Sunday, October 25, 2009

The Guru's BlogPoll Ballot (Week 8)

The Guru's Week 8 BlogPoll ballot, with more in-depth explanations below:

RankTeamDelta
1 Alabama
2 Florida
3 TCU 8
4 Texas 1
5 Boise State 1
6 Oregon
7 USC 2
8 LSU
9 Iowa
10 Cincinnati
11 Georgia Tech 1
12 Penn State 1
13 Utah 2
14 Houston 4
15 Oklahoma State 4
16 Pittsburgh 4
17 West Virginia
18 Virginia Tech 2
19 Miami (Florida) 12
20 California
21 Ohio State
22 Arizona
23 Central Michigan 1
24 Notre Dame
25 South Carolina

Last week's ballot

Dropped Out: Brigham Young (#14), Texas Tech (#22), South Florida (#23), Kansas (#25).


* First a disclaimer, since some of the new readers seem a bit confused about the difference between BCS Standings Projections and the BlogPoll ballot: The projections are my estimations on what the upcoming BCS Standings will look like, based on projected computer ratings and an educated forecast of the two polls. It's NOT an opinion. The BlogPoll ballot, which is new for this season, is my own personal vote and is COMPLETELY my opinion.

* Alabama is still No. 1 on my ballot, the close game with Tennessee notwithstanding. It amuses me that somehow Florida's performance this week was viewed more favorably than Alabama's, causing voters to flop their picks. For the second week in a row, the Gators needed favorable calls by SEC's suspect officiating crews to salt away a game against a heavy underdog. The Tide had a late-game mini-meltdown, but they held on without anybody's help.

* TCU is No. 3. To be honest, I haven't seen the Frogs all that much this season, but the BYU game opened my eyes. The way they dominated the Cougars, in Provo, no less, was impressive. That, coupled with Miami's overtime loss to Clemson - a team TCU beat on the road - made the case for the Frogs to jump all the way to No. 3 (even if we're not in Calaveras County).

* Boise State stays ahead of Oregon, that's a slam dunk. The Broncos' schedule may not be awe-inspiring, but they still have the second-most impressive victory of the season among the top five teams (behind only Florida's win at LSU).

* Oregon jumps USC this week after yet another throttling of a Pac-10 foe, this time a rout of Washington in Seattle - a feat that eluded the Trojans. The Ducks have been easily the most impressive team after conference play began, outscoring their opponents, 161-38. They also have two wins over top 25 teams to their credit (Utah and Cal).

* Somehow, the fact that USC is among the top teams in any poll seems somewhat controversial to some people. Why? Surely this is not a vintage Trojans team, with a sometimes-sputtering offense and a suddenly mortal defense, but they have earned their ranking by merit. They have beaten three top 25 teams (Ohio State, Notre Dame, Cal), all on the road, in three of the most hostile environments. That's more than anybody in the top 25 has done (including Oregon, whose two best wins were both at home). A victory over the Ducks next week would make it four.

* Iowa may be 8-0, but I'm not quite willing to take the Hawkeyes seriously yet. Their entire resume is built on the upset win over Penn State in Happy Valley, but I have my doubts about just how good the Nittany Lions really are, too.

* Cal is back in the top 25, its only losses were to teams ranked sixth and seventh (even if the Bears were blown out in both games). Arizona also deservedly makes into the ballot, with its only losses to Iowa and Washington (caused by a bizarre play late in the game). The Wildcats handed Central Michigan its only loss of the season, while the Chippewas (No. 23) won at East Lansing in the same fashion that Iowa did.

19 comments:

Dustin said...

The BCS Guru likes to predict the future with is upcoming BCS standings, so let me make a little prediction of my own.

One Month from now, this will be seen on the BCS Guru's BlogPoll Ballot: Iowa #9
"Yes Iowa is 12-0, but I'm still not convinced. I mean they have only beat Penn St. Wisconsin, Mich. St., Ohio St. on the road, and Arizona, Michican at home, Lets see if they can have a tougher schedule and be undefeated for a few years in a row, then I may move them up a few spots"

Dustin said...

Iowa's opponents combined record: 38-22. Only one Iowa opponent currently has a losing record (arkansas st.). No other team can claim anything close to that.

LAprGuy said...

One month from now, this will be seen on the BCS Guru's BlogPoll Ballot: Boise State #2
"Yes, their schedule isn't awe-inspiring, but they did beat Oregon at home, in a game where both teams looked like crap."

