Dullard's College Football Ratings - Week 13

[dullard]

Ok here is the ratings for both sets of optimization settings. One optimization predicts the next week's winner quite well (75%-81% success) and next week's score isn't predicted as well (~13 point average error), the other set predicts the scores better (~12 point average error) but the winner prediction isn't as good (72%-78% success).

Win Place / Win Rating ( Score Place / Score Rating ) Team name
01 / 161.1 ( 01 / 142.0 ) Oklahoma
02 / 158.5 ( 02 / 138.6 ) Southern California
03 / 152.0 ( 04 / 133.9 ) Georgia
04 / 151.4 ( 03 / 136.2 ) LSU
05 / 146.9 ( 10 / 128.5 ) Ohio St.
06 / 146.8 ( 05 / 131.4 ) Michigan
07 / 146.4 ( 11 / 128.1 ) Washington St.
08 / 145.7 ( 08 / 131.2 ) Texas
09 / 145.3 ( 07 / 131.2 ) Florida St.
10 / 144.9 ( 06 / 131.2 ) Kansas St.
11 / 143.8 ( 09 / 129.3 ) Miami Ohio
12 / 143.5 ( 16 / 126.1 ) Purdue
13 / 143.3 ( 14 / 127.3 ) Iowa
14 / 143.3 ( 17 / 125.8 ) Tennessee
15 / 142.0 ( 13 / 127.3 ) Miami Florida
16 / 140.4 ( 19 / 125.1 ) Utah
17 / 140.3 ( 12 / 127.5 ) Arkansas
18 / 140.0 ( 15 / 126.9 ) Virginia Tech
19 / 137.6 ( 20 / 124.8 ) Nebraska
20 / 137.6 ( 21 / 124.3 ) Florida
21 / 137.4 ( 18 / 125.8 ) Minnesota
22 / 136.3 ( 23 / 123.9 ) Mississippi
23 / 136.2 ( 22 / 124.0 ) Boise St.
24 / 135.4 ( 24 / 123.7 ) Oklahoma St.
25 / 134.3 ( 28 / 120.7 ) Wisconsin
26 / 133.8 ( 25 / 122.5 ) Oregon St.
27 / 133.7 ( 26 / 121.9 ) Maryland
28 / 133.4 ( 30 / 120.1 ) TCU
29 / 133.3 ( 29 / 120.1 ) Pittsburgh
30 / 132.9 ( 33 / 119.7 ) West Virginia
31 / 132.4 ( 34 / 119.6 ) Michigan St.
32 / 131.4 ( 31 / 119.9 ) Texas Tech
33 / 131.2 ( 27 / 121.1 ) California
34 / 130.2 ( 36 / 119.2 ) North Carolina St.
35 / 129.7 ( 39 / 118.2 ) Clemson
36 / 129.4 ( 35 / 119.6 ) Missouri
37 / 129.1 ( 32 / 119.7 ) Auburn
38 / 129.0 ( 37 / 118.5 ) Colorado St.
39 / 128.7 ( 38 / 118.2 ) New Mexico
40 / 128.6 ( 48 / 115.7 ) Oregon
41 / 128.6 ( 40 / 117.5 ) Air Force
42 / 128.1 ( 43 / 116.4 ) Bowling Green
43 / 127.4 ( 45 / 115.9 ) Southern Miss
44 / 126.6 ( 51 / 115.1 ) Georgia Tech
45 / 125.6 ( 42 / 116.5 ) Alabama
46 / 125.1 ( 49 / 115.5 ) South Carolina
47 / 125.1 ( 46 / 115.7 ) Connecticut
48 / 124.6 ( 41 / 117.2 ) Virginia
49 / 124.0 ( 55 / 112.5 ) Notre Dame
50 / 124.0 ( 44 / 116.2 ) Memphis
51 / 123.5 ( 47 / 115.7 ) Syracuse
52 / 123.3 ( 50 / 115.4 ) Northern Illinois
53 / 123.3 ( 53 / 113.3 ) Louisville
54 / 122.9 ( 52 / 113.8 ) Boston College
55 / 121.6 ( 58 / 111.8 ) Marshall
56 / 121.6 ( 54 / 112.7 ) North Texas
