Conf......vs FBS.....vs P5

B1G....... .788 .... .667
SEC....... .741 .... .500
ACC....... .704 .... .583
PAC....... .640 .... .455
AAC....... .571 .... .353
B12....... .545 .... .333
MAC...... .424 .... .308
MWC..... .323 .... .176
SBC...... .269 .... .091
USA...... .125 .... .059

Man C-USA... wow.

Awful season for the MWC. Other than Boise State, SDSU and Air Force everyone in the conference is down. Props to the AAC for outperforming the Big 12.

The B1G is definitely top dog so far this season. Pretty good gap after the SEC & ACC to the PAC as well. I would say this ranking is pretty consistent with what we are seeing from on the field performance thus far.

(09-25-2016 07:36 AM)ken d Wrote:  Conf......vs FBS.....vs P5

B1G....... .788 .... .667
SEC....... .741 .... .500
ACC....... .704 .... .583

My P5 conference comparison has the B1G at #3, not #1, even though your stats show the B1G with the best overall OOC record and best vs other P5 as well. Why? Because my system does discriminate among quality, not all P5 wins are created equal nor are all G5 wins, my system gives more credit for beating ranked vs unranked teams.

FWIW, I looked at Massey's composite rankings, which combines 68 computer rankings. It has the B1G at #4.

Yep, beating the #25 according to some arbitrary measure is much higher quality than beating #26 according to the same arbitrary measure.

Not arbitrary at all.

(09-26-2016 11:32 AM)MplsBison Wrote:  Yep, beating the #25 according to some arbitrary measure is much higher quality than beating #26 according to the same arbitrary measure.

Not arbitrary at all.

Well actually, in my system you get win credit for beating G5 if ranked in "others receiving votes" as well, so maybe the dividing line is more like 33 vs 34? But I see your point, which is why my system isn't perfect. If it were perfect, credit would seemlessly reduce or increase depending on how good the other team is, rather than dropping off a cliff once an arbitrary value is reached.

But can you also see mine? I mean, if right now, school A were 2-1 vs a schedule of Alabama, Ohio State, and Louisville, we'd probably agree that school was better than a team B that was 3-1 (higher win %) against a schedule of wofford, FAMU, ECU, and Charleston Southern. So it makes even less sense to count all wins the same.

I'd say the fact that my system is a lot closer to the Massey rankings (which do include computers with seemless formulas) than is the raw win-loss % reported by ken_d indicates that it's better to err on the side I err on.

I see all the viewpoints. No system is even close to perfect.

To be fair, I completely misjudged the tone of your post. I apologize for that. It seems you were closer to "here's another system" than "my system is better than yours".

Keep in mind that I have no system, nor am I making any value judgment. All I am doing is counting raw wins and losses in OOC games and doing the math. I don't believe that's a particularly good way to rate conferences, and it's especially bad when some conferences have largely finished their OOC slate while others have a lot of OOC games yet to play (not to mention post season games which are almost all OOC.

FWIW, the Sagarin ratings, which won't be considered by Jeff Sagarin to be valid until after this week's games are complete, rank the conferences as follows:

SEC....79.8
PAC....78.4
B12....78.3
ACC....77.4
B1G....75.2
AAC....69.2
MWC...61.0
MAC...60.2
MVC...57.2 **
SBC....55.5
USA....55.0

** Missouri Valley Conference (FCS)

If the reason for lack of validity in early season is due to including carry over from last year as some say it does, then I would expect the Big XII to slip below the ACC next week (but probably not below the B1G). But computers can be inscrutable, so we'll just have to wait and see.

One conclusion you might draw from this is that B1G ADs are helping their schools' bowl prospects by astute scheduling.

I'm guessing that the difference between the top P5 conference and the bottom P5 is about as small as it's ever been. Should make CFP selection interesting this year, IMO.

(09-26-2016 12:11 PM)ken d Wrote:  Keep in mind that I have no system, nor am I making any value judgment.

