Blog Entry

Blogpoll Week 1

Posted on: September 9, 2009 12:18 pm
Edited on: September 9, 2009 12:20 pm
Rank Team PPB StdDev Delta
1 Florida (92) 24.5 2.7 --
2 Texas 23.2 2.7 --
3 Southern Cal (2) 22.4 2.1 --
4 Alabama (5) 22.3 1.4 1
5 Oklahoma State 20.0 2.0 5
6 Penn State 18.7 3.1 1
7 California 17.7 2.6 5
8 Mississippi 16.9 3.1 3
9 Brigham Young (1) 15.5 4.0 11
10 Ohio State 14.8 4.3 4
11 LSU 14.7 4.4 3
12 Boise State 14.5 4.1 3
13 Oklahoma 11.8 4.1 9
14 Georgia Tech 11.5 4.4 --
15 Virginia Tech 10.2 4.6 6
16 TCU 8.1 4.1 1
17 Miami (Florida) 6.5 5.3 8
18 Notre Dame 6.2 5.2 6
19 Utah (1) 5.7 5.0 1
20 Georgia 5.6 4.0 7
21 North Carolina 5.3 4.1 --
22 Cincinnati 4.7 4.7 4
23 Nebraska 4.6 3.4 --
24 Missouri 3.8 5.1 2
25 Kansas 2.9 3.1 1

Also Receiving Votes: Oregon State(2.1), Michigan State(1.7), Florida State(1.6), Clemson(1.3), Pittsburgh(0.8), Baylor(0.7), Oregon(0.7), Texas Tech(0.6), Iowa(0.5), Arkansas(0.3), South Carolina(0.3), Michigan(0.3), Tennessee(0.3), South Florida(0.2), Colorado State(0.2), UCLA(0.2), Arizona(0.1), Auburn(0.1), Southern Miss(0.1), West Virginia(0.1), Bowling Green(0.1), Northwestern(0.1), Miami (Ohio)(0.1), East Carolina(0.0), Stanford(0.0), Texas A&M(0.0), Navy(0.0), Air Force(0.0), North Carolina State(0.0), Houston(0.0), Wisconsin(0.0), Kentucky(0.0), UAB(0.0), Tulsa(0.0),

Total Ballots: 101

Votes by blog here , votes by team here .

Ah, welcome to the week one Blogpoll, when no one can decide just what the hell they're supposed to be doing. Some people make adjustments to their preseason ballot. Some people throw it away entirely . And some people regard their ballot as an exercise in epistemology by ranking the entirety of D-I based on nothing more than first-week victory margin, thus throwing your poll editor into a tailspin of philosophical agony over spiking the thing. It's a stunt ballot, but two other ballots are in striking distance in Mr. Bold. It maintains internal consistency. It's dumb. It's modern art, or something. I bet this never happens in the basketball poll.

Anyway… yes, spiked. It's kind of smug to make some sort of point about how no one really knows anything after week one, isn't it? It's doubly so when the poll doesn't require anyone to submit a ballot until after week three. Maybe you're taking the polling thing a little too seriously if you can't just wait, Che. No stunt ballots. (DocSat ballots and other pure resume ballots are not stunt ballots, they're deeply neurotic ballots.)

The poll itself features a couple of relatively large AP gaps this week, possibly because of all the nuts:

  • Bloggers love: Cal (7th instead of 10th), Miami (17th instead of 20th).
  • Bloggers hate: many teams very slightly: PSU, OSU, Mississippi, etc.

Wack Ballot Watchdog

I get moving your first place ballot to someone else after Florida took on the fightin' Chesneys in their opener, but… Utah? That's the sage decision of The Lion's Den for reasons unknown. I'm sure the BYU vote is some resume guy or another. This feature will get ramped up more over the next couple weeks as there's more hard evidence to dispute fancy notions. A quick scan of Oregon's votes shows no one with Boise State under them, so that's good… and that's about the only thing we can say at this point since the BYU game was a nail-biter in which OU played without that Heisman guy.

Now on to the extracurriculars. First up are the teams which spur the most and least disagreement between voters as measured by standard deviation. Note that the standard deviation charts halt at #25 when looking for the lowest, otherwise teams that everyone agreed were terrible (say, Eastern Michigan) would all be at the top.

Most Disagreement

# Team StdDev
1 Miami (Florida) 5.3
2 Notre Dame 5.2
3 Missouri 5.1
4 Utah 5.0
5 Cincinnati 4.7

Mid-majors are surprisingly almost excised from the most controversial teams in the poll, and Miami takes the crown for good reason. Every year after the FSU-Miami game we have to figure out whether the winner is legit or not; recently the answer has been "not."