57 / 121.0 ( 57 / 112.3 ) UCLA
58 / 120.5 ( 56 / 112.4 ) Wake Forest
59 / 119.6 ( 66 / 109.4 ) Northwestern
60 / 119.5 ( 60 / 111.4 ) Colorado
61 / 119.2 ( 59 / 111.4 ) Washington
62 / 118.0 ( 61 / 110.7 ) Tulsa
63 / 117.9 ( 68 / 109.2 ) Fresno St.
64 / 117.8 ( 63 / 110.0 ) Toledo
65 / 117.7 ( 70 / 108.6 ) Stanford
66 / 117.6 ( 65 / 109.9 ) Kansas
67 / 117.2 ( 67 / 109.4 ) UNLV
68 / 116.9 ( 62 / 110.6 ) Navy
69 / 116.1 ( 71 / 108.5 ) Hawaii
70 / 115.5 ( 64 / 110.0 ) Penn St.
71 / 113.3 ( 73 / 106.9 ) Arizona St.
72 / 113.3 ( 72 / 107.5 ) Wyoming
73 / 112.6 ( 79 / 105.0 ) Texas A&M
74 / 112.5 ( 77 / 105.5 ) UAB
75 / 112.5 ( 78 / 105.3 ) Houston
76 / 112.3 ( 69 / 108.9 ) Kentucky
77 / 112.1 ( 76 / 105.5 ) Brigham Young
78 / 111.8 ( 74 / 106.3 ) San Diego St.
79 / 111.6 ( 75 / 106.0 ) Rutgers
80 / 110.0 ( 81 / 104.3 ) Louisiana Tech
81 / 109.2 ( 80 / 104.8 ) South Florida
82 / 108.4 ( 84 / 103.3 ) Nevada
83 / 107.7 ( 82 / 103.6 ) Cincinnati
84 / 107.3 ( 83 / 103.3 ) Akron
85 / 105.0 ( 091 / 98.8 ) Arizona
86 / 104.7 ( 85 / 102.0 ) Western Michigan
87 / 103.6 ( 089 / 99.3 ) Troy St.
88 / 103.2 ( 088 / 99.6 ) Iowa St.
89 / 102.3 ( 086 / 99.7 ) North Carolina
90 / 102.1 ( 090 / 99.1 ) Duke
91 / 101.4 ( 092 / 98.1 ) Tulane
92 / 100.2 ( 096 / 96.8 ) Mississippi St.
93 / 100.0 ( 087 / 99.7 ) Vanderbilt
94 / 100.0 ( 100 / 95.5 ) Baylor
095 / 99.4 ( 095 / 96.9 ) Illinois
096 / 98.7 ( 098 / 95.8 ) Rice
097 / 98.6 ( 093 / 97.5 ) Kent St.
098 / 98.4 ( 099 / 95.7 ) Ball St.
099 / 97.8 ( 097 / 96.6 ) Utah St.
100 / 97.7 ( 094 / 96.9 ) Middle Tennessee
101 / 96.8 ( 101 / 95.1 ) San Jose St.
102 / 93.9 ( 104 / 93.2 ) Louisiana-Lafayette
103 / 93.7 ( 102 / 94.8 ) Temple
104 / 93.2 ( 105 / 92.6 ) Indiana
105 / 91.9 ( 103 / 93.7 ) New Mexico St.
106 / 90.1 ( 106 / 92.3 ) Idaho
107 / 90.0 ( 108 / 90.4 ) Arkansas St.
108 / 88.0 ( 107 / 91.8 ) Ohio
109 / 86.6 ( 109 / 89.5 ) UCF
110 / 84.4 ( 111 / 88.0 ) East Carolina
111 / 83.3 ( 112 / 85.5 ) Eastern Michigan
112 / 83.1 ( 110 / 88.2 ) Central Michigan
113 / 81.9 ( 114 / 84.4 ) Buffalo
114 / 81.6 ( 113 / 85.3 ) Louisiana Monroe
115 / 78.9 ( 116 / 81.7 ) UTEP
116 / 77.9 ( 115 / 82.1 ) SMU
117 / 75.1 ( 117 / 80.3 ) Army

 

[sciencewhiz]

Even with the win optimization, it looks like your computer predicts that Michigan will win (I assume you still adjust for home field advantage)

 

[Ogg]

aaah so LSU will beat Ole Miss then beat Arkansas and then we should have a shot to jump up to #2 to play for the championship right?

