Carlton Supporters Club

It has to be about more than access to statistics. If our coaching staff have the same footy knowledge and footy IQ as a few big mouths on social media, then we are seriously wasting our time and money, and the club should pack it in and go and play in some dust bowl league out in the boonies, where the local waste collection guy is the coach, physio, high performance manager, list boss………

Yep
Of course it's more than stats....even advanced ones.
Naturally they play a part in some decisions and the coaches have access to better ones than we have but...

Our head coach has 289 playing games.
He has around 15 years experience as a senior or assistant coach.
Add the experience of our other assistants.

Then add an in depth knowledge of the playing group.

-their strengths and weaknesses
-their injury and form situation at any one time
-priorities they may be working on at any given time.
-what their training loads look like
- development plans.
-how players fit in the team structure and roles they play
-team balance considerations
- opposition strengths and weaknesses in up coming games
-proposed match-ups
-personal issues players may be dealing with (relationships and family illness)
etc  etc.

We think we might know.
We actually have little idea

Yes. I agree. Put any of us in the coaches’ box on game day and we would be found out in the first five minutes.

Few collingwood supporters on this forum it appears, everything is black and white.

Nobody is saying statistics is the be all and end all. If it was, we'd have computers as head coaches.

However, people saying statistics don't see this, or can't see this and are of no use are equally wrong.

The truth is in the middle.

When Trigg came to the club he lost his $h!t because we didn't have any kind of analytics in place with our drafting and recruiting. That is, 100% trusting peoples 'eye' and not using statistical evidence to back it up and compare.
Clearly that was wrong.....and we have our poor recruiting/development as proof.

Since then we've come on board with the modern game and have an abundance of statistical information to back our coaches up.
We;ve had a few very good 'moneyball' recruits off the back of it - Think Elijah Hollands as an example.
Alternatively, when you look at what the old fashioned 'eye' test tells us, which got Vossy undone in Brisbane, it also hasn't worked here with Fantasia......but the personal relationships got that done more than statistical.

Coaches have favourites, they are only human.
Fans too.

Players play on emotion....coaches and MC can't afford too.

As i said earlier, ignore statistics at your own peril......we have 20+ years of dark ages as evidence.

I dont think anyone said ignore the stats, but most have said beware of putting too much stock into them.

I dont think anyone said ignore the stats, but most have said beware of putting too much stock into them.

MBB posted an interesting "statistical analysis" of Tom De Koning's game against Essendon that ventured into the one or two rucks territory.

Champion Data notes De Koning has averaged 21.1 AFL Player Ratings points, 22.3 disposals, 14.7 contested possessions, 9.7 clearances, 4.7 centre-bounce clearances and 6.7 score involvements in that stretch. It ranks him No.1 among the season averages of all rucks throughout that run.

Why then does Tom not appear in the AFL's mid year All-Australian team?  I'm pretty sure that his name wasn't mentioned during the  televised All-Australian selectors' meeting.  Perhaps Stats Insider's AFL Player Ratings can shed some light.  They use a combination of votes, performance assessments and statistical data to rank players.  Tom is currently the ninth ranked ruckman in their ratings (Goldstein is 16th, one behind Pittonet).  I think that by considering factors other than statistics, Stats Insider's ratings better reflect players' performances and influence.  It could be argued that Stats Insider's ratings are less objective than Champion Data's numbers-based system but Champion Data's statistical categories, definitions, weightings, etc are subjective.

Then there's;

Carlton is a much better turnover team when De Koning is playing as a sole ruck. According to Champion Data, the team's scores from turnover differential is -5.3 when Pittonet is paired with De Koning, but +25.4 when De Koning plays alone.


If that's what the statistics tell us, then it must be true.  BUT, is there a causal relationship between scores from turnover and whether  Tom plays a lone hand or with Pitto?  Possibly, but there are other factors like the opposition, the form of other players, umpiring and our other scoring processes.  If we are dominating the clearances and scoring heavily from stoppages, there will limited opportunity to score from turnovers.  And that's where relying of football statistics can lead you astray.

As flawed as the statistical categories and their definitions may be, footy stats may help to identify trends, what's working and what's not, and where game plans or structures might need a tweak, but only if the limitations of the data are understood.  Furthermore, those statistics should not be viewed in isolation but in the context of the full range of factors that influence footballers' performances and the results of games.

Adam Saad has played eight games this season for seven wins.  He was subbed off in the game we lost.  Statistically, if Adam plays the whole game, we will win.  Let's see how that holds up 

Why then does Tom not appear in the AFL's mid year All-Australian team?  I'm pretty sure that his name wasn't mentioned during the  televised All-Australian selectors' meeting.  Perhaps Stats Insider's AFL Player Ratings can shed some light.  They use a combination of votes, performance assessments and statistical data to rank players.  Tom is currently the ninth ranked ruckman in their ratings (Goldstein is 16th, one behind Pittonet).  I think that by considering factors other than statistics, Stats Insider's ratings better reflect players' performances and influence.  It could be argued that Stats Insider's ratings are less objective than Champion Data's numbers-based system but Champion Data's statistical categories, definitions, weightings, etc are subjective.

the vibe i got from watching that AA TV thing, was that there wasnt a heap of stats/science put into it (maybe aside from Buckley) - there was a lot of "jeez, he has been pretty good the last couple of weeks" as the basis for putting a name up.   I am sure i heard someone say of Mac Andrew "did you see his first quarter on Curnow, last week" - if that is the basis for making the squad, then it obvious why they have so many misses