The optimal way of doing things would be to include both mean and variance (variance = standard deviation squared and is used by statisticians rather than the SD). High mean is good, low variance is good - so perhaps mean divided by variance would be a better guide to batting ability? I'll do some more research with my powerful stats software soon!Originally Posted by Pratyush
As for medians? Er... no! They're not as good as the mean as they have absolutely no indicator of how someone performs at either end of the range. Granted, most measures of central tendency hide as much as they show, but I'd have got that wrong and I'm a maths genius
I'll show this question to my statistics lecturer this morning, see what he thinks to it.
@ Jungle Jumbo - did you get my e-mail?