 

[dullard]

Originally posted by: sciencewhiz
Even with the win optimization, it looks like your computer predicts that Michigan will win (I assume you still adjust for home field advantage)

The scores above do not take home field advantage into account, you must add that to the home team. So the win optimization says Ohio State and Michigan are basically tied. But if they play at Michigan, I'd give Michigan about a 3 point advantage. The reason I don't put home field advantage in the scores is since I do not feel it is a fair thing to add in using a formula - some teams play great on the road, others don't.

 

[hdeck]

i don't like your rankings anymore

 

[dullard]

Originally posted by: awg
aaah so LSU will beat Ole Miss then beat Arkansas and then we should have a shot to jump up to #2 to play for the championship right?

I still think the voters will vote OSU into 2nd place (2nd place of the BCS) if OSU beats Michigan. I don't think LSU will get a chance unless OSU and USC both lose.

 

[dullard]

Originally posted by: hdeck
i don't like your rankings anymore

Which team got shafted this time?

 

[Tom]

Just so you know, the Buckeyes are playing Michigan at Michigan..

 

[NeoV]

I'm a buckeye fan, and I'm ok with your rankings now...though Georgia isn't that good......voters don't determine the BCS, that is only 1 part of it...

Here is what I hope happens - Oklahoma wins their regular season and Big 12 championship game - very likely..
Ohio St beats Michigan
LSU wins out
USC wins out..

now, that would likely leave the Sugar bowl as Oklahoma - Ohio St, and another bowl could have USC - LSU - all that complaing about who is #2 vs #3 vs #4 would hopefully end all this BCS crap and get a real playoff system going!

 

[Dudd]

Originally posted by: NeoV
I'm a buckeye fan, and I'm ok with your rankings now...though Georgia isn't that good......voters don't determine the BCS, that is only 1 part of it...

Here is what I hope happens - Oklahoma wins their regular season and Big 12 championship game - very likely..
Ohio St beats Michigan
LSU wins out
USC wins out..

now, that would likely leave the Sugar bowl as Oklahoma - Ohio St, and another bowl could have USC - LSU - all that complaing about who is #2 vs #3 vs #4 would hopefully end all this BCS crap and get a real playoff system going!

Throw in an OU loss somewhere along the way, and then we'll be in complete agreement. Oh yeah, and have TCU win out and then be able to complain of getting the shaft. If that doesn't kill the stupid BCS system, then nothing will.

 

[poopaskoopa]

NCSU should be higher and FSU should be lower.... That damn kicker...

 

[dullard]

Originally posted by: Dead Parrot Sketch
Just so you know, the Buckeyes are playing Michigan at Michigan..

I always try to say "if" when talking about future events. Who knows a tornado might go through the Michigan field during the week...

I'm a buckeye fan, and I'm ok with your rankings now...though Georgia isn't that good......voters don't determine the BCS, that is only 1 part of it...

In this year, the voters may very well deterimine the BCS result. Look at the 1 loss teams - quality wins will have a negligible effect, # losses will be the same, SOS will be quite close (not for LSU I realize), and computers are quite close between Ohio State, USC, and LSU. So in this case, the voters have a significant impact on the final results. If they swap OSU and LSU for 3rd and 4th place (quite a likely occurance looking at the chance for OSU to defend the national championship, looking at LSU's strength of schedule, and looking at the computer rankings telling the voters that OSU is better), then USC has no chance at all of getting into the national championship game.

 

[sciencewhiz]

Originally posted by: dullard

Originally posted by: hdeck
i don't like your rankings anymore

Which team got shafted this time?

Looking at his signature, he doesn't think that Michigan, Georgia, and WSU are better then Texas.

