Plural
Registered User
- Mar 10, 2011
- 33,722
- 4,879
Personally I don't really have any objections when someone has a drastically different POV and/or weighs player contributions outside of the average line. It's fun to see people taking different approaches and finding new (or in this case, possibly old ) angles to tackle these lists.
The problem I have is that very simplistic statistical model was offered as an explanation for some of the rankings when it's clear that the statistical model is not giving the desired results most of the time. I get that context and other views were considered and the statistical model was probably just a crude example on the listing. But the reality is that the model does not work even with modest consistency, therefore rendering itself useless to this exercise. It's not that anyones opinion is wrong, it's that using a statistical tool that wouldn't pass the smell test within thousand yards from a class room is dubious at best. If you apply one model for several cases to draw conclusions you need to have the exact same set of rules for all cases. Otherwise you're losing integrity of the comparison and any conclusions drawn are automatically invalid.
The problem I have is that very simplistic statistical model was offered as an explanation for some of the rankings when it's clear that the statistical model is not giving the desired results most of the time. I get that context and other views were considered and the statistical model was probably just a crude example on the listing. But the reality is that the model does not work even with modest consistency, therefore rendering itself useless to this exercise. It's not that anyones opinion is wrong, it's that using a statistical tool that wouldn't pass the smell test within thousand yards from a class room is dubious at best. If you apply one model for several cases to draw conclusions you need to have the exact same set of rules for all cases. Otherwise you're losing integrity of the comparison and any conclusions drawn are automatically invalid.