Dullard's College Football Week 13

[dullard]

Not much mixing up going on these days. This always happens as there are more games. One game averaged in among 10 other games has little effect. Other than the 3 decent teams that were upset in the top 25, not much happened.

Place : Win Rating / Score Rating ( W , L ) Team name
01 : 83.0 / 85.5 ( 10 , 0 ) Southern Cal
02 : 79.2 / 78.4 ( 11 , 0 ) Oklahoma
03 : 78.4 / 83.5 ( 9 , 1 ) California
04 : 75.6 / 77.2 ( 11 , 0 ) Utah
05 : 72.5 / 71.5 ( 9 , 1 ) Texas
06 : 72.1 / 69.3 ( 8 , 2 ) Arizona St
07 : 72.0 / 72.0 ( 11 , 0 ) Auburn
08 : 66.9 / 67.2 ( 8 , 2 ) Miami FL
09 : 66.0 / 62.3 ( 10 , 0 ) Boise St
10 : 66.0 / 65.3 ( 7 , 3 ) Texas A&M
11 : 65.4 / 69.0 ( 8 , 1 ) Louisville
12 : 63.5 / 64.2 ( 8 , 2 ) Georgia
13 : 63.4 / 59.8 ( 9 , 2 ) Iowa
14 : 62.4 / 62.3 ( 7 , 3 ) Oklahoma St
15 : 62.4 / 62.7 ( 8 , 2 ) Virginia
16 : 61.3 / 58.5 ( 9 , 2 ) Michigan
17 : 60.7 / 61.3 ( 8 , 3 ) Florida St
18 : 60.5 / 56.8 ( 8 , 2 ) Tennessee
19 : 60.4 / 58.5 ( 8 , 2 ) LSU
20 : 57.8 / 58.6 ( 8 , 2 ) Virginia Tech
21 : 56.4 / 56.6 ( 6 , 5 ) Oregon St
22 : 56.0 / 52.4 ( 9 , 2 ) Wisconsin
23 : 55.8 / 55.6 ( 6 , 4 ) Texas Tech
24 : 55.0 / 58.4 ( 7 , 4 ) Purdue
25 : 54.7 / 57.0 ( 6 , 4 ) UCLA
26 : 53.8 / 53.7 ( 6 , 4 ) Notre Dame
27 : 53.6 / 50.9 ( 8 , 2 ) UTEP
28 : 53.3 / 57.2 ( 7 , 4 ) Florida
29 : 53.0 / 49.8 ( 6 , 4 ) Colorado
30 : 52.9 / 52.0 ( 7 , 4 ) Ohio State
31 : 52.8 / 51.0 ( 8 , 2 ) Boston College
32 : 52.7 / 51.8 ( 8 , 2 ) West Virginia
33 : 52.0 / 50.0 ( 7 , 4 ) New Mexico
34 : 51.4 / 54.0 ( 7 , 3 ) Fresno St
35 : 50.1 / 54.0 ( 5 , 5 ) Arkansas
36 : 49.6 / 45.9 ( 6 , 5 ) North Carolina
37 : 49.5 / 45.0 ( 6 , 4 ) Iowa St
38 : 47.7 / 46.5 ( 6 , 5 ) Clemson
39 : 46.8 / 44.0 ( 6 , 4 ) Georgia Tech
40 : 46.5 / 41.1 ( 8 , 2 ) Navy
41 : 46.1 / 42.9 ( 6 , 5 ) Northwestern
43 : 46.0 / 41.4 ( 7 , 3 ) Alabama-Birmingham
42 : 46.0 / 45.7 ( 5 , 6 ) Brigham Young
44 : 45.9 / 47.4 ( 5 , 6 ) Oregon
45 : 45.4 / 49.7 ( 4 , 7 ) Stanford
46 : 45.1 / 48.1 ( 6 , 5 ) Alabama
47 : 45.1 / 45.2 ( 6 , 5 ) South Carolina
48 : 44.6 / 47.9 ( 8 , 2 ) Bowling Green
49 : 44.6 / 41.5 ( 6 , 3 ) Pittsburgh
50 : 44.5 / 43.5 ( 5 , 6 ) Washington St
51 : 44.4 / 46.0 ( 6 , 5 ) Minnesota
52 : 43.5 / 42.5 ( 6 , 4 ) Cincinnati
53 : 43.1 / 41.3 ( 6 , 5 ) Wyoming
54 : 42.6 / 38.8 ( 7 , 3 ) Memphis
