naicos, who gathers the ball but is immediately tackled for a second ball-up.

Naicos is accredited with a contested possession, but no disposal.

I did not know that. Thanks for clearing up something that I haven't been able to. Possessions should be regarded as inferior to disposals then. Interesting

 

[gouki88]

I did not know that. Thanks for clearing up something that I haven't been able to. Possessions should be regarded as inferior to disposals then. Interesting

I think it's mostly useful for distinguishing how one first gets possession (contested vs uncontested), but the conflation is annoying

 

[doppleganger]

I think it's mostly useful for distinguishing how one first gets possession (contested vs uncontested), but the conflation is annoying

Even that doesn't tell the full story.

An intercept possession is one of the most important parts of the game in modern footy, but they are included as uncontested possessions.

And free kicks for are included as a contested possession.

IMHO the Pies current game would place most emphasis on intercept possessions (uncontested).

 

[gouki88]

Even that doesn't tell the full story.

An intercept possession is one of the most important parts of the game in modern footy, but they are included as uncontested possessions.

And free kicks for are included as a contested possession.

IMHO the Pies current game would place most emphasis on intercept possessions (uncontested).

Yeah, there's some fun depth you can dive into with possessions. It's why I like wheelo so much - sorting by several of the different possession stats easily, and with little explanations down the bottom. So much easier to use than the clunky AFL site

 

[PonsfordMagpie]

Yeah, there's some fun depth you can dive into with possessions. It's why I like wheelo so much - sorting by several of the different possession stats easily, and with little explanations down the bottom. So much easier to use than the clunky AFL site

Agree. Wheelo is an excellent resource.

 

[doppleganger]

I did not know that. Thanks for clearing up something that I haven't been able to. Possessions should be regarded as inferior to disposals then. Interesting

They just need to be understood and used correctly, as the overwhelming majority of instances a player both possesses the ball and disposes people use the terms interchangeably, but they are different stats

IMHO yes, how you use the ball (disposals) is the most important thing.

But a player taking the kick-in from a behind gets credited with a disposal these days...some players can launch scores from kick-ins, but majority are just defensive chips to the pocket.

Don't think many would value that higher than Sidey running onto a loose ball and winning possession and then getting a hand ball off that sets up a score or Dan McStay taking a big contested pack mark (a contested possession) down the line from a kick-in.

And our game relies on plenty of uncontested possession chains where we just run and link up the ground and eventually hit a forward on a lead for an uncontested mark.

 

[No SPIN]

Agree. Wheelo is an excellent resource.

do they have stats broken down by QTR?

 

[Finnishpie]

But that is the point with all stats, majority are not important and most are not even understood properly by the people using them.

Plenty of fans and commentators don't even appear to understand that a possession and disposal are two different things!!

It is good that people are delving more into the analysis side of things, but have a long way to go.

Well, it's not exactly the point with all stats. For example, to use the Josh Daicos example again: if we said he averages 25 possession a week in the centre and 15 on the half-back line that tells us something real, because it's a statistic about only him. It's not a statistic about him that we are pretending is about the entire team -- i.e. when he plays half-back we win 58% of the time.

 

[doppleganger]

Well, it's not exactly the point with all stats. For example, to use the Josh Daicos example again: if we said he averages 25 possession a week in the centre and 15 on the half-back line that tells us something real, because it's a statistic about only him. It's not a statistic about him that we are pretending is about the entire team -- i.e. when he plays half-back we win 58% of the time.

Still not really sure what your point is?

All the stat you are quoting is giving is a win ratio. It appears that you are inferring more from the stat about Josh's positioning than what it is saying.

 

[Gammalite44]

Against Carlton we lost Clearances by 9, but were +27 on post clearance possession ( post game on FOX).
That's a great stat but have not seen it on the usual stats sites.

Daniel Hoyne is a gun with the analytic interpretation and footy knowledge
Just remember from the 900 minutes of footy played each week he is responsible for directing the narrative for so much of media. So take all of those ex players / coaches now on the gravy train of a plethora of football shows during the week and remember not all of them are contestants of Mastermind or hard working. Its no small job covering 18 clubs and 800 odd players and trying to put in into media friendly format.

Also those media outlets pay CD big $$$ to provide the footy starved masses (i.e like those on these forums!!) with the data stuff. Your right so much is just selective analytical nonsense. But without it the footy shows would have no content, Ex players and coaches would be queuing at Centrelink and the dull masses would be feeling neglected. Hoyney and CD are doing exactly what the customer pays them to provide!

