I calculate a 29 point win next week.

11/22/2008 10:46:29 PM

So it is graphed, so it shall be








[Edited on November 22, 2008 at 10:48 PM. Reason : ]

11/22/2008 10:48:27 PM

haha nice

11/22/2008 10:54:14 PM

this is obviously assuming that our weekly margin is a 2nd order polynomial function of the calendar week of the game.
Which i see no reason shouldn't be true!!

11/22/2008 10:56:00 PM

interesting.

11/22/2008 10:56:45 PM

change SFU if you want me take your jew math seriously

11/22/2008 10:59:04 PM

seeing as the r² isn't very close to 0.99 on that, i would say the next datapoint won't land on the regression line

so ill say 26-pt win

11/22/2008 11:01:49 PM

This thread confused wut. He dropped out of engineering because graphs and regressions are difficult

11/22/2008 11:07:46 PM

lol wut?

hahahah

11/22/2008 11:08:46 PM

^^ haha

11/22/2008 11:18:51 PM

Quote :

"seeing as the r² isn't very close to 0.99 on that, i would say the next datapoint won't land on the regression line"

you're trying too hard.

[Edited on November 22, 2008 at 11:35 PM. Reason : and really, why not just say close to 1?]

11/22/2008 11:35:20 PM

the 0.99 is just stuck in my head from six sigma classes

11/22/2008 11:44:36 PM

six sigma. lol.

11/23/2008 12:16:54 AM

sample space is too small

11/23/2008 12:17:39 AM

11/23/2008 12:26:12 AM

what happens when you remove the bye weeks from your model
does the software you used to create that support a loess fit or a penalized b spline fit, that would be cool to see what they predict

11/23/2008 8:31:29 AM

What is the R-squared??

11/23/2008 1:33:28 PM

i bet that shit is like 0.8 LOL

11/23/2008 1:34:24 PM

lets get a graph when we have both irving and wilson in

11/23/2008 1:36:49 PM