Champion Data

[Brian Wilson 331]

I think I could pick a pretty decent first 22 without looking at the stats as any football observer with their salt would. Having said that of course coach's look at the stats but stats alone don't really tell the story and along with a little faulty analysis give some of the anomalous results that posters are complaining about.

I agree. The criticism should be about the analysis of the statistics, not the statistics themselves.

The 'shelter shed' test is a little misleading but I'll take up the argument. I think that the first dozen picks or so would be relatively easy but that the remaining picks would become increasingly difficult and the further it went, the more statistical analysis would become relevant. One point in CD's analysis is that Sydney's bottom group is one of the most impressive. Depth is important and those final slots in the team can be the difference.

In any case, all captains would pick an excellent 22, but I doubt that any two captains would pick the same one.

[Chook 14,988]

I nearly choked on my bar b qued alpaca . . .

? ? ?

[Bitter but optimistic 22,057]

? ? ?

Why the ?????. If you prefer roasted or sauteed just say so!

[rjay 24,577]

I agree. The criticism should be about the analysis of the statistics, not the statistics themselves.

The 'shelter shed' test is a little misleading but I'll take up the argument. I think that the first dozen picks or so would be relatively easy but that the remaining picks would become increasingly difficult and the further it went, the more statistical analysis would become relevant. One point in CD's analysis is that Sydney's bottom group is one of the most impressive. Depth is important and those final slots in the team can be the difference.

In any case, all captains would pick an excellent 22, but I doubt that any two captains would pick the same one.

...and that's the beauty of it.

[Chook 14,988]

Why the ?????. If you prefer roasted or sauteed just say so!

No, I'm just freaked out that you needed to use three words to spell "barbecue".

Edited March 8, 2013 by Chook

[Bitter but optimistic 22,057]

No, I'm just freaked out that you needed to use three words to spell "barbecue".

Well I suppose it is controversial but Macquarie allows both barbeque and bar-b-q. Stephen Murray - Smith however insists on your spelling.

[Moonshadow 17,471]

Yeah, the Jones omission shows how bad the ranking system is.

I'll have mine roasted, thanks BBO

[Winners at last 5,150]

I think I've said something like this before but if you picked teams using the old school method of lining everyone up against the shelter shed and the captains selecting their teams one at a time does anyone think they would care too much about what champion data has to say?

I was always one of the last picked. Thanks for reminding me.

[Demon Dynasty 17,091]

Poor Shannon. Melbourne

ELITE

Jack Grimes defender

BELOW AVERAGE

Rohan Bail midfielder

Jake Spencer ruck

James Sellar defender

Tom Gillies defender

Sam Blease forward

Neville Jetta midfielder

Luke Tapscott defender

James Strauss defender

POOR

Shannon Byrnes forward

http://www.theage.com.au/afl/afl-news/tiger-sixpack-rates-a-mention-20130306-2flk2.html#ixzz2MnOiGnKw

Not worth the $$ required to collect the data.

The rep from Champion stated on SEN this week that to qualify for the rankings, a player had to have played only 5 matches over a period of 2 seasons of football.

That's approximately only 10% game time over 2 seasons for some players' (data), included as part of the overall sample date.

As Sheldon from BBT would be happy to espouse, using such a small amount of data on any one player (or more) within the sample (or in this case sample data ie., as a percentage of the overall possible sample) is a very risky strategy IMO. In general, the larger the sample size N (of the included player's data, in this case the number of games of data used for the statistical analyes/rankings), the smaller the sampling error should (in theory) be.

In short, poor form from Champion data and any outcomes/rankings derived from said data should be taken with a severe amount of skepticism

At best, this is something you might wish to wrap your fish & chips in and quickly move on

[Barney Rubble 1,576]

Also there are 2 other categories that "Caros whipping boy"

forgot to put in his very selective article.

"ABOVE AVERAGE & "AVERAGE"

[ThreeOneSix 954]

Pearce Hanley is considered Elite for Brisbane - and he is an out and out spud.

Yet Nathan Jones doesn't get a mention for us!

Either way Richmond won't be troubling the top sides in a hurry, it's not in their make up.

[dpositive 1,649]

Do you think Neeld and Craig ignore statistical analysis? Does Longmire? What about Clarkson?

I think you will find that every coach is interested in statistical analysis. Of course, some put more weight in it than others, and different coaches are interested in different measures. But the point is that every club uses statistical measures as a means of helping analysis. And yes, Champion Data is relevant here.

I think we need to separate the data from the analysis. The data itself is very useful. In fact, it's a necessity in modern footy. However, the analysis is what is really important.

Assessing the players without statistical analysis would be like David Misson assessing their running fitness without a stopwatch. He would still have a pretty good idea but I'm sure he'd like the insight that can be gained from the data.

The old "Data Continuum"

Data is not Information

Information is not Knowledge

Knowledge is not Wisdom"

[rjay 24,577]

Pearce Hanley is considered Elite for Brisbane - and he is an out and out spud.

Yet Nathan Jones doesn't get a mention for us!

Either way Richmond won't be troubling the top sides in a hurry, it's not in their make up.

Pearce Hanley is a very good footballer, would love to have him at Melbourne.