55 : 41.9 / 44.9 ( 4 , 7 ) Kansas
56 : 41.9 / 43.8 ( 4 , 6 ) North Carolina St
57 : 41.7 / 40.1 ( 5 , 5 ) Nebraska
59 : 41.1 / 41.4 ( 5 , 6 ) Air Force
58 : 41.1 / 40.0 ( 7 , 4 ) Troy St
60 : 40.9 / 42.9 ( 5 , 6 ) Michigan St
61 : 40.1 / 42.0 ( 4 , 7 ) Kansas St
62 : 39.3 / 40.3 ( 4 , 6 ) Maryland
63 : 38.8 / 38.1 ( 6 , 4 ) Connecticut
64 : 38.7 / 36.2 ( 5 , 5 ) Syracuse
65 : 38.6 / 37.8 ( 8 , 3 ) Northern Illinois
66 : 38.4 / 41.0 ( 4 , 6 ) Missouri
67 : 38.3 / 35.4 ( 5 , 5 ) TCU
68 : 37.4 / 37.5 ( 4 , 7 ) Colorado St
69 : 37.3 / 29.3 ( 7 , 4 ) North Texas
70 : 37.2 / 33.3 ( 5 , 4 ) Southern Miss
71 : 36.7 / 35.8 ( 8 , 3 ) Miami OH
72 : 36.7 / 40.8 ( 4 , 7 ) Penn St
73 : 36.5 / 31.0 ( 5 , 6 ) Louisiana Tech
74 : 35.6 / 36.2 ( 4 , 6 ) Wake Forest
75 : 35.2 / 33.4 ( 7 , 3 ) Toledo
76 : 33.1 / 33.1 ( 4 , 7 ) San Diego St
77 : 33.0 / 34.6 ( 2 , 8 ) Arizona
78 : 32.6 / 32.1 ( 3 , 7 ) Mississippi
79 : 32.2 / 27.8 ( 5 , 6 ) New Mexico St
80 : 31.5 / 32.7 ( 6 , 5 ) Marshall
81 : 31.3 / 28.6 ( 4 , 5 ) South Florida
83 : 30.9 / 25.0 ( 3 , 8 ) Baylor
82 : 30.9 / 27.1 ( 4 , 5 ) Hawai`i
84 : 29.6 / 30.6 ( 3 , 7 ) Mississippi St
85 : 29.5 / 26.8 ( 4 , 6 ) Rutgers
86 : 27.0 / 24.4 ( 4 , 5 ) Tulane
87 : 26.9 / 28.6 ( 3 , 8 ) Indiana
88 : 26.7 / 27.5 ( 3 , 8 ) Illinois
89 : 26.0 / 24.1 ( 3 , 8 ) Houston
90 : 25.8 / 21.4 ( 5 , 6 ) Middle Tenn St
91 : 24.9 / 20.8 ( 6 , 5 ) Akron
92 : 24.2 / 16.7 ( 5 , 6 ) UL-Monroe
93 : 24.0 / 22.9 ( 3 , 7 ) Rice
94 : 23.9 / 26.0 ( 1 , 10 ) Washington
95 : 23.5 / 22.8 ( 2 , 9 ) UNLV
96 : 23.2 / 26.1 ( 2 , 9 ) Vanderbilt
97 : 23.0 / 23.0 ( 2 , 8 ) Kentucky
98 : 22.8 / 21.9 ( 2 , 9 ) Duke
99 : 20.9 / 17.7 ( 5 , 6 ) Nevada
100 : 20.8 / 14.4 ( 3 , 8 ) SMU
101 : 20.8 / 19.9 ( 3 , 8 ) Tulsa
102 : 20.7 / 23.2 ( 4 , 6 ) Kent St
103 : 20.0 / 17.6 ( 4 , 7 ) UL-Lafayette
104 : 19.7 / 19.7 ( 2 , 8 ) Army
105 : 18.3 / 13.0 ( 3 , 8 ) Utah St
106 : 17.5 / 16.4 ( 2 , 9 ) Temple
107 : 17.4 / 14.7 ( 2 , 8 ) East Carolina
108 : 16.1 / 10.8 ( 3 , 8 ) Arkansas St
109 : 14.4 / 7.6 ( 3 , 8 ) Idaho
110 : 14.0 / 14.3 ( 4 , 7 ) Ohio U.
111 : 12.0 / 10.3 ( 2 , 8 ) San Jos? St
112 : 10.0 / 6.6 ( 4 , 7 ) Eastern Michigan
113 : 9.4 / 6.1 ( 4 , 7 ) Central Michigan
114 : 7.4 / 7.1 ( 2 , 9 ) Ball St
115 : 1.7 / 0.0 ( 2 , 9 ) Buffalo
116 : 0.3 / 1.9 ( 0 , 10 ) Central Florida
117 : 0.0 / 0.9 ( 1 , 10 ) Western Michigan

