Illinois unranked in final AP Poll, #25 in final Coaches Poll

Status
Not open for further replies.
#26      
How does Sagarin work again? I'm not going to get into a long debate over it, but it's odd to me that Illinois is #27 and Tennessee is #23 when according to these numbers, Tennessee had a weaker schedule than Illinois AND did worse in said schedule. Not to mention Illinois beat Tennessee. I just don't understand the algorithm.

Sagarin's ratings are weighted toward strength of schedule and margin of victory. I infer that the latter was a smidge stronger for Tennessee even though its SOS was lower than ours.

My point, however, is that you can barely shim a cigarette paper between the two teams (80.32 v. 79.28) just as there's no great discernible difference between #21 Michigan and #30 Virginia. Big difference between # 1 and #10. Smaller difference between #11 and #20. Little difference between #21 and #30.

The same delta exists between #1 Indiana and #10 Penn State as does between #21 Michigan and #79 UCLA.

Something something Pareto distribution.
 
#27      
How does Sagarin work again? I'm not going to get into a long debate over it, but it's odd to me that Illinois is #27 and Tennessee is #23 when according to these numbers, Tennessee had a weaker schedule than Illinois AND did worse in said schedule. Not to mention Illinois beat Tennessee. I just don't understand the algorithm.

I believe (someone can correct me if they know more) he uses standard statistical measures (e.g. ordinary least squares method for score differential), rather than the highly advanced metrics a lot of sites use now. You can play a weaker slate of teams and beat a couple of them by larger margins and come out ahead in his rankings. Or have the a similar average but have a great game against a stronger opponent, which moves your overall metric more. Sagarin was big when it came out, but I think his method is long out of date. It's still pretty good directionally.

Btw, Indiana 5-0 vs top 10 is crazy good.
 
#31      
I believe (someone can correct me if they know more) he uses standard statistical measures (e.g. ordinary least squares method for score differential), rather than the highly advanced metrics a lot of sites use now. You can play a weaker slate of teams and beat a couple of them by larger margins and come out ahead in his rankings. Or have the a similar average but have a great game against a stronger opponent, which moves your overall metric more. Sagarin was big when it came out, but I think his method is long out of date. It's still pretty good directionally.

Btw, Indiana 5-0 vs top 10 is crazy good.
Sincere question: how can a poll compiled using sports "journalist" picks or the tallies of coaches who can't possibly do their jobs competently and watch college football games in-season possibly be a better predictor of team strength?

I don't pay much attention to what's out there. What method has more descriptive power than Sagarin? Next season I'd like to use whatever is better.
 
#32      
So many bowl results went against us. It felt like every single bowl was a competitor for a slot in the 20's winning, usually beating a team ranked too high to fall below us. I thought SMU and Washington would snake us too, I'm actually surprised we got so close.

Houston-LSU, Iowa-Vandy, North Texas-SDSU, Virginia-Mizzou, also all really close games that sneak us in at 25 if they tilted the other way.
Yet we are ahead of UW who beat us.... go figure...
 
#33      

I agree a lot with his ranking. He has us 16 shows a lot of respect
 
#34      

I agree a lot with his ranking. He has us 16 shows a lot of respect
The tears of the SEC fans upset the didn't include/didn't rank them highly enough are delicious.
 
Status
Not open for further replies.
Back