 

[Alkaline5]

Originally posted by: NeoV
...all that complaing about who is #2 vs #3 vs #4 would hopefully end all this BCS crap and get a real playoff system going!

But then what would we have to talk about?

 

[sciencewhiz]

Originally posted by: Alkaline5

Originally posted by: NeoV
...all that complaing about who is #2 vs #3 vs #4 would hopefully end all this BCS crap and get a real playoff system going!

But then what would we have to talk about?

That's exactly why I like college football

 

[Nitemare]

Originally posted by: poopaskoopa
NCSU should be higher and FSU should be lower.... That damn kicker...

NCSU had them on the ropes and let them get back in it...imho Austin Herbert, the entire defense and all those that fumbled the ball.

I mean 272 yards rushing allowed as well as a 71 yard in which at least 6 people should have tackled Booker

It's Amato's fault for letting Norm Chow go as well as not having a decent backup for an injury prone TA McLendon.

Too fix them for next season.
-Hire a defensive coordinator
-Your offensive coordinator is a tailbacks coach..so make him work with Josh Brown, Tramaine Hall & Cotra Jackson and spend less time with TA, let TA heal.
-Jay Davis will probably be pretty good at QB, but tough shoes to fill.
-Get a good backup QB and red shirt him.
-The wide-outs that are returning will probably be best in the ACC and hopefully help J.Davis
-The defense can't get any worse, let's hope that they learned this year and a defensive coordinator can whip them in shape
-Get rid of the stupid penalties. FSU's defense can commit personal fouls because they good you are suck, if it was not for UNC it would be league worse

 

[sciencewhiz]

Originally posted by: dullard
In this year, the voters may very well deterimine the BCS result. Look at the 1 loss teams - quality wins will have a negligible effect, # losses will be the same, SOS will be quite close (not for LSU I realize), and computers are quite close between Ohio State, USC, and LSU. So in this case, the voters have a significant impact on the final results. If they swap OSU and LSU for 3rd and 4th place (quite a likely occurance looking at the chance for OSU to defend the national championship, looking at LSU's strength of schedule, and looking at the computer rankings telling the voters that OSU is better), then USC has no chance at all of getting into the national championship game.

Well, I beleive you even less then when you said this last week ;-) LSU and Ohio State are now seperated by 39 points in the coaches poll, but it was only 29 points last week. In the AP Poll, they are seperated by 43 points, up from 32 points last week.

However, if LSU loses one of it's remaining games and OSU wins, then USC has no chance.

 

[dullard]

Originally posted by: sciencewhiz
Well, I beleive you even less then when you said this last week ;-) LSU and Ohio State are now seperated by 39 points in the coaches poll, but it was only 29 points last week. In the AP Poll, they are seperated by 43 points, up from 32 points last week.

However, if LSU loses one of it's remaining games and OSU wins, then USC has no chance.

Like I said before, the voters will do this at the end of the regular season IF OSU isn't in the top 2. Right now, they have no reason since OSU is in the top 2 and since OSU still has a major game ahead. I'd love to see OSU win out (including the national championship), but I see many possibilities that USC may get back in:
1) OSU will get a bit better in SOS from now until the end, but USC will gain a lot of ground now that Arizona is out of the picture.
2) USC is almost guaranteed a bigger quality win bonus since Washington St WILL move up in the BCS if beats an easy Washington team.
3) OSU cannot any more ground in the computers (unless OU loses but that helps USC just as much in the computers), but USC will be helped if any of these teams lose: OSU, LSU, Texas, TCU, or Florida (as it is Florida will have no effect, since one computer is tossed out, but if things change a Florida loss will help USC). In fact USC will likely be helped by the computers even if TCU wins out (which I think won't happen, TCU plays their toughest team of all with just 4 days rest after Cincinnatti).
4) USC will be helped in the polls if LSU loses. And LSU has a tough schedule ahead (3 hard games and I'm not sure if LSU can win all of those).

For all of these reasons it is quite possible that USC will jump back into 2nd. If that occurs, and only if it occurs, I think the voters will vote OSU at least into 3rd place to keep OSU in the national championship game.

