Jump to content

Featured Replies

 

When I was a number loving nerd-child, it would have been my dream job to work for Champion Data and invent new statistics for the AFL.

As a number loving nerd-adult I find it just about impossible to gauge how meaningful these complex statistics are when there are so many components of it that are subjective. For example, in some particular situations, "corralling" is really the most effective method pressure available - if you're standing several meters from your opponent who beats you to the ball and you're a bit flat footed but your opponent is mobile, and you try to apply direct physical pressure, you risk over-committing and having your opponent leave you eating their dust. And oft times you see a player lay a tackle, which would amount to "physical" pressure, but the tackle is hopeless and the player gets the ball away easily.

And don't even get me started on the fact that DE% is an input in to the calc.

I mean I guess the weightings of each type should take these vagaries in to account, but I'd need to see this particular stat get used a lot more before I was comfortable with how well it is calibrated against how high a team's pressure "looks". If I were a coach I'd be pretty skeptical about using it for any sort of analysis until I'd seen it play out a bit.

The longer I go on, the more I'm convinced the only statistic that really matters is the W or L in the result column. The simple, counting stats can be used to debunk obviously wrong criticisms of players ("Joe Bloggs just kept getting caught holding the ball. Oh, only one free kick against for the whole game? Guess not.") but I think for the average fan, the complex ones are just gobbledygook that confuse things more than they make them clear.

2 minutes ago, Nasher said:

...

I think for the average fan, the complex ones are just gobbledygook that confuse things more than they make them clear.

I think Champion Data need to release this kind of stuff because they keep all their real info secret so they can sell it for big money.

In American sports where lots of statistics are released, 3rd party analysts can find real and meaningful correlations, and the public can come up with their own personal key stats.

Because the stats aren't available, they need to make up some kind of over complicated metrics to convince they are getting the info.

 
26 minutes ago, deanox said:

I think Champion Data need to release this kind of stuff because they keep all their real info secret so they can sell it for big money.

In American sports where lots of statistics are released, 3rd party analysts can find real and meaningful correlations, and the public can come up with their own personal key stats.

Because the stats aren't available, they need to make up some kind of over complicated metrics to convince they are getting the info.

The betting agencies dont want their most precious stats to get out unto the public domain. Those that indicate key stats that determine the outcomes of games.

I could almost guarantee they pay Champion Data to supress the most important parts of their data collection and key game indicators.


The only pressure stat that counts is "did the applied pressure adversely affect the player with the ball" so that he loses control of the ball, is penalised, miss-kicks, or misses a target that should otherwise have been readily achievable, or leads to an interception and so on.

So long as the player is prevented from an unfettered disposal then the pressure has been effective.

I will lead it to others to try and apply an objective test to the degree of pressure and try to apply some formula but for me all pressure actions are equal. The aim is the stop or prevent clean disposal.

 

 

 
Just now, Fanatique Demon said:

Um, I think Champion Data is a business.  Why would they give away the information they generate? 

The afl allow them to monopolise the market. It is outrageous 

Open up the market and allow competition 

For wins/losses the three most significant stats for the past 2 decades are contested possessions, inside 50 and marks inside 50. Not rocket science. Win the ball, kick it forward, mark it. 

On any one game, pressure ratings, free kicks, 1 percenters etc can have an influence but over the course of a season i believe those are the only three to be 'statistically' important across all teams' results over multiple years. 

 


Is that true? Or has Champion Data bought a licence from the AFL? Or is it just that the barriers to entry are too high for competition?

What is to stop me from counting kicks etc and publishing them? 

18 minutes ago, Win4theAges said:

The betting agencies dont want their most precious stats to get out unto the public domain. Those that indicate key stats that determine the outcomes of games.

I could almost guarantee they pay Champion Data to supress the most important parts of their data collection and key game indicators.

You're probably right.

10 minutes ago, Fanatique Demon said:

Um, I think Champion Data is a business.  Why would they give away the information they generate? 

The AFL own 49% of Champion Data. The whole thing is a racket.

These two bits of info make sense. Betting agencies won't be paying champion data to suppress the stats, it will be part of their sponsorship agreement.

12 hours ago, Fanatique Demon said:

Is that true? Or has Champion Data bought a licence from the AFL? Or is it just that the barriers to entry are too high for competition?

What is to stop me from counting kicks etc and publishing them? 