Point of the story - don't take any of the footy stats too seriously or get hung up why it doesn't feel correct. My guess only at club land where their analysts work on the totality of the data where the true value of footy analytics works.

 

[sr36]

And the really hard thing is trying to attribute individual player impact in the team game...by there nature they award the same ratings for all acts on the ground in all games and teams, despite teams not all playing with the same game styles or strategies.

Yep. Dogs players do really well in Champion Data stats because as a team the Dogs run forward so hard, so they score really well for their involvements in goals, but they don't and can't measure the impact of those keeping their defensive shape really well that helps a team stop the other team from scoring.

 

[Harry O]

Hoyne must hate Fly - his coaching system cannot be understood by Champion and their ratings

 

[adamov13]

Does anyone have an idea what Josh Daicos

7

Josh Daicos

Age

27

Ht

178cm

Wt

77kg

Pos.

D/M

Career

Season

Last 5

D

15.6

4star

K

8.3

3star

HB

7.3

5star

M

3.3

3star

T

3.6

5star

MG

230.5

4star

D

16.2

4star

K

9.2

4star

HB

7.0

4star

M

3.3

3star

T

3.5

4star

MG

230.5

4star

D

13.8

4star

K

7.6

3star

HB

6.2

4star

M

3.2

4star

T

4.2

5star

's champion data rating was for the final quarter on Friday night? I read somewhere that his rating for the match was 25 odd but just curious to know what his last quarter was.

 

[Finnishpie]

Still not really sure what your point is?

All the stat you are quoting is giving is a win ratio. It appears that you are inferring more from the stat about Josh's positioning than what it is saying.

My point is that when Hoyne talks about those sorts of stats, he is the one making the implication, and Healy etc assume it to be an argument for the implication -- i.e., if Josh Daicos in the centre corresponds to a 100% loss rate, then never have him in there.

 

[PatrickPiesFan]

Does anyone have an idea what Josh Daicos's champion data rating was for the final quarter on Friday night? I read somewhere that his rating for the match was 25 odd but just curious to know what his last quarter was.

I don't know the exact rating, but Tom Morris referenced on Monday that it was the best single quarter anyone has played this year in terms of rating.

 

[HeathySure]

MY theory is that when we move the ball its so clean we don't have to get contested possessions. Most of the possessions are uncontested and that's why we are better than the other sides.

Its just a theory from my eye test. Could be wrong







mo

 

[Gone Critical]

Daniel Hoyne is a gun with the analytic interpretation and footy knowledge
Just remember from the 900 minutes of footy played each week he is responsible for directing the narrative for so much of media. So take all of those ex players / coaches now on the gravy train of a plethora of football shows during the week and remember not all of them are contestants of Mastermind or hard working. Its no small job covering 18 clubs and 800 odd players and trying to put in into media friendly format.

Also those media outlets pay CD big $$$ to provide the footy starved masses (i.e like those on these forums!!) with the data stuff. Your right so much is just selective analytical nonsense. But without it the footy shows would have no content, Ex players and coaches would be queuing at Centrelink and the dull masses would be feeling neglected. Hoyney and CD are doing exactly what the customer pays them to provide!

Point of the story - don't take any of the footy stats too seriously or get hung up why it doesn't feel correct. My guess only at club land where their analysts work on the totality of the data where the true value of footy analytics works.

Rather than give Hoyne too much credit I think he is better explained by the saying
"In the land of the blind the one eyed man is king"

He is hailed as an expert in an industry that essentially has no knowledge of what he is selling. AFL does stats, or as it really should be called data, very badly. They are always open to self proclaimed gurus. Hoyne reads far.too much meaning into the data he collects. He will be serendipitously correct some of the time but he won't know which times those are. If a real statistician looked at his data and analysis they would tell.him its interesting but essentially meaningless from a statistical point of view.

That doesn't mean it isn't useful but it should be analysed from a view to help confirm benefits of performance, game plans, tactics etc with the background thought that the conclusions you take from it are potentially very fallible.

I don't know what Hoynes background is but I doubt its in stats. Like Dank he is a man with not expert knowledge who sells himself to an industry who doesn't understand the difference between an expert view and someone who knows a bit more about the topic than they do.

 

[HoraceStrasbork]

Mine said I was elite at eating, drinking and making a

I've been trying to work out how you get a possession without a disposal and I can't work it out. The only thing I can think of is demon possession