Remove this Banner Ad

Analysis Champion Data articles

🥰 Love BigFooty? Join now for free.

SJ said:
I believe this is the paper which was the first work in the underlying basis to AFL Player Ratings: http://www.rankingsoftware.com/research/PossessionVersusPosition.pdf

From a set shot:

View attachment 133756

From open play but not under immediate pressure:

View attachment 133757

From open play and under immediate pressure:

View attachment 133758

These are significantly different charts that show how many points (on average) having the ball is at a certain position on the ground, in different circumstances. As you said, turning the ball over from a kickout leading into an intercept mark is one of the worst options as practically anywhere within 25 metres of the posts is practically worth a goal to the opposition. However turning it over practically anywhere inside your own forward half with the opposition under immediate pressure is either a neutral outcome or indeed worth more to your team than the opposition.
Click to expand...
If you turn that last map sideways, it looks like Homer Simpson trying to be scary :)
I would beg to differ on your interpretation of turnovers in our defensive 50- sadly, it appears that our opponents are more likely to create inside 50s and score goals from our turnovers there, because we leave ourselves so wide open. Perhaps over all the clubs the trend is different, as that heat map shows, but from what I can see, Geelong are always bucking the trends :(

That's really interesting though- it shows how much backline pressure influences your opponents' scoring shots and probably why we seem to be able to force our opponents to be inaccurate. Something I think we have been a lot better at this year, with the inclusion of Jed Bews and, now, Thurlow, though the old stalwarts have been- and continue to be- very good at it as well. Last year (looking at Footywire stats), our opponents scored a total of 284 goals and 285 behinds against us. That's about 50/50, not including missed opportunities. Geelong scored 315.303, also not including missed opportunities- slightly better than 50/50. Our accuracy was 15th overall, at 47.4%, well behind the Hawks who were 57.1% and West Coast (52.8%), according to AFL stats.

So far this year, our opponents have kicked 76.85 against us, while we've kicked 69.64. That's not enough scoring shots on our part but we've done OK with what we've had, with an improved accuracy of 49.3% (7th overall), still a long way behind the Eagles (57.9%) and Freo (56%), though. Interestingly, the Hawks have started the season poorly, with a goal kicking accuracy of 48.2%- ranked 11th. Hawks are ranked 1st for I50s, though, while we are ranked 14th- this suggests that we are making a better fist of it than they are, at the moment.
 
twarby said:
i am not sure what they publish in their book, I source a few team related stats from someone and scoring sources is one, although I am pretty sure this gets published in Mondays paper as well.
Click to expand...
Oh thanks- yes, I've seen the scoring from stoppages/turnovers, etc. stats at the bottom of the stats sheet in our Sunday paper. We never seem to have the whole story, though- i.e. we don't get those stats for all of the weekend's games- and I haven't collected all the stats in each paper to try and work out a story about it all. Hubby would go mad if I did - I already have a few copies of this year's stats squirrelled away, that he has helpfully offered to throw out. I keep telling him that I haven't read the papers, though, so he can't chuck them out yet. ;)

Thanks for sharing your info- you're more than welcome to post any other tidbits that you might see over the course of the year. :)
 
AM said:
It's not only the stats that one can avail themselves of it's the capacity to interpret and weight them that equally - some would say more importantly - counts. Which is where the clubs have it over we punters and the commentariat.
Click to expand...
Well, and they get access to vastly more data than we do.
 

Log in to remove this Banner Ad

AM said:
That was part and parcel of my point ,t.j. Clubs gather their own stats in addition to those they get from other sources as well.
Click to expand...
I meant that what they get from CD is vastly greater than what the public gets. Certainly, I agree that clubs are in a better position to use that data. But the same was true in the US, and yet amateurs ran rings around MLB, NBA and NHL clubs in their analysis (and ultimately changed all three sports). Be interesting to know just how much investment AFL clubs make in the use of what they've got.
 
thejester said:
I meant that what they get from CD is vastly greater than what the public gets. Certainly, I agree that clubs are in a better position to use that data. But the same was true in the US, and yet amateurs ran rings around MLB, NBA and NHL clubs in their analysis (and ultimately changed all three sports). Be interesting to know just how much investment AFL clubs make in the use of what they've got.
Click to expand...
How does that model work? Do the leagues themselves fund the collection and publication of the data for analysis?
 
