Sagarin says

[NDSUstudent]

Liberty might have gotten in, they were 8-3(all DI) with a win vs Ball St.

[Bisonguy]

Liberty might have gotten in, they were 8-3(all DI) with a win vs Ball St.

Yup, and they couldn't beat a 6-5 CAA team last year.

There's a reason why it says "may" instead of "shall" when mentioning 7 DI wins for at-large selection. It should read "Not having 7 DI wins if you're not from one of the top 4 conferences shall remove you from the at-large selection process".

[Hammersmith]

Just looked and the numbers appear to have shifted... I see Wofford (48) and GSU (46) ahead of NDSU (63) and UNI (66).

Also the conference rankings appear to have shifted as well.

I always use the predictor column rather than the rating column. The predictor number is the core of Sagarin. The elo chess number is forced onto him by the BCS(it's pure wins/losses and doesn't take into account margin of victory). The rating column is a synthesis of the two.

From the explanation on his webpage:

In ELO CHESS, only winning and losing matters; the score margin is of no consequence, which makes it very "politically correct". However it is less accurate in its predictions for upcoming games than is the PURE POINTS, in which the score margin is the only thing that matters. PURE POINTS is also known as PREDICTOR, BALLANTINE, RHEINGOLD, WHITE OWL and is the best single PREDICTOR of future games. The ELO CHESS will be utilized by the Bowl Championship Series(BCS).

For the conference standings, I recalculate the central mean column using the predictor numbers. Also, the predicted margin of victory in my top charts already includes the home advantage(currently 3.08). Bisonaudit gives an extra 3.08 to teams coming off a bye week, but I'm too lazy to do that.

[Twentysix]

I always use the predictor column rather than the rating column. The predictor number is the core of Sagarin. The elo chess number is forced onto him by the BCS(it's pure wins/losses and doesn't take into account margin of victory). The rating column is a synthesis of the two.

From the explanation on his webpage:


For the conference standings, I recalculate the central mean column using the predictor numbers. Also, the predicted margin of victory in my top charts already includes the home advantage(currently 3.08). Bisonaudit gives an extra 3.08 to teams coming off a bye week, but I'm too lazy to do that.

Ill just let you post numbers, while I nod and smile.

[Tatanka]

I always use the predictor column rather than the rating column. The predictor number is the core of Sagarin. The elo chess number is forced onto him by the BCS(it's pure wins/losses and doesn't take into account margin of victory). The rating column is a synthesis of the two.

From the explanation on his webpage:


For the conference standings, I recalculate the central mean column using the predictor numbers. Also, the predicted margin of victory in my top charts already includes the home advantage(currently 3.08). Bisonaudit gives an extra 3.08 to teams coming off a bye week, but I'm too lazy to do that.

Well then.

[bisonaudit]

11-0 31.6%
10-1 44.5%
9-2 20.1%
8-3 3.5%
7-4 0.3%
6-5 0.0%

[westnodak93bison]

NDSU 68.11; Wabbits 46.04 plus 3 SDSU so Bizon by 19.07
Go Bizon!

[Hammersmith]

Sunday morning comes again and so do this week's Sagarin results. As always, I use the predictor column(furthest right) for everything and recalculate the conference standings using central mean.

In this episode, NDSU finally moves past UNI for top spot in the conference, but both teams get passed up by Wofford and Georgia Southern. Don't ask me how Wofford loses, yet takes the top spot; I don't have an answer. Illinois State gets a big boost from their win over Indiana State, but it doesn't change their standing within the conference. Missouri State moved up several spots due to the clustering in the bottom third of the MVFC. It also is clear that three of our four toughest games are yet to be played(according to the numbers).

In conference ratings, the SoCon solidifies its spot at the top of the AQ pack, but the Valley is making a charge and has taken second place over the CAA and Big Sky.

Week 8

			rate   predict actual	diff
61  Wofford		70.32
66  Georgia Southern	69.99
67  North Dakota State	69.90
75  Northern Iowa	68.45  + 4.59  
106 Illinois State	60.90  +12.14  +10.00   -2.14
122 Indiana State	58.88  + 7.88  
135 Youngstown State	57.24  +15.80  
141 Minnesota		56.27  +10.49  +13.00   +2.51
157 Southern Illinois	52.31  +14.45  + 6.00   -8.45
174 Missouri State	49.02  +24.02  +30.00	+5.98
176 South Dakota State	48.76  +18.00  +24.00	+6.00
181 Lafayette		47.51  +25.53  +36.00  +10.47
193 Western Illinois	45.55  +21.21  
237 Saint Francis-Pa	28.06  +44.98  +53.00   +8.02

Week 7

			rate   predict actual	diff
70  Northern Iowa	68.22  + 2.97  
71  North Dakota State	68.11
120 Illinois State	57.00  +14.19  +10.00   -4.19
131 Indiana State	55.29  + 9.74  
142 Minnesota		53.13  +11.90  +13.00   +2.10
135 Youngstown State	54.91  +16.28  
165 Southern Illinois	48.34  +16.69  + 6.00  -10.69
171 South Dakota State	46.04  +18.99  
181 Lafayette		43.70  +27.49  +36.00   +8.51
182 Western Illinois	43.46  +21.57  
184 Missouri State	43.13  +28.06  +30.00	+1.94
240 Saint Francis-Pa	20.58  +50.61  +53.00   +2.39

Conferences

Week 8
1. SoCn	59.94
2. GWC	59.25
3. MVFC	56.42
4. CAA	55.28
5. BSC	54.67

Week 7
1. GWC	55.49
2. SoCn	54.94
3. CAA	54.12
4. MVFC	53.31
5. BSC	53.24

[bisonaudit]

11-0 34.1%
10-1 44.8%
9-2 18.5%
8-3 2.5%
7-4 0.1%

SoCon conference increased rating 5 full points and Clemson is still undefeated so I'm thinking that SOS is floating the Wofford boat this week.

[X-Factor]

NDSU's SOS was better with the Kansas win and Montana also did not have an exclusively DI schedule which is another one of the guildelines. If NDSU would have lost at Kansas, there would have been a 6 DI win team most likely hosting a first round playoff game. I'd much rather have a 6 DI win SoCon, CAA, MVFC, or even Big Sky team than a 7 DI win Pioneer team.

A 7 win Pioneer will never make the playoffs. A 6 win SoCon, CAA, MVFC, Big Sky is better than a 10 or even 11 win Pioneer, and IMO the selection committee already knows that.