 

[dullard]

Division I-AA top 25

Place : Score Rating ( W , L ) Team name
01 : 94.7 ( 10 , 0 ) Harvard
02 : 94.5 ( 9 , 2 ) Georgia Southern
03 : 94.2 ( 10 , 1 ) Southern Illinois
04 : 90.7 ( 9 , 2 ) New Hampshire
05 : 88.5 ( 9 , 2 ) William & Mary
06 : 86.4 ( 9 , 2 ) James Madison
07 : 86.3 ( 9 , 2 ) Furman
08 : 86 ( 8 , 3 ) Eastern Washington
09 : 84.7 ( 8 , 3 ) Delaware
10 : 83.5 ( 9 , 2 ) Cal Poly SLO
11 : 83.5 ( 9 , 2 ) Montana
12 : 82.7 ( 6 , 5 ) Villanova
13 : 81.7 ( 7 , 3 ) Florida Atlantic
14 : 81.5 ( 9 , 2 ) Sam Houston St
15 : 80.8 ( 7 , 4 ) Northern Iowa
16 : 80.6 ( 8 , 2 ) Pennsylvania
17 : 80 ( 9 , 2 ) Western Kentucky
18 : 78.9 ( 9 , 1 ) Jacksonville St
19 : 77.9 ( 7 , 4 ) Portland St
20 : 77.4 ( 9 , 2 ) Lehigh
21 : 77.2 ( 5 , 6 ) Maine
22 : 77.1 ( 5 , 6 ) Hofstra
23 : 76.2 ( 8 , 3 ) Northwestern St
24 : 76 ( 6 , 5 ) Massachusetts
25 : 75.9 ( 5 , 6 ) Northeastern

 

[kalster]

wow, ucla at 25,

 

[Alkaline5]

dullard, why do you thing Texas A&M is still ranked so high even after the loss to Baylor? I know they've had a great schedule, but your rankings have them placed about the same as the BCS computers' and I'm just wondering how a loss to such a crappy team doesn't penalize them more...

Originally posted by: dullard
Division I-AA top 25

01 : 94.7 ( 10 , 0 ) Harvard

I won't claim to know anything about I-AA football, but I don't see how the selection committee could leave the only undefeated I-AA team out of the playoffs this year.

 

[wyvrn]

After the schedule Auburn has had to play, why in the hell are they ranked as low as 7? That's just insanely stupid. I don't have any affiliation with Auburn, but I think they have proven they are as good as anyone this year.

Why is Notre Dame so high? They are my favorite college team, but I don't think they deserve to be 26th in the nation with the stupid losses they have had. I know their schedule is tough, but they are almost a .500 ballclub.

 

[sciencewhiz]

Remember that good wins help to cancel out stupid losses (that applies to both Notre Dame and Texas A&M).

 

[dullard]

Originally posted by: wyvrn
After the schedule Auburn has had to play, why in the hell are they ranked as low as 7? That's just insanely stupid. I don't have any affiliation with Auburn, but I think they have proven they are as good as anyone this year.

