Quote:
Originally Posted by butterfly
If that was a good approach with respect to commenting on competitive balance, wouldn't you expect to see a notable increase in the σ of league winning percentage as you've observed in the σ of league goal differential?
|
Standard deviation is not the be-all and end-all. It doesn't work well when you get away from bell curves and into bimodal distributions, for which the central tendency actually goes negative. That's what we're seeing here.