It's impossible to make a better model/analysis when you don't have access to the data. The ones we do have access to aren't very useful and are often a combination of a lot of smaller categories.
 
SJ said:
How does that model work? Do the leagues themselves fund the collection and publication of the data for analysis?
Click to expand...
Yeah, but the data being used was incredibly basic - the innovation was in how it was used, not what it was. Moneyball covers it in detail but the early sabremetrics guys like Bill James just gleaned data from the basic box scores published in the paper. One of the key insights that drove the A's success (that on-base percentage was the most important measure of a batter's ability) didn't come from new data, but an aerospace engineer just thinking about the game critically.

I'm more familiar with hockey, but it's basically the same story. Maybe a decade ago smart guys who liked hockey started arguing on the internet about it. They came up with a variety of metrics, all of which are drawn from very basic stats, that greatly increased understanding of the game. Obviously in hockey measuring possession in the way that you can in footy or soccer is impossible, but some guys figured out that shot attempts were a good proxy for it; and that that proxy was a good predictor of a teams ability to win consistently. NHL.com publishes SA data as part of its play-by-play updates on Gamecentre; websites like war-on-ice automatically harvest it, punch it into an equation, and produce shot attempt metrics (known as Corsi and Fenwick, in the latter case after the blogger who invented it). Similar story with PDO (which measures shooting and save percentage), and quality of competition. The data is really basic, and was being published by the NHL for other reasons, but in the right hands could be really interesting.

Now, clearly CD are well ahead of that stage. But it'd be interesting to know how much sway their work has within clubs.
 

Remove this Banner Ad

SJ said:
How does that model work? Do the leagues themselves fund the collection and publication of the data for analysis?
Click to expand...

synergy sports handle the nba and it seems they are separate from the league itself (obviously funded by the teams paying for the info):
http://corp.synergysportstech.com/

also, there is a site called basketball reference where anybody can glean stats about players past and present (the info on it is expansive):

http://www.basketball-reference.com/players/g/griffbl01.html
 
thejester said:
Yeah, but the data being used was incredibly basic - the innovation was in how it was used, not what it was. Moneyball covers it in detail but the early sabremetrics guys like Bill James just gleaned data from the basic box scores published in the paper. One of the key insights that drove the A's success (that on-base percentage was the most important measure of a batter's ability) didn't come from new data, but an aerospace engineer just thinking about the game critically.

I'm more familiar with hockey, but it's basically the same story. Maybe a decade ago smart guys who liked hockey started arguing on the internet about it. They came up with a variety of metrics, all of which are drawn from very basic stats, that greatly increased understanding of the game. Obviously in hockey measuring possession in the way that you can in footy or soccer is impossible, but some guys figured out that shot attempts were a good proxy for it; and that that proxy was a good predictor of a teams ability to win consistently. NHL.com publishes SA data as part of its play-by-play updates on Gamecentre; websites like war-on-ice automatically harvest it, punch it into an equation, and produce shot attempt metrics (known as Corsi and Fenwick, in the latter case after the blogger who invented it). Similar story with PDO (which measures shooting and save percentage), and quality of competition. The data is really basic, and was being published by the NHL for other reasons, but in the right hands could be really interesting.

Now, clearly CD are well ahead of that stage. But it'd be interesting to know how much sway their work has within clubs.
Click to expand...


We'd need to by a fly on the wall during Match Committee heart-to-hearts to know the emphasis that's placed on stats and how each one is weighted.

There are clues to the value clubs place on team and player stats though. One is the number of people beavering away on keyboards behind coaches in the coaching box on match days. Recall counting 8 in a West Coast box last year.

I'm told there are also personnel within clubs whose role is to refine the data and produce it in usable chunks.

Then there's coaches post game pressers. It's rare for a coach not to refer to stats during that time, particularly losing coaches.

I see a lot of local footy and even at that level there are a number of people recording stats. The half a dozen key indicators appear on whiteboards at breaks and it's generally the first thing the senior coach - even the line coaches - refer to when addressing the players.

