+- Carlton Supporters Club (http://new.carltonsc.com)
+-- Forum: Princes Park (http://new.carltonsc.com/forum-4.html)
+--- Forum: Robert Heatley Stand (http://new.carltonsc.com/forum-14.html)
+--- Thread: Stick it on the lockers!! (/thread-2536.html)
Re: Stick it on the lockers!! - LP - 02-02-2016
(02-02-2016, 05:59 AM)DJC link Wrote:If Yarran's above average ranking is legitimate, most of our list would have to be average or above average.
In that statement lies the truth about the folly of a purely statistical analysis.
A player can be ranked highly but have no "value" at all!
Re: Stick it on the lockers!! - jeza - 02-02-2016
(02-02-2016, 08:56 PM)LP link Wrote:In that statement lies the truth about the folly of a purely statistical analysis.
A player can be ranked highly but have no "value" at all!
Must have taken his northern blues form into account.
Re: Stick it on the lockers!! - Thryleon - 02-02-2016
(02-02-2016, 05:59 AM)DJC link Wrote:Jon Ralph's article relies on Champion Data's statistical analysis which concludes that we have "no elite players and only Kade Simpson and Marc Murphy as above average" and Patrick Cripps is average.
However, in today's Age, Jon Pierik reviews Richmond through the use of Champion Data's statistical analysis and that places Chris Yarran in the above average category. I understand that Champion Data uses statistics gathered over two seasons and Yarran had a good year in 2014 and an above ranking for that season is fair enough. His form last season stank and must have dragged his ranking down to below average.
If Yarran's above average ranking is legitimate, most of our list would have to be average or above average.
Statistical bias must be looked at here.
We have data for 2014, where Yarran would have rated either above average or higher.
In 2015, he actually didnt play many games which might further skew the actual rating.
Statistical analysis, always need to be measured relative to something else for them to actually have any relevance which this champion data rating seems to ignore.