"Champion Data compares players in the same position and age categories over a number of years to give each a relative rating, which then determines the strength of a club's list."
That's how they do it.
They're comparing players with similar ages and positions against each other using their huge number of metrics, most of which aren't available to the general public... we only get to see the basic stuff outside of when the commentators drop obscure stats mid-game or when Kingy get's in the lab...
Remember not so long ago when CD announced that Collingwood had the best midfield in the comp? It's a similar thing... they're not looking at the mids in the best 22, they're looking at the sum of all the midfielders on the list.
When you look at our list demographic it's not surprising that we land down the bottom. We've got 8 players yet to debut, they effectively amount to nothing as there is no data on them. There's another 11 with less than a full season under their belt who are still finding their feet. 27 players have yet to notch 50 games. That's part of the reason why they're calling us last on overall list ranking but somewhat higher for 22 (as Freo is last). Then again, we all know what most media/pundits best 22s are like so who knows who they've decided makes the cut for that. Or perhaps it's simply the 22 statistically highest rated players on the list who make the cut.
I don't really have an issue with the premise of the article per say, or the way the stats are used in the study. My big issue is the way afl.com.au are taking those results and implying that they indicate talent, which isn't what those stats explicitly show - it's a combination of talent/ability and experience across a list.