Why is Notre Dame so high? They are my favorite college team, but I don't think they deserve to be 26th in the nation with the stupid losses they have had. I know their schedule is tough, but they are almost a .500 ballclub.

Strength of schedule is the biggest factor in my ratings. Here is how mine and many other computers work:
[*]Even case: If (1) your average opponent rating is at 50 points, (2) you win 6 games and lose 6 games, and (3) the average score difference in all your wins and all your losses is zero points, then you are rated at 50 points. I don't think anyone would disagree with this judgement.

[*]SOS case: If (1) your average opponent rating is at 55 points, (2) you win 6 games and lose 6 games, and (3) the average score difference in all your games is zero points, then how should you be rated? Obviously the looks of things are the same as in the even case above - zero point margin on average and same number of wins and losses. However, you played better opponents. Thus you definately should be rated higher. And using the same logic as in the even case, you deserve to be rated at 55 points.

[*]Win case: If (1) your average opponent rating is at 50 points, (2) you win 7 games and lose 5 games, and (3) the average score difference in all your games is zero points, then how should you be rated? Obviously there is a difference between this case and the case above. While this team on average played the same, they happened to pull out a win in a case that the team above didn't. Thus this team should be rated higher than 50. How much higher is the key.

[*]Score case: If (1) your average opponent rating is at 50 points, (2) you win 6 games and lose 6 games, and (3) the average score difference in all your games is +3 points on your side, then how should you be rated? On average, this team scored more points and did a bit better than the teams above. Thus this team should be rated higher than the even team. Thus this team should be rated higher than 50. How much higher is the key. And whether or not this better scoring justifies rating higher than the team with one more win is debatable.

So in 2 of the 4 cases, the rating is easy to see. SOS is the major factor in a computer rating. How you weight the bonus for an extra win and how you rate the bonus for more points score (or conversely, less points allowed) will vary from computer program to computer program. And in fact, I give two different weightings above.

Despite what you think, Auburn has played a fairly weak SOS. Division I-AA Citadel (a crappy I-AA team by the way), bottom of the I-A barrel UL-Monroe and Kentucky, and two additional teams with only 3 wins. All together, my computer says Aurburn's SOS is at #68. That means 67 teams played tougher schedules. That is devistating in any computer program. Compare that to #8 SOS for USC, #17 SOS for Oklahoma, #54 SOS for Utah, and #82 SOS for Boise St. Yes, Utah is given a tougher schedule than Auburn in computer programs. Playing Tennessee again will signficantly help here and a win by more than 12 points may very well put Auburn into the #5 spot (if all other teams play as expected).

Notre Dame has played a fairly tough schedule (#21 in my program). I don't think they are rated too much higher than they should be. Parity is finally setting in one the middle of college football. Top rated teams and bottom rated teams haven't shown parity issues yet (as seen by undefeated teams and an unwinning team). But the middle of the pack is huge this year with nearly all teams with about a .500 record. Someone with a nearly 0.500 record must sit on top of that mushy middle section. Here are the teams Notre Dame is ranked significantly ahead of that have fewer losses: Fresno St, Navy, Memphis, UAB, Bowling Green, Miami OH, and Toledo. Which of these teams would you rate above Notre Dame? And Why?

 

[dullard]

Originally posted by: Alkaline5
dullard, why do you thing Texas A&M is still ranked so high even after the loss to Baylor? I know they've had a great schedule, but your rankings have them placed about the same as the BCS computers' and I'm just wondering how a loss to such a crappy team doesn't penalize them more...

I won't claim to know anything about I-AA football, but I don't see how the selection committee could leave the only undefeated I-AA team out of the playoffs this year.

My computer puts Texas A&M at the #3 toughest schedule. That is big for them, especially considering that they play Texas next. That SOS will remain high.

Texas A&M lost 3 games. Two of them were to undefeated teams that A&M should have lost. The oddest thing to many people is that a team isn't harmed much at all if they lose to a team that is ranked ahead of them. This is true with just about any computer program. What this means though, is a loss to Texas won't harm Texas A&M much unless the loss is big. So don't expect computers to drop A&M much next week.