Ballot Math

First up are "Mr. Bold" and "Mr. Numb Existence." The former goes to the voter with the ballot most divergent from the poll at large. The number you see is the average difference between a person's opinion of a team and the poll's opinion.


Mr. Bold

# Blog Team   Avg Error
1 Dr. Saturday Southern Miss 8.95
2 SimonOnSports Connecticut 7.09
3 Bleed Scarlet Rutgers 4.71
4 Dan Shanoff Florida 4.50
5 The Rakes of Mallow Notre Dame 4.23

The aforementioned Che ballot is this week's nominal Mr. Bold winner, but amongst the non-spiked ballots it's now-traditional week one victor Doctor Saturday, who is one of those guys—the foremost guy, in fact—who discards the preseason ballot wholly for stuff that happened in week one. This results in oddities like BYU #2, Boise #4, and Texas and Florida (DocSat's preseason 1 and 2) unranked.

I still disagree with this sort of balloting but I've given up trying to change anyone's minds.

Mr. Numb Existence

# Blog Team   Avg Error
1 SOX & Dawgs Connecticut 1.34
2 Block C Clemson 1.40
3 Block U Utah 1.49
4 NittanyWhiteOut Penn State 1.56
5 The Only Colors Michigan State 1.61

New UConn blog Sox & Dawgs is your Mr. Numb Existence winner. This early no one gets an exact ballot , but Sox and Dawgs is pretty close this week. Also note that having the word "Block" in your blog name turns you into a lifeless shell incapable of feeling anything.

Next we have the Coulter/Kos Award and the Straight Bangin' Award , which are again different sides of the same coin. The CKA and SBA go to the blogs with the highest and lowest bias rating, respectively. Bias rating is calculated by subtracting the blogger's vote for his own team from the poll-wide average. A high number indicates you are shameless homer. A low number indicates that you suffer from an abusive relationship with your football team.

The Coulter/Kos Award

# Blog Team   Bias
1 Subway Domer Notre Dame 8.78
2 RazorBloggers Arkansas 7.65
3 Mountain West Connection TCU 5.86
4 Anton Azucar Miami (Florida) 5.48
5 Double Extra Point Nebraska 5.36

All right .

The CK Award continued its reign of terror by condemning by-the-numbers week one winner Tomahawk Nation , which shot Florida State way high in a preseason ballot that was based solely on schedule strength, to suffer humiliating defeat at the hands of an arch-rival. (Last week's blogpoll actually skipped over that in favor of ND blog Rakes of Mallow , which actually expressed opinions about teams in his poll; the CK Award has let it be known that the numbers are not to be defied.) By the numbers, the award is 1-0 against the spread and straight up.

This week your winner is Subway Domer, which places Notre Dame #11, almost nine points higher than poll consensus. Notre Dame plays Michigan this week; they're favored by about three. Mwa ha ha.

The Straight Bangin' Award

# Blog Team   Bias
1 Tilting at WindMills Oklahoma -11.83
2 Dawg Sports Georgia -5.62
3 Hey Jenny Slater Georgia -5.62
4 CollegeGameBalls Virginia Tech -4.20
5 And The Valley Shook! LSU -3.67

Wow. One of the blogpoll's finest traditions is downvoting your team spectacularly after a disappointing loss, and indeed the top four in this category are all bloggers supporting teams that lost disappointingly in week one action. And though the fifth didn't lose they did get Life on the Margins-ed. New Oklahoma blog Tilting at Windmills  is just going to, like, die it's life is totally over . Accordingly, it ranks the Sooners… well… not at all.

Swing is the total change in each ballot from last week to this week (obviously voters who didn't submit a ballot last week are not included). A high number means you are easily distracted by shiny things. A low number means that you're damn sure you're right no matter what reality says.

Mr. Manic-Depressive

# Blog Team   Swing
1 Dr. Saturday Southern Miss 418
2 SimonOnSports Connecticut 310
3 The Rakes of Mallow Notre Dame 222
4 One Bronco Nation Under God Boise State 204
5 mgoblog Michigan 200

Ah, first-week swing, the province of dedicated resume-rankers like now-traditional week one Mr. Manic Depressive winner Doctor Saturday. And, uh… MGoBlog , your poll host. Don't look at me, I don't vote anymore.