 

[sciencewhiz]

I can guarantee that USC won't get back to #2 until their Dec 6th game with OSU. At that point, the will have beaten a pretty good Oregon State team, that will help SOS (and maybe a computer or two). Also, several teams above WSU will lose in Conference Championships, giving USC a bigger QW. But, at that point, Voters won't have time to correct, because that's also the last poll.

But we'll see. We still have 3 weeks left to argue about it

 

[tigerbait]

and Georgia is ahead of LSU how?




the following is from an LSU fan website

YOU HEARD IT HERE, FOLKS...

"If Michigan beats Ohio State and USC wins out and if LSU wins out,
including a victory in the SECCG, LSU will be ranked #2 in the BCS."

Here's how:
#1) Ohio State must lose to Michigan this weekend and LSU must win out.

#2) LSU currently at 12.21, is 4.50 points behind USC at 7.71. LSU sos will
be drastically helped over each of the next THREE weeks. They currently have
a 2.88 sos rank, which is determined by dividing 25 into whatever schedule
rank we are (our current sched.rank = 72).
The final records of our next three opponents will be most likely be: OM
(9-3), Ark (8-4), and UGA or UT (10-2), a combined opponent record of 27-9.
The win a/g OM is a/g the #15 bcs team, on the road. Win a/g ark will be a/g
possibly #25 bcs. Win a/g UT or UGA will be a/g #6 0r #7 bcs.
Taking just these factors into consideration, our schedule rank is predicted
to move up into the 30's. That would give us a sos rank of about 1.2-1.5,
approximately 1.25 less than our current 2.88. Another words 1.25 closer to
USC

#3) USC sos will be somewhat hurt over each of the next two weeks. USC plays
UCLA who will finish at (6-6) and Oregon St. who will finish at best at
(8-4), a combined opponent record of 14-10. USC sched. rank is currently at
37, giving them a sos rank of 1.48. Some predictions put their sced. rank up
to near 50 by the end of their season, giving them a sos rank of around 2.
Thus, their BCS total goes up by about .52 as a result of this factor.

#3) After Ohio State loses this weekend and LSU and USC win, both LSU and
USC will move up in most of the computer rankings. LSU is currently at 5.83,
which is the average of the best 6 of the 7 computer ranking services used.
USC is currently at 3.33, 2.5 points better than LSU. IF these two teams are
vying for BCS #2 after the OSU loss, there is much more room for LSU to move
up than USC. 2 is the highest either can attain with Oklahoma undefeated and
at #1. LSU will never move up to a 2 computer average unless USC loses, but
point is they will gain more than USC will gain after the OSU loss and at
season's end, considering the sos argument above. USC can move up 1.33
points max and LSU 3.83 - a 2.5 point difference. LSU will gain at least
half that by season' end, putting LSU 1.25 points closer to USC as a result
of this category.

#4) By season's end, USC will be 11-1, with its last three wins over
unranked Arizona, UCLA, and Oregon St. LSU will be 12-1, having beaten #15
Ole Miss, #25 Arkansas, and #6 Georgia or #7 Tennessee. Since the AP and
ESPN polls are measured by the subjective votes of humans who are free to
place their vote with whomever they choose, I see the above factors as
influencing enough voters to give LSU more votes than USC, thus moving LSU
into #2 in the AP and ESPN and giving LSU a poll average of 2, up from their
current average of 3. In so doing, USC would necessarily drop from 2 to 3 in
poll average, thus netting LSU a 2 point gain in the BCS.

#5) As lagniappe, if LSU beats UT in the SECCG, they will acquire additional
quality win point deduction based on where UT ends up in the BCS after
sustaining that loss. If UT beats out UGA by tiebreaker to get to SECCG, UGA
will not drop, so we will still have the QW point deduction from beating UGA
earlier in the year. After the loss, UT would probably end up BCS #9 or #10,
giving LSU another .1 or .2 less in total BCS points.

