Sagarin says

[Hammersmith]

Crap, I thought I actually understood this and now I see that it changed. How the heck did I miss all this last year?

#oldguyprobs

If I remember right, he changed things several weeks into last season.

Since the regular rating is now an average of two score-based systems, I'm just going to go with that(like we agreed toward the end of last season). It should cause a lot less confusion.

[thebootfitter]

If I remember right, he changed things several weeks into last season.

Since the regular rating is now an average of two score-based systems, I'm just going to go with that(like we agreed toward the end of last season). It should cause a lot less confusion.

Not sure if this will sway your judgment, but per the definitions that I copied from Sagarin, the Rating metric is actually a weighted average of all three metrics. More weight is placed on the two models that are better predictors.

I still think for simplicity, the Rating is probably best to use, unless we wanted to come up with our own metric that uses an average of just the two measures deemed to be better predictors. That just seems like a lot of unnecessary complication though.

[bisonaudit]

I'm confused. Look at Oregon. Their RATING is higher than any of the other three methods? How does that work?

[thebootfitter]

I'm confused. Look at Oregon. Their RATING is higher than any of the other three methods? How does that work?

Could be a mistake. Remember when he published some erroneous figures last year and then republished with corrected figures?

I didn't look closely enough to catch that, but I'll take a look too.

[Hammersmith]

And here we go. Four touchdown favorite going into Weber State. Closest game prediction is UNI by a TD. MVFC top three, four of top six, and six of top twelve. MVFC also a TD favorite over the next best conference.


Sagarin's Week 2 Predictions

			       Sagarin	  Predict    Result	 Diff  
 40	North Dakota State	76.45			
 70	Iowa State		68.39	   +4.90     +20.00	+15.10
197	Weber State		44.78	  +28.51		
218	Incarnate Word		39.27	  +40.34		
103	Montana			62.13	  +17.48		
177	Western Illinois	49.09	  +24.20		
 90	Southern Illinois	64.87	  +14.74		
188	Indiana State		46.54	  +33.07		
156	South Dakota		53.27	  +20.02		
 79	South Dakota State	67.08	  +12.53		
 82	Northern Iowa		66.57	   +6.72		
133	Missouri State		57.30	  +15.99		
107	Youngstown State	62.25	  +17.36		
					
108	Illinois State		62.23			
						Total Diff:	+15.10
	Home Field		 3.16

Sagarin FCS Predictor Top-25

 1   40  North Dakota State
 2   79  South Dakota State
 3   82  Northern Iowa
 4   84  Eastern Washington
 5   89  New Hampshire
 6   90  Southern Illinois
 7   94  Villanova
 8   99  Jacksonville State
 9  103  SE Louisiana
10  106  McNeese State
11  107  Youngstown State
12  108  Illinois State
13  109  Chattanooga
14  110  Montana
15  116  Tennessee State
16  119  Princeton
17  121  Richmond
18  123  Eastern Illinois
19  127  William & Mary
20  128  Harvard
21  130  Montana State
22  132  Coastal Carolina
23  133  Missouri State
24  134  Sam Houston State
25  136  Central Arkansas

Sagarin Predictor Top-5 FCS Conferences

1.  MVFC	60.86
2.  Colonial	53.62
3.  Ohio Valley 53.22
4.  Big Sky	52.26
5.  Southland	50.29

[bisonaudit]

And here we go. Four touchdown favorite going into Weber State. Closest game prediction is UNI by a TD. MVFC top three, four of top six, and six of top twelve. MVFC also a TD favorite over the next best conference.

That's pretty nuts.

[thebootfitter]

Sagarin Predictor Top-5 FCS Conferences

1.  MVFC60.86
2.  Colonial53.62
3.  Ohio Valley 53.22
4.  Big Sky52.26
5.  Southland50.29

Wow, that's quite a separation between MVFC and the rest.

[Hammersmith]

That's pretty nuts.

Yeah. Never seen anything close to that. Normally a big gap is three or four points. I think the overall suckage of FCS vs. FBS in week 1 is part of it. Only the MVFC as a conference really stood up to their FBS foes.

Oh, and that puts the MVFC above C-USA, MAC and Sun Belt.

[bisonaudit]

Based on Sagarin's RATING numbers (3.16 home field) and a 2.61 pt kicker for coming off a bye week, here's the outlook for the remainder of the regular season:

Weber State 98.2% (+0.1% from last week)
Incarnate Word 99.8% (+0.1%)
Montana 90.1% (+1.9%)
Western Illinois 97.5% (+0.7%)
Southern Illinois 86.2% (-3.5%)
Indiana State 99.2% (+0.1%)
South Dakota 93.0% (+0.6%)
South Dakota State 82.3% (+3.5%)
Northern Iowa 69.1% (+1.6%)
Missouri State 88.1% (+0.8%)
Youngstown State 90.0% (+2.2%)

12-0 30.9% (+15.4% from last week)
11-1 40.4% (+5.8%)
10-2 21.4% (-9.3%)
9-3 6.1% (-8.3%)
8-4 1.0% (-3.0%)
7-5 0.1% (-0.6%)
6-6 0.01% (-0.1%)
5-7 0.0003%
4-8 0.00001%
3-9 0.0000001%
2-10 0.0000000003%
1-11 0.0000000000003%

Assumes a normal distribution of outcomes w/ Sagarin's projected spread as the mean and a standard deviation of 13.86. In round numbers this means that if the spread is 14 the model predicts that roughly 2/3 of the outcomes will be between a tie and a 28 point victory. Half of the remaining outcomes or about 1/6 of all the outcomes would be a victory by more than 28 and there would be roughly a 1 in 6 chance that the team favored by 14 would lose.

[thebootfitter]

I'm confused. Look at Oregon. Their RATING is higher than any of the other three methods? How does that work?

I just took a closer look. I was assuming that the RATING was a weighted average, but the description states that it is a synthesis of the three different methods. I'd sure be interested to see how that formula works. Probably not just a simple weighted average. Or maybe there really is an error in his figures.

I ran a quick regression analysis to see if I could identify any patterns or consistencies between the three independent measures and the RATING. I may have missed something, but I wasn't able to identify anything. So my guess is that there's some additional variable(s) in the synthesis of the measures.