Anonymous said...

The Iowa situation is really interesting this year. On one hand, you have BCS Guru (and most human polls) saying that Iowa is roughly 8 or 9, whereas five of the six BCS computer models have Iowa at #1! That's a big difference.

I think this is because the computer models are really measures of how much each team has accomplished so far, in hopes that past accomplishments are a good predictor of future performance. Iowa has accomplished a lot: as Dustin notes, only one of its opponents now has a losing record (compare to, say, Florida: four of their seven opponents have losing records). But the models by design can't take all information (i.e., margin of victory) into their measure of accomplishment, so they are blind to Iowa's "ugly" wins. Do ugly wins mean that the victor is not very good? I have no idea. At the least, however, I am skeptical of BCS Guru's argument that Iowa's resume is thin -- the computer models belie that argument. I suspect instead that style points are playing a huge role here, for better or for worse. The same goes for the AP, Harris, and USA Today polls.

Jonathan said...

"The Tide had a late-game mini-meltdown, but they held on without anybody's help."

Ignoring, of course, the no-call on Cody pulling his helmet off in the middle of a play, one of the few cut-and-dry calls that can be made in the fuzzy world of SEC unsportsmanlike conduct rulings.

I realize that the Tide took the game fair and square for the most part, but no one seems to be pointing out that little issue.

Anonymous said...

@Jonathan

Chris Dufresne of the LA Times also noted this. Although, even if they had called it, I'm sure refs would have given it an "after the play" flag ... can't let the meal ticket lose.

Reid said...

I'm not too sure about Oregon being ahead of USC(disclaimer: usc fan posting). Oregon doesn't have an impressive road win this season. I also don't believe it was surprising that Oregon St had such a good game against USC, because they have been giving USC a run for their money for quite a bit now.

I'm also not too sure about TCU or BYU right now I'm afraid. After BYU's blowout loss to Florida St, I can't say I am too impressed by BYU for this season.

Anonymous said...

Horned Frog Nation loves the BcSGuru.

Riff Ram!

mike said...

The whole point of the computers is to give some kind of objective value to a system where comparing records simply can't be done. I don't care how much you want to debate one team's schedule versus another, the fact is we don't know what is more impressive. For almost every team that USC plays that game is the biggest on their schedule. Nobody can argue that emotion is not an important part of sports.

Iowa has the advantage over USC in the computer polls. But people who watch both teams cannot possibly believe that Iowa would beat USC more often than not. The point of the BCS is to pit the two best teams against each other. Not the two teams with the best records. I wish it was more like the top 4 teams to compete but it is what it is. Just because a Cinderella team makes it undefeated through the whole year doesn't mean I want to see them playing for the championship. I want the best game and I think without question USC could provide that better than Iowa.

mike said...

Sam,

I'm not sure how you can in the same poll give TCU a huge amount of credit for beating BYU, while dropping BYU completely out of the rankings. How impressive can it be that TCU beat an unranked WAC team? How much should BYU be punished for losing to the #3 team in the nation?

Victor said...

Unfortunately, the actual purpose of the BCS is something that is only loosely defined, and voters get very little in the way of guidelines. The most basic way to sum up the problem is to think about who is the "strongest" team vs who is the "most accomplished" team.

Both concepts require wins, but the former rewards recruiting, athletic ability and coaching, and is more forgiving of upset losses. However, it's also heavily biased towards traditionally strong teams, as forgiving an upset loss also means denying an upset win. Voting for "most accomplished" rewards easier (aka "championship") schedules, gritty wins, and most importantly, consistency. It penalizes losses heavily, but also leaves a chance for smaller, underdog programs to compete for bigger money.

The distinction is subtle, but these two sets of criteria don't mix well at all. Unfortunately, voters mix them every week - they measure some teams against one standard, and other teams against the second standard. The result is the inconsistent rankings we see. Even the computers have these biases built in, based on what they are and aren't allowed to calculate.

The result is that concepts like strength of schedule, consistency, margin of victory, etc. are applied only when convenient. As an example from the Guru's poll, USC is ahead of LSU, Iowa and Cincy based on "strength" (despite Cincy faring better against a common opponent and LSU's only loss coming to Florida). Yet Boise and TCU are ahead of all the above based on "accomplishment."