The stats are quite specific these days even at local level.

Extremely important tool is my take.
 
AM said:
We'd need to by a fly on the wall during Match Committee heart-to-hearts to know the emphasis that's placed on stats and how each one is weighted.

There are clues to the value clubs place on team and player stats though. One is the number of people beavering away on keyboards behind coaches in the coaching box on match days. Recall counting 8 in a West Coast box last year.

I'm told there are also personnel within clubs whose role is to refine the data and produce it in usable chunks.

Then there's coaches post game pressers. It's rare for a coach not to refer to stats during that time, particularly losing coaches.

I see a lot of local footy and even at that level there are a number of people recording stats. The half a dozen key indicators appear on whiteboards at breaks and it's generally the first thing the senior coach - even the line coaches - refer to when addressing the players.

The stats are quite specific these days even at local level.

Extremely important tool is my take.
Click to expand...
When my son was playing U17s, there was a young lad who was in the districts team (East Freo squad, whatever you call it) who wasn't getting games in the Colts side most weekends so he played U-17s most of the time. His Colts coach had asked his parents to collect his stats every week, so he could see how he was progressing.

I didn't see what stats they were, though. He was a ruckman so obviously hitouts and possessions but things like clearances and tackles, goals etc would've been written down also, I expect.

So that was happening at junior club level about 4 years ago- with 16-17 year old kids. I'm sure he wouldn't be the only player being monitored, either. The stats aren't the be-all and end-all of the game, obviously but they are an important tool, as you said, in understanding the influence and development of each player.
 
Champion Data is including a nice little set of herbs and spices in its game statistic tables in the Herald Sun this year.

One addition they've made is add in an 'expected score', alongside the actual score. This 'expected score' utilises accuracy over time factoring in difficulty (distance, angle, applied pressure) of shot and so is a better measure of comparing accuracy of the team on the day.
 
SJ said:
Champion Data is including a nice little set of herbs and spices in its game statistic tables in the Herald Sun this year.

One addition they've made is add in an 'expected score', alongside the actual score. This 'expected score' utilises accuracy over time factoring in difficulty (distance, angle, applied pressure) of shot and so is a better measure of comparing accuracy of the team on the day.
Click to expand...
Yeah, I'm loving the extra data; the tables also include intercepts and intercept marks, metres travelled, score involvements and sources of scores (kick in, turnover, etc.). It gives us number crunchers happy dreams. :D
 

🥰 Love BigFooty? Join now for free.

Rourke said:
There is an update to those pictures in this year's AFL Prospectus. The raw data has been smoothed in a sensible way too, so you can calculate the equity using a parametrised formula based on X-Y and pressure on the player
Click to expand...
images


Far too advanced for me.... :confused:
 
Here's a tasty little article on free kicks- specifically the WCE home ground advantage when it comes to free kicks.

WEST Coast is the most protected team in the AFL.

The Eagles have won a staggering 349 more free kicks than they have conceded in the past five years — a differential three times bigger than their nearest rival.

But it’s at Subiaco, at the aptly nicknamed “House of Pain”, where West Coast gets the biggest advantage from whistleblowers.

Since the start of 2011, the Eagles have lost the free kick count in only four of 46 matches on home soil against interstate sides — a success rate of 91 per cent.

http://mobile.news.com.au/sport/afl...ground-advantage/story-fndv8w9k-1227383490467
 
Not an article but had to laugh listening to Ken Hinkley on SEN on Friday night. Mooney asked him if he was still as much of a stats nerd as he had been at Geelong, and Hinkley replied that he wasn't because the game was changing so much week-to-week...and deployed as an example of that the fact Port were giving up huge numbers of I50s early in the season and that it was a problem that was going to explode in their face, which it duly did. Which to me seems to prove the exact opposite of what Hinkley was trying to say.
 

Remove this Banner Ad

Analysis Champion Data articles

Similar threads

cats_09
  • Sticky
Mod. Notice Please Read re Copyright: Posting Newspaper Articles
Replies
1
Views
2K
iameviljez
iameviljez

🥰 Love BigFooty? Join now for free.

Back
Top