Adding it all up (difference of end of season bcs rankings as compared to
current):
LSU sos minus 1.25
USC sos plus .52
LSU computer average, minus 1.25
LSU poll average, minus 1
USC poll average, plus 1

TOTAL = 5.02
Since LSU is 4.50 points behind USC right now, these factors combined would
put LSU at BCS #2, with a .52 point cushion for error, but not factoring in
any additional QW deduction LSU would get from beating UT in the SECCG.


Just my guesses - I don't purport to be a mathemetician. In fact, does
anyone really get the BCS?

:beer::beer:

 

[dullard]

Originally posted by: tigerbait
and Georgia is ahead of LSU how?

In the score rankings, LSU is soundly ahead of Georgia - meaning that LSU is predicted to score more points than Geogia if they played. In the Win optimization, they are basically tied (anything less than 1 point is meaningless). So I wouldn't say Georgia is ahead of LSU in my computer. My thoughts on the rest of your post:

#1) Ohio State must lose to Michigan this weekend and LSU must win out.

A given. Without that happening LSU has no chance.

Taking just these factors into consideration, our schedule rank is predicted to move up into the 30's. Another words 1.25 closer to USC.

You have to also include all the opponents that LSU already played and how well they do in the rest of the season. I think the quote above assumes that everyone else that LSU already played wins out as well. While that can happen, I don't think it will. I'd expect LSU right around 40. not in the lower 30s. But we will see.

Some predictions put their sced. rank up to near 50 by the end of their season, giving them a sos rank of around 2. Thus, their BCS total goes up by about .52 as a result of this factor.

Again that ignores how well all of USC's previous opponents play in the rest of the season. Most of the predictions I see is for USC's strength of schedule to get better not 0.52 points worse.

#3) After Ohio State loses this weekend and LSU and USC win, both LSU and USC will move up in most of the computer rankings. LSU will gain at least half that by season' end, putting LSU 1.25 points closer to USC as a result of this category.

If USC and LSU win out and OSU loses, I'd expect USC to move up to 2.3 and LSU to move up to 3.8. That would give a boost of 1 to LSU. The 1.25 estimate is also possible. I'll give you that.

I see the above factors as influencing enough voters to give LSU more votes than USC, thus moving LSU into #2 in the AP and ESPN and giving LSU a poll average of 2, up from their current average of 3. In so doing, USC would necessarily drop from 2 to 3 in poll average, thus netting LSU a 2 point gain in the BCS.

That is quite wishful thinking. Polls are subjective, yes, but there is a mile between USC and LSU. 80 points in the polls are very difficult to make up. That would require virtually all of the voters switching their viewpoint on USC and LSU. I don't think it will happen.

#5) After the loss, UT would probably end up BCS #9 or #10, giving LSU another .1 or .2 less in total BCS points.

A UT loss will give them at least 1 point from the loss, as well as at least 5 points from the computer and polls. If UT scraps by with just a 5 point loss in those two sections, that would put them at a tie for #10 with FSU - small little factors are too difficult to estimate now as to whetere UT would be ranked #10 or #11 (would a 2 loss Florida state be ranked behind UT?). Lets give them the benefit of the doubt and put UT at #10. But that would mean Washington St jumps over UT giving USC an equivalent 0.1 point boost. Of course UGA would probably go up one spot jumping OSU - giving LSU a 0.1 point advantage.

Adding it all up I'd give LSU this:
SOS - 1.0 boost (including both teams)
Computer average - 1.25 boost
Poll average - no boost
Quality wins - 0.1 boost

TOTAL = 2.35 - not enough. I think there are just too many ifs in this post to happen. Lets talk about it in 2 weeks.

 

[sciencewhiz]

I've seen LSU getting to the low 40s for SOS, but that's for the regular season. The SEC Championship should get them better. Assuming the team they beat is 10-3 afterwards, and LSUs Opponent's Opponents keep the same winning percentage, LSU would project to the high 20s/low 30s in SOS.

Still, LSU would need to jump USC in the polls, and that will only happen if LSU wins big in all 3 games and USC squeaks out wins against UCLA and Oregon State.

 

[Tom]

Keep it up. The more of y'all that are counting on the Buckeyes losing the more likely it is they will win.