So basically that puts A&M with one loss that is important. That one game they definately should have won. But they didn't. However in a computer program that is one bad performance averaged in with 9 good performances. The averaging math just makes one game have little impact. I've considered weighting some games more than others, but I just haven't had the time the last 2 years to implement things like that. I guess I'll have to post the unfinished code here at the end of the year and let ATOT readers do the improvements.

As for I-AA, I don't pay any attention at all. I don't know how the teams are chosen.

 

[ElFenix]

after examining the bcs i've determined that the only way texas can get into a bcs bowl is by beating the crap out of a&m with cal barely beating southern miss. this wouldn't send cal falling down in the polls far enough for boise state to jump into a bcs bowl, but it might be enough for texas to squeek past cal.

 

[ElFenix]

Originally posted by: dullard
The oddest thing to many people is that a team isn't harmed much at all if they lose to a team that is ranked ahead of them.

the oddest thing to me is how the wire polls send a team dropping down the ranks when they lose a close game to a team ranked 3 or 4 places ahead, when that should be confirmation that the poll is right. meh.

 

[Nitemare]

Still don't get why you have Auburn so low and Arizona State so high

 

[Alkaline5]

Originally posted by: dullard
So basically that puts A&M with one loss that is important. That one game they definately should have won. But they didn't. However in a computer program that is one bad performance averaged in with 9 good performances. The averaging math just makes one game have little impact. I've considered weighting some games more than others, but I just haven't had the time the last 2 years to implement things like that. I guess I'll have to post the unfinished code here at the end of the year and let ATOT readers do the improvements.

I'm sure that wouldn't be necessary and didn't mean for my comments to sound like a criticism of your rankings (if that's how you took it). I've been working on my own (flawed) rating program this season and just find it interesting to discuss the idiosyncrasies inherent to such things.

 

[dullard]

Originally posted by: Alkaline5
I'm sure that wouldn't be necessary and didn't mean for my comments to sound like a criticism of your rankings (if that's how you took it). I've been working on my own (flawed) rating program this season and just find it interesting to discuss the idiosyncrasies inherent to such things.

Criticism is good, comments are good. Without those, I'd get a lot fewer ideas. We'd love to see your rating program's results whenever you get it to a point that you like it.

 

[dullard]

Originally posted by: Nitemare
Still don't get why you have Auburn so low and Arizona State so high

They are tied. I could just as easilly swap them around when I typed that up. No program is good down to the 0.1 point prediction. Arizona St. has by far the toughest schedule. And they did quite well with that schedule. And because of that, they deserve to be at least at the head of the 2 loss teams. If we blindly just put all 0 loss and 1 loss teams ahead without regard for schedule difficulty or performance in those games, then Arizona St drops 3 spots to #9 (Auburn, Boise St, and Louisville move up). I just don't see the overrated part here. Their only losses were on the road at Cal and on the road at USC (both teams which deserve to be in major BCS bowls).

The Arizona game coming soon can only harm Arizona St. Since my program limits the number of points you can win by, Arizona St. cannot win by enough points to move up. Of course, teams around Arizona St could lose, making Arizona St. go up by default. Texas for example could lose big to Texas A&M, which would probably move Arizona St. up one spot.

 

[wyvrn]

Originally posted by: dullard

Originally posted by: wyvrn
After the schedule Auburn has had to play, why in the hell are they ranked as low as 7? That's just insanely stupid. I don't have any affiliation with Auburn, but I think they have proven they are as good as anyone this year.

Why is Notre Dame so high? They are my favorite college team, but I don't think they deserve to be 26th in the nation with the stupid losses they have had. I know their schedule is tough, but they are almost a .500 ballclub.

Strength of schedule is the biggest factor in my ratings. Here is how mine and many other computers work:
[*]Even case: If (1) your average opponent rating is at 50 points, (2) you win 6 games and lose 6 games, and (3) the average score difference in all your wins and all your losses is zero points, then you are rated at 50 points. I don't think anyone would disagree with this judgement.