If you're using mixed standards, you're effectively biasing the polls to include teams that you want to see play in BCS games. That's certainly a valid way to go, if that's the stated goal, but let's not pretend it's anything than a ranking of favourite teams.

Ute Fan said...

Wow. It's beat up on the guru day. I think your rankings are well thought out and you are to be lauded for your openness.

That said, it just goes to show how subjective any rankings are. The only way to tell if one team will beat another is to have them play it out on the field. "Would beat" means nothing. And even "more often then not", Mike. Who cares if USC could beat Iowa more often then not, we all know that. Can they beat them now? That's what matters and only a bout on the field could tell us.

It is interesting to see MWC (not WAC) TCU get more respect than a BCS conference Iowa. I agree that TCU deserves the respect, can't comment on Iowa as I haven't seen them play.

Donny said...

How in the world do you drop Texas after an impressive 41-7 road win versus Missouri? At the same time you do NOT DROP Alabama who would have lost if Tennessee's kicker wouldn't have missed 3 fg's on the game. Does. Not. Compute.

Texas has the #1 scoring offense and the #2 overall defense in the country... And yet you dropped them to #4 after a dominating win. Please reconsider... At least Texas can score offensive touch down's. Something Bama has done twice in their last 3 games.

LAprGuy said...

@Victor - Great post!

I don't even mind that there's bias out there -- makes it interesting.

What troubles me with the system is that BCS commissioners herald the computers as being the bias-free input ... and yet the public isn't privliged with information on how each programmer assesses the initial rankings -- which are AMAZINGLY close to the preseason writers' polls! -- and, especially, how they calculate strength of schedule. Because without margin of victory, then SOS and W-Ls are the ONLY factors that each computer should be calculating -- and yet each computer has wide discrepancies in their team rankings.

Jonathan said...

@LAprGuy: I think only one or two of the BCS computer programs have initial rankings other than starting everyone equal. Of course, "equal" might mean different things to different people, and you're right, there should be more transparency to these algorithms. That's why I like the Colley Matrix so much (well, transparency PLUS it seems like a strong basis for a ranking system).

@Victor: Your final statement might be a bit harsh, but overall, nice critique of the inherent tension of CFB polling.

Bizarro Ocho said...

Looks like the Guru actually watches the games unlike many of the pollsters.

Every team in the Top 10 has flaws so each one can be picked apart by anyone with an agenda.

It is ashamed the BCS is a beauty contest controlled like the Olympics were by the old Iron Curtain judges.

Some of the reasons why Florida, ALA, and TX stay at the top are:

Iowa suffers from playing in a league where only 3 programs (Penn St, Ohio St, Iowa) appear to concentrate on defense. The Big 10's recent poor Bowl record is killing them.

Cincinnati hails from a league that is under ridicule as the worst of the Big Six and is not a program with a storied past. Just my personal opinion but I think their high school looking uniforms hurt them also.

TCU lives in the shadow of the giant alumni bases of Texas, Texas A&M, and Texas Tech. Having the Dallas Cowboys 15 miles away does not help either. Their recruiting classes don't get hyped because they recruit mostly fast high school runningbacks and move them to other positions. The Verses DirecTV fued is also keeping their exposure down.

Boise suffers from the ineptitude of their conference to develop another serious program outside of Fresno St. They virtualy have 7 gimmies a year.

mike said...

"Who cares if USC could beat Iowa more often then not, we all know that. Can they beat them now? That's what matters and only a bout on the field could tell us."

Well we call them "Rankings" so basically the whole idea is that you rank who you think is the best team. If we called it the "Human Standings" then all undefeated teams should be ahead of teams with a loss. So the voters should care whether they think USC is better than Iowa, because that's how you rank who is better. I think, ideally, the voters take what they see and think "would team A beat team B". This can be based on reputation, performance, whatever. If you want to argue that because Iowa hasn't lost a game, that's why you think they would beat USC, then rank them higher.

Anonymous said...

---> Jonathan... the non-call in the UA TN game was a good no call. Even if a flag was thrown, the penalty for removing a helmet on the field wouldn't have mattered because its enforced as a Dead Ball foul... ALABAMA recovered the live ball after the kick was blocked...

Jonathan said...

@ Anonymous: I found that out later and wasn't sure if I should bother correcting myself on here. My own ignorance gets to shine through a bit. You're right. It was a good no-call.

Google