GET YOUR GEEK ON

May I direct your attention to this http://www.cnn.com/2009/LIVING/06/04/real.life.superheroes/index.html?iref=mpstoryview#cnnSTCPhoto

An SEC coaches’ justice league seems too good to pass up

My stats training is cursory and aged, but I’d say that – even without the added visual benefit of a “fit line” – that we have no relationship here.

Not that I’m suggesting this… ok, maybe I am.

Shouldn’t this look at previous season comparisons to examine the delta between starts lost and change in production between seasons? So if there are N offensive starts lost, then a team experienced a change of (+/-)K points avg per season. A plot like that would so the overall effect of the starts lost

Naturally stats could be determined any way you want, but the correlation between starts lost and change in pts avg would be more indicative of a change.

Of course, then the “interpretation” of stats comes into play, and the punditry of “but Arkansas lost RUN-DMC and changed coaches” comes into play. But that’s the whole point of college football (VIVA LA BCS REVOLUTION!)

Forget about the stats, check out the superheroes at the Crimson Commodore’s link. I think the superpower of the woman from Ocala, FL would be stubbing out cigarettes on the bad guys.

My guess is that starts lost isn’t a good indication of experience lost. you could lose one guy with 4 years worth of starts or a bunch of guys with a handful of starts in their four years and either case would put you near the top of the list. I would think that stars lost / starters lost would be more indicative, with a sweet spot for damage done somewhere in the middle.

1

Yes, the possibilities are endless.

GIGGETY MAN!!! Able to rescue female sportscasters in a single bound.

Westsiiiiiiiiiidddddddddeeeeeeeeee!!!!!!!!

Ah, yes, the mean streets of Ocala, Florida.

Where you can get mugged and beaten up with regularity? Seriously?

Borrrring….. Wake me up when we see some SAS code and some regression analysis with a double blind test with a high beta. Well fuck me, I guess I did learn something in college!

There seems to be an ever so slight correlation, if you actually put a fit line in there. Not as strong as many would expect, though.