[*]SOS case: If (1) your average opponent rating is at 55 points, (2) you win 6 games and lose 6 games, and (3) the average score difference in all your games is zero points, then how should you be rated? Obviously the looks of things are the same as in the even case above - zero point margin on average and same number of wins and losses. However, you played better opponents. Thus you definately should be rated higher. And using the same logic as in the even case, you deserve to be rated at 55 points.

[*]Win case: If (1) your average opponent rating is at 50 points, (2) you win 7 games and lose 5 games, and (3) the average score difference in all your games is zero points, then how should you be rated? Obviously there is a difference between this case and the case above. While this team on average played the same, they happened to pull out a win in a case that the team above didn't. Thus this team should be rated higher than 50. How much higher is the key.

[*]Score case: If (1) your average opponent rating is at 50 points, (2) you win 6 games and lose 6 games, and (3) the average score difference in all your games is +3 points on your side, then how should you be rated? On average, this team scored more points and did a bit better than the teams above. Thus this team should be rated higher than the even team. Thus this team should be rated higher than 50. How much higher is the key. And whether or not this better scoring justifies rating higher than the team with one more win is debatable.

So in 2 of the 4 cases, the rating is easy to see. SOS is the major factor in a computer rating. How you weight the bonus for an extra win and how you rate the bonus for more points score (or conversely, less points allowed) will vary from computer program to computer program. And in fact, I give two different weightings above.

Despite what you think, Auburn has played a fairly weak SOS. Division I-AA Citadel (a crappy I-AA team by the way), bottom of the I-A barrel UL-Monroe and Kentucky, and two additional teams with only 3 wins. All together, my computer says Aurburn's SOS is at #68. That means 67 teams played tougher schedules. That is devistating in any computer program. Compare that to #8 SOS for USC, #17 SOS for Oklahoma, #54 SOS for Utah, and #82 SOS for Boise St. Yes, Utah is given a tougher schedule than Auburn in computer programs. Playing Tennessee again will signficantly help here and a win by more than 12 points may very well put Auburn into the #5 spot (if all other teams play as expected).

Notre Dame has played a fairly tough schedule (#21 in my program). I don't think they are rated too much higher than they should be. Parity is finally setting in one the middle of college football. Top rated teams and bottom rated teams haven't shown parity issues yet (as seen by undefeated teams and an unwinning team). But the middle of the pack is huge this year with nearly all teams with about a .500 record. Someone with a nearly 0.500 record must sit on top of that mushy middle section. Here are the teams Notre Dame is ranked significantly ahead of that have fewer losses: Fresno St, Navy, Memphis, UAB, Bowling Green, Miami OH, and Toledo. Which of these teams would you rate above Notre Dame? And Why?

UTEP and Boston College.

I understand that Auburn playing some weak teams this year has hurt them. But they have still played a killer conference schedule and if there was a playoff system, there would zero doubt in my mind they were one of the final 2 teams standing. Or at least they would be as effective as OK and USC (meaning if they played in the runner up game, they would beat the #4 team handily). I think there is a dropoff after Auburn, OK, and USC to the rest of the division.

 

[sciencewhiz]

When you compare Dullard's rankings with other computers that factor in MOV (specifically Sagarin's predictor and Massey's power), his results are close.

Sagarin has USC and Cal within a point of each other, then approximately a 6 point drop to Oklahoma and Utah, then a 4 point drop to the next group which contains Auburn (who is rated 8th). Massey has USC, Cal, Oklahoma, and Utah within 2 points of each other, and then a 3 point drop to Auburn who is 5th. In other words, the dropoff occurs before Auburn, not after.

As far as SOS goes, Sagarin has USC 9, Oklahoma 18, Utah 62, Auburn 74, and Boise State 96. Massey has USC 4, Oklahoma 9, Utah 49, Auburn 58, and Boise St 84.

If you liked the SOS component from the BCS ratings last year, you have Oklahoma 11, USC 26, Auburn 42, Boise St 67, Utah 71,

 

[dullard]

Originally posted by: wyvrn
UTEP and Boston College.

Those two are tied with Notre Dame in my program and aren't on the list I gave you.