Nothing is stopping you from counting kicks at the footy.

Each club collects their own stats on specific indicators i think, albeit a a small number. They obviously get the full data set from Champion data and then i guess the skill is mining that data to customize it for their own KPIs etc.

And there are some examples of journos and others creating their own analytics based on publicly available data - for example the excellent articles by Cody Atkinson (and someone else?) for ABC news.  

But even if you collected your own stats I'm not sure you would be able to publish them as there are licensing issues no doubt.

However, even if you could publish them, there is no way you - or more pertinently a start up who set out to collect and publish (or sell) their own AFL data   - could get  comprehensive useful data in the first place. So therefore no way you could monetize it. 

I'm totally guessing here (someone will know specifics i'm sure), but i assume Champion data have access to the GPS numbers, are provided a box, or similar, with all the necessary infrastructure (power, internet access, monitors etc etc) and access to the grounds is facilitated (eg right passes to get into the grounds, parking spots, moving equipment about etc etc). 

Only Champion data gets all this from the AFL, effectively making it impossible for another organization to compete. 

There is another data tool which the AFL own and control - AFL Stats pro. I assume they use this to generate the data they provide on their website (and the club's websites too). It appears to be fully automated ie the data is pulled form the vision of the game directly. No doubt Champions data have access to this too. 

Edited by binman

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Unfortunately, your content contains terms that we do not allow. Please edit your content to remove the highlighted words below.
Reply to this topic...

Featured Content

  • PREGAME: Richmond

    The Dees are back at the MCG on Thursday for the annual blockbuster ANZAC Eve game against the Tigers. Can the Demons win back to back games for the first time since Rounds 17 & 18 last season? Who comes in and who goes out?

      • Like
    • 61 replies
    Demonland
  • PODCAST: Fremantle

    The Demonland Podcast will air LIVE on TUESDAY, 22nd April @ 8:00pm. Join Binman, George & I as we analyse the Demons first win for the year against the Dockers. Your questions and comments are a huge part of our podcast so please post anything you want to ask or say below and we'll give you a shout out on the show. If you would like to leave us a voicemail please call 03 9016 3666 and don't worry no body answers so you don't have to talk to a human.

      • Thanks
    • 14 replies
    Demonland
  • POSTGAME: Fremantle

    A undermanned Dees showed some heart and desperation to put the Fremantle Dockers to the sword as they claimed their first victory for the season winning by 10 points at the MCG.

      • Haha
      • Love
      • Like
    • 293 replies
    Demonland
  • VOTES: Fremantle

    Max Gawn is leading the Demonland Player of the Year award from Christian Petracca followed by Ed Langdon, Jake Bowey & Clayton Oliver. Your votes for our first victory for the season. 6, 5, 4, 3, 2 & 1.

      • Like
    • 42 replies
    Demonland
  • GAMEDAY: Fremantle

    It’s Game Day, and the Demons return to the MCG wounded, undermanned and desperate. Still searching for their first win of the season, Melbourne faces a daunting task against the Fremantle Dockers. With key pillars missing at both ends of the ground, the Dees must find a way to rise above the adversity and ignite their season before it slips way beyond reach. Will today be the spark that turns it all around, or are we staring down the barrel of a 0–6 start?

      • Haha
    • 634 replies
    Demonland
  • PREVIEW: Fremantle

    A month is a long time in AFL football. The proof of this is in the current state of the two teams contesting against each other early this Saturday afternoon at the MCG. It’s hard to fathom that when Melbourne and Fremantle kicked off the 2025 season, the former looked like being a major player in this year’s competition after it came close to beating one of the favourites in the GWS Giants while the latter was smashed by Geelong to the tune of 78 points and looked like rubbish. Fast forward to today and the Demons are low on confidence and appear panic stricken as their winless streak heads towards an even half dozen and pressure mounts on the coach and team leadership.  Meanwhile, the Dockers have recovered their composure and now sit in the top eight. They are definitely on the up and up and look most likely winners this weekend against a team which they have recently dominated and which struggles to find enough passages to the goals to trouble the scorers. And with that, Fremantle will head to the MCG, feeling very good about itself after demolishing Richmond in the Barossa Valley with Josh Treacy coming off a six goal haul and facing up to a Melbourne defence already without Jake Lever and a shaky Steven May needing to pass a fitness test just to make it onto the field of play. 

    • 0 replies
    Demonland