I kept that in mind, that's why I didn't say "my system is better than yours" or something to that effect. I assumed you were just providing win-loss information, I didn't assume that you regarded that information as a 'system' you were endorsing for ranking the conferences.

PS - good point about Sagarin, and surely many of those other Massey computers, still including information from last year in their present rankings. IMO that is totally wrong and i don't do that. So maybe my system is more accurate than Massey right now, LOL.

(09-26-2016 11:59 AM)MplsBison Wrote:  I see all the viewpoints. No system is even close to perfect.

To be fair, I completely misjudged the tone of your post. I apologize for that. It seems you were closer to "here's another system" than "my system is better than yours".

Fair enough. I'll take some of the blame as well, because I didn't mean for my original post to have a "my system is better ..." tone. I assumed all along that ken was just providing the facts, not endorsing any particular interpretation of them. In fact, I gave him rep points for doing so before I even made my reply because I found the discrepancy between his facts and my system to be interesting and worth talking about.

Sorry about the initial confusion.

The problem with Sagarin's ratings right now is it still has last years data in it.

Here's Massey's ratings for FBS conferences:
1 SEC 1.91
2 P12 1.81
3 ACC 1.80
4 B10 1.78
5 B12 1.71
6 AAC 1.58
7 MAC 1.30
8 MWC 1.29
9 SBC 1.22
10 USA 1.06

Even when Sagarin no longer includes last year's ratings, there aren't enough games for the ratings to be very accurate.

would love to see the P5's road OOC record broken down.

(09-28-2016 12:04 AM)perimeterpost Wrote:  would love to see the P5's road OOC record broken down.

ACC
PAC 0-1
B12 0-1
B1G 3-1
SEC 3-2
ND 1-0
P5 7-5
AAC 4-3
MAC 1-0
USA 4-0
SBC 2-0
IND 1-0
FBS 19-8

PAC
ACC 1-3
B12 3-0
B1G 1-3
SEC 0-2
P5 5-5
MWC 6-3
USA 1-0
SBC 2-0
IND 2-1
FBS 16-9

B12
ACC 1-0
PAC 0-3
B1G 0-2
SEC 1-1
ND 1-0
P5 3-6
AAC 2-2
MWC 1-0
MAC 0-2
USA 4-0
SBC 1-0
IND 1-0
FBS 12-10

B1G
ACC 1-3
PAC 3-1
B12 2-0
SEC 1-0
ND 1-0
P5 8-4
AAC 4-1
MWC 6-0
MAC 5-2
USA 2-0
SBC 1-0
FBS 26-7

SEC
ACC 2-3
PAC 2-0
B12 1-1
B1G 0-1
P5 5-5
AAC 1-0
MAC 3-0
USA 5-1
SBC 4-1
IND 2-0
FBS 20-7

Ask and he shall receive.

Any way you slice it -- on pure W-L alone -- the Big Ten is the top dog, for now. Except, naturally, that we're 1-3 versus the second best overall conf.

Illinois got smoked by North Carolina, no surprise, but disappointed to see Indy lose to Wake and Penn St lose to Pitt -- both were very close scores.

those numbers aren't right. Those are overall games, not road OOC games.

(09-28-2016 10:43 AM)stever20 Wrote:  those numbers aren't right. Those are overall games, not road OOC games.

You are right. I didn't see the "road" qualifier. I'll see if I can get that (not immediately, because I have someplace to be this afternoon). But my spreadsheet doesn't generate that automatically (it would have to be way bigger than I am prepared to make it).

Why does road matter?? I think ken has done enough work. He's not your slave boy.

True road games (neutral site games excluded) against P5 opponents(all FBS road record in parentheses):

ACC...4-3 (9-5)
PAC...0-3 (3-5)
B12...0-4 (2-6)
B1G...2-2 (5-2)
SEC...1-2 (3-2)

Pretty small sample size.