Mr. Stubborn

# Blog Team   Swing
1 The Bull Gator Florida 48
2 The Daily Gopher Minnesota 54
3 Fantasy College Blitz Oklahoma State 60
4 Orange::44 Syracuse 60
5 SpartyMSU Michigan State 64

On the flipside of things, these are the preseason poll believers that only lightly dabbed at their near-perfect ballot after a week of chaos. The Bull Gator was exceptionally generous to Oklahoma, dropping them only five slots and actually keeping BYU eight slots behind their opening-week victim. Similarly, Oregon did not get summarily dismissed after an opening game that went as disastrously as possible—Boise, at least, is in front of them—and Virginia Tech got a light tickling for their Alabama loss, dropping only two spots.

Category: NCAAF
Tags: blogpoll

Since: Sep 9, 2009
Posted on: September 11, 2009 12:54 pm

Blogpoll Week 1

Based only on the scores, its true that OU might be terrible, and that BYU is only very slightly less terrible.  Whereas Iowa and UNI might be the two best teams in the country.  But is there really any other reason to believe either of those statments?  We have tons of other information to go on.  We know that OU looked pretty darn good last year and we know which of their players returned for this year.  Do we have a reason to suspect that their returning running backs got significantly worse?  We know something about the kind of recruits they have.  We know what BYU looked like last year, and we know that they return a pretty decent QB, a 1000 yard receiver and a 1000 yard rusher.  Likewise, we know that UNI wasn't terribly good last year.  Do we have reason to suspect UNI got significantly better over the summer?

Finally, to look only at margin of victory belies one of our greatest abilities: the ablility to use a complex advanced brain to look at a complex question.  The MOV arguement breaks the entire sport down to a ranking system that a 4 year-old could both devise and master.  Instead, we have the opportunity to bring so much more analytic capability to the table.

I'm a big fan of resume rankings.  I wish everyone could forget about what conference someone plays in, or how many championships they won in the 1930s.  But just as jobs aren't given out on bare numbers on a resume alone, neither should poll rankings.

But hey, if anyone wants to get into a specific discussion of whether or not Air Force is the best team in the nation, lets go for it.  But ask yourself, which side do you think will come out on top?

Since: Feb 27, 2007
Posted on: September 10, 2009 2:58 am

Blogpoll Week 1

In the case of the ballot that rated everyone only on margin of victory, the rator is ignoring the opponent.
What do we know about said opponent besides "lost by a gazillion to so-and-so"? (If you're going to do the whole resume-ranking thing after week 1, there's really nothing else to go on because everyone with a win beat someone who's 0-1.) Yes, we think Oklahoma is probably better than most of the teams that got clobbered, but all we really know about them at this point is "lost to BYU". After all, Alabama vaulted up the polls last year after annihilating Clemson ... which turned out to be impressive but nowhere near worthy of the "holy @#$%, Alabama must be the BEST TEAM EVAR!!!" reaction that it got at the time.

I wouldn't vote that way if it were me, but I can see the logic in it. It's taking the resume thing a bit farther than most of the other resume voters (who do make assumptions about opponent strength which are, as of yet, not justified by results).

Since: Sep 9, 2009
Posted on: September 9, 2009 6:49 pm

For Spiking Ballots

Spiking ballots is acceptable any time someone is obviously or openly ignoring data that the vast majority of people agree is both useful and pertinent in judging a team.  The primary judging criteria have to be: win/loss record, margin of victory and strength of schedule.  Other factors can certainly be included, such as yardage totals, game sites, and injuries. In the case of the ballot that rated everyone only on margin of victory, the rator is ignoring the opponent.  BYU pulled out a very close game over OU.  The 1 point margin of victory that was still in doubt until the closing minutes makes it one of the closest games of the weekend.  Using his system, BYU must be one of the very worst of the 1-0 teams in the nation.  My point is, he is blatantly ignoring useful information.  There must be some limit to how much an entry can deviate, otherwise there is nothing to stop someone from rating teams entirely on field goal accuracy, or passes defended.  It is not too much to ask that a ballot be reasonable, and that it include those three major points, because the basic questions in rating a team ought to be "Did you win? Who did you beat? and By how much?"

Since: Sep 9, 2009
Posted on: September 9, 2009 3:29 pm

Against Spiking Ballots

Your decision to spike ballots shows a limited tolerance towards rigorous debate and ideological diversity. After one week of play, there is limited data available. It should come as no surprise that bloggers interpret that data in non-conformist ways. Spiking ballots for running away with the "Mr. Bold" award says, in essence, too much non-conformity will not be tolerated.

You have enough submissions that you don't need to be concerned about a few outliers to your blogpoll. If a blogger is truly randomly selecting which team to vote, then sure, I see why you would spike the ballot. Short of that, you shouldn't be spiking ballots because you are opposed to the methodology the voter used.

The views expressed in this blog are solely those of the author and do not reflect the views of CBS Sports or