[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: Celtics just lucky?



At 09:40 AM 3/26/2003 -0500, Eggcentric@AOL.com wrote:
< Put me in the camp that thinks the article stupid as well.
This is exactly why I hate statistics and don't even bother to read
them most of the time. > - Josh

Well Josh and Shizz and dajubo and dliles -  this column really
raised your hackles. Sure it is not about to be nominated for a Pulitzer,
but equally as sure is that it incorporates all those distasteful stats
that no Celt true-blue wishes to own up to.

< It would be more rewarding to think of a few statistics
that actually tell you something than ones that don't. > -Josh

These stats, like your prejudices, are but one man's slant. But who is
to say that these stats are actually misleading or stupid?
Of course the stats aren't, they never are. It's all in the presentation and selection of which ones you do/don't use. And those are pretty definitely handled here with an intent to mislead to suit an agenda and show how 'brilliant' this guy is in being able to quantify things.

First, this didn't raise my hackles over showing our warts, because I'm good with stats and data analysis and can come up with a bunch of worse ones if I wanted. I can make those suckers dance and sing and play the ukelele. But c'mon Egg --- his article about his own 'brilliant' deduction that the Cs are just lucky derives from one of those superficially enticing assumptions for idiots misleading points - if you win you score more points, therefore average points scored should be better than average points given up or the winner is just lucky. That's THE standard basis for misleading stats -misusing/misapplying average values. As Mark pointed out days ago, that totally ignores the skewing effect of even a few aberrations (and this season has been full of them) by giving more weight to one historically bad loss than dozens of the more usual tight score wins in evaluation a team. Bad analysis.

Kim