In partnership with CBSSports.com
Online Now 1598
Online now 1079 Record: 10351 (3/11/2012)
The largest and most active MSU Spartans board on the web
The place to ask questions to SpartanTailgate's recruiting experts
"The Duff" is dedicated to Michigan State football recruiting discussion
"The Bres" is dedicated to Michigan State basketball recruiting discussion
This is your pulpit to preach to the masses about everything from politics to religion
The place to buy, trade or sell Michigan State tickets
For fantasy football and other fantasy sports discussion
You have no favorite boards.
The most viewed topics.
The most replied to topics.
The most up-voted topics.
The most down-voted topics.
The most up-voted posters.
The most down-voted posters.
The most followed posters.
As we say in the profession, your model is severly underspecified and the relationship you observed between remaining schedule and final conference standing is highly spurious. To think the expected probabilities are invariant among certain classes of opponents is ridiculous. Good effort, but as a journal reviewer, I'd reject the manuscript.
This post has been edited 2 times, most recently by Radio Raheem 14 months ago
Make your own damn model then.
You guys are just like the people I used to.work with. I would make 50 pages of materials and they wouldnt do crap except to show up to the meeting and try to poke holes in my work. "No I actually intended to oversimplify that assumption because I knew your lazy ass wouldnt be able to understand it"
Just bustin your chops, bro. Oversimplifying is dangerous because it often leads to the wrong conclusions, but you often have to dumb things down to get people to understand it. Damned if you do, damned if you don't. Just remember: it's always easier to be a critic.
Much more meaningful analysis at The Only Colors using prediction models:
Either this website doesn't exist or is not currently available.
A clear favorite has emerged, one of the former favorites for the top three is no longer a favorite to earn a bye, Illinois appears to be pulling a repeat of last years collapse ... and poor, poor Penn State.
This post has been edited 2 times, most recently by TheDudeAbides 14 months ago
I liked reading what you put together.
If that guy was able to obtain point spreads for all these games and represent them with probabilities then that would be the way to do it.
But as I said, the biggest leap in my model might be putting MSU as equals to IU and UM. My model was initally constructed to re-evaluate the race and see what the strentgh of the positions really are in light of the scheduling and big games coming up. In that regard I think I was successful.
Obivously, IU and UM are #1 and #3. I'm guessing the spread for IU-UM is lower than IU-MSU.
But I will challenge his model head to head. If I win I am starting my own blog.
This post was edited by Big Green Stick 14 months ago
But it would be stronger if you used a model to obtain predicted probabilities rather than fixing your betas at pre-specified values.
I could use your help on my new blog.
I already have a couple of other feature stories in mind.
You couldn't afford my consulting rate.
I've often found that historically calculated betas, while institution specific, lead to higher deltas than industry standard data.
So you thought arbitrarily chosen parameters would be more defensible?
Well, we shall see. Score one for the BGS model. The Ony Colors predicts 15-16 wins for UM. I aint about to crown them yet.
Well, it looks like my model tore "The Only Colors To Shreds" which predicted 15 wins for UM and 12 for MSU. I think the problem with those kinds of models is that they are too backward looking. UM was #1 in the country and those models become a self-fulfilling prophecy of putting such a team as too heavy a favorite, rather than looking at the reality of what is going to come.
BGS model forecast 9.25-2.75
Revised forecast 14.50-3.50
BGS model forecast 9-3
Revised forecast 13-5
BGS model forecast 8.75-2.25
Revised forecast 14-4
Note that I see no need to revise my model as the variances are within acceptable ranges. Anyone could be correct by updating their models until the end. But I thought it might help to provide an update to again point out where the standings are in light of the remaining schedules. MSU holds a 1/2 game edge over IU and 1.5 game over UM. But of course that could change quickly with an unexpected loss. All teams within one game of their original 14-4 forecast.
Great job, BGS!!
You seriously should change your thread title to start with "Warning Science"...because this is legit!
In reality some luck or even dumb luck was involved. But I think we know how hard it is to win at places like Illinois, UW and Purdue. Although Purdue has tanked. But the Illini and UW have come back strong.
I think it's funny how some people really enjoy adding and subtracting and dividing and multiplying numbers.
I hate numbers. I like dunks and steals and wins and boobies.. I'm just saying
There is romance in all of us
Still trying to push your model, huh?
I like numbers, but I like dunks, steals and boobies as well.
They are not mutually exclusive. In fact, these days it seems like there is now momentum for the math nerds getting the chicks.
Just wait until I come out with my recruiting rankings analysis and I am bound to score.
247Sports In partnership with CBS Sports