WTF is wrong with the wink icon?

 

[Alkaline5]

Originally posted by: dullard
Criticism is good, comments are good. Without those, I'd get a lot fewer ideas. We'd love to see your rating program's results whenever you get it to a point that you like it.

I've had to make adjustments for variations in # of games played, but the core calculations of my rankings haven't changed since about week 7. I decided to base it solely on win/loss record for two reasons: 1) input simplicity and 2) the generally poor quality of existing record-only rankings compared to existing rankings that include MoV. My implementation has it's drawbacks, of course, but for the most part I usually like it better than at least a couple of the BCS' computer rankings.

 

[AgentEL]

my rankings

Arizona State ahead of Cal? boo.

 

[Alkaline5]

Originally posted by: AgentEL
Arizona State ahead of Cal? boo.

Yeah, that just happened this week. The trouble with them is that they've lost to two extremely good teams and also really lucked out by beating two ranked OoC opponents. (Cal has only beaten one ranked team, btw.) Hopefully playing Arizona will hurt their schedule enough to drop them a spot next week.

 

[kalster]

Originally posted by: Alkaline5

Originally posted by: dullard
Criticism is good, comments are good. Without those, I'd get a lot fewer ideas. We'd love to see your rating program's results whenever you get it to a point that you like it.

I've had to make adjustments for variations in # of games played, but the core calculations of my rankings haven't changed since about week 7. I decided to base it solely on win/loss record for two reasons: 1) input simplicity and 2) the generally poor quality of existing record-only rankings compared to existing rankings that include MoV. My implementation has it's drawbacks, of course, but for the most part I usually like it better than at least a couple of the BCS' computer rankings.

shouldnt the fact that sc beat 2 out of the top 6 teams in your ranking, make it higher than Auburn atleast, OK has beaten a top 6 team too (texas)

 

[dullard]

Originally posted by: Alkaline5
I decided to base it solely on win/loss record for two reasons: 1) input simplicity and 2) the generally poor quality of existing record-only rankings compared to existing rankings that include MoV. My implementation has it's drawbacks, of course, but for the most part I usually like it better than at least a couple of the BCS' computer rankings.

Thanks for posting that. I haven't looked at mine with MoV ignored in years. Maybe I'll compare mine this holiday weekend with yours and post if anything interesting appears.

As for input, how do you do it? I used to type them in by hand, now I've gotten it down mostly to one copy and one paste. Of course there is the occasional exception when the website I copy from messed up the score (such as xx for the score instead of a number). If you are doing it by hand, I can give you my input code (nothing fancy, just brute force).

 

[Alkaline5]

Originally posted by: kalster
shouldnt the fact that sc beat 2 out of the top 6 teams in your ranking, make it higher than Auburn atleast, OK has beaten a top 6 team too (texas)

Auburn is helped somewhat by the fact that my formula ignores I-AA teams unless you lose to one. I haven't been able to reliably rate inter-divisional teams against each other due to a lack of cross-reference games so I just ignore the game if it's a win. My rankings work on an accumulation basis (rather than an average) so in it's 10 I-A games Auburn has accumulated a slight lead despite USC's victories over Az St and Cal. USC's upcoming games provide a good opportunity for them to pass Auburn, however, and the fact that they play 12 I-A opponents in the regular season should help.

 

[Tom]

Ohio State is 30th, where were they before they POUNDED Michigan ? 🙂

 

[Alkaline5]

Originally posted by: dullard
As for input, how do you do it? I used to type them in by hand, now I've gotten it down mostly to one copy and one paste. Of course there is the occasional exception when the website I copy from messed up the score (such as xx for the score instead of a number). If you are doing it by hand, I can give you my input code (nothing fancy, just brute force).

We must use the same source because that "xx" has screwed up things more than once for me (usually in the late-running Hawaii games). Thanks for the offer. My program has access to specific scores, but I just ignore them for now (all game results are read into an array straight from a text file). I'm still surprised that I've been able to get recognizable results at all with my formula and I'm sort of afraid that trying to add in MoV calculations will be more trouble than I can handle.