Tuesday, August 12, 2008

The Poll

Wide receivers are not The Poll, but here you go anyway. First, a piece on former walk-on receiver Cary Koch. Koch, as you probably know, is our resident Katrina refugee: specifically, he is a transfer from Tulane, which dropped his major following the hurricane. I like this article because it speaks to how walk-ons are treated in our program: very well. Koch, Staton Jobe, and Hall Simmons have earned themselves schollies this year after walking on. Not under-talented, just under-evaluated; those are Groh's own words, and I think the program benefits greatly from that attitude.

Speaking of benefitting greatly, here's some fluff on another guy we can really use this year: Kevin Ogletree. Since arriving on Grounds as a wide-eyed first-year, I can honestly say I have not been treated to spectacular wide receiver play from the football team. Billy McMullen is the only one that us Wahoos could really get excited over. Now the return of Ogletree to the team has most fans thrilled that the offense may once again involve the ball traveling more than five yards through the air. Not that Groh has ever been a big fan of the old hey-why-not fly pattern on first and 10, but Ogletree brings a position back to the offense that nobody really noticed last year.

OK. The Poll. First, a word on methodology. Guidance has been handed down from On High regarding how the dungeonmaster would like us to vote, and it's located here. There is only one ironclad rule, which boils down to: No funny stuff. Meaning, no look-at-me voting, no designer polls to prove a point. So, even though this is the only Virginia blog participating, that means you won't see me voting UVA #1 just because, nor will I even be including UVA just because.

(That thing about this being the only Virginia blog in the poll? Pretty significant, by the way, as it's essentially the primary reason your humble blogger even has a seat at the adult table.)

So, how is this poll going to shake out over the season? The rules that this pollster will follow are listed. Consider them bylaws, almost.

1. The first three polls of the season are the only "predictive" polls. There isn't a body of work that's big enough to judge the teams on until they've all played about three games. The first couple weeks of the season is when I'll allow (mis)conceptions to have a much greater role in shaping my judgment than what has actually happened on the field.

2. The Week 3 poll will shift almost exclusively to "resume" voting. For reference, this year, Saturday #3 is September 13. And yes, this means teams that had a bye week in the first couple weeks will be somewhat penalized. Doesn't matter - if they're any good, they'll catch up. Why Week 3? I point to last season at Michigan. After the third game (in which they demolished Notre Dame after two galactically embarrassing losses) was when people started to realize that, OK, the Wolverines are not going to be that bad, but yes, the Irish really are. After three games, I don't care if you're Florida International, if you're 3-0, you're probably one of the top 25 teams in the country. (Actually, especially FIU, as it means they'll have knocked off Kansas, Iowa, and USF.)

After three games is also a perfect time to shift gears to resume voting, because that is, for most teams, the time of transition from the non-conference slate to the conference schedule.

3. All undefeated teams that have won at least 7 games will be ranked higher than any team with a loss. This goes directly against the cautionary Hawaii tale explicitly pointed out in the philosophy, but the cautionary Hawaii tale goes against my notion that undefeatedness deserves to be rewarded. If you've played a majority of your schedule and not lost, that is worth something. The truth is, Hawaii aside, most teams that "don't deserve" a ranking that high will prove that by season's end. College football's like that. Somebody's gunning for you. In fact, they all are, if you draw attention to yourself by refusing to lose. So staying on that perch gets harder and harder.

Yes, this may potentially result in a precipitous rise in some WAC or ConfUSA team's ranking. You may see somebody go from #17 to #6 in one week because of this rule. Again though: if they don't deserve it, they'll show it. More specifically, someone will expose it.

4. After Week 8, no team that is leading its conference or division will be placed lower than a team in the same conference that is not. Why Week 8? Due to byes, most teams will not play 7 games in 7 weeks. Week 8 is when most teams will have played A) 7 games and B) 4 conference games, which in most conferences is half the schedule. Saturday #8 this year is October 18.

Again, this is resume voting by this time, and not predictive. I don't really care if a team has the monster half of its schedule coming up and is likely to drop. That's not the point. The poll is a snapshot of what we're looking at now, not some Frankensteinish hybrid of what we see now plus what we think we'll see in the future. I don't use tiebreakers to determine who's on top. If two or more teams are tied, as is likely that early, then I get to just pick which one I think has done better; however, chances are pretty good that if one has a better nonconference record than the other, or has, you know, actually beaten the other, then that team will be ranked higher.

What this doesn't mean is that all conference leaders will line up for the top spots. Conference leaders will be placed where they belong, not #5 because they lead the fifth-best conference. But no team from their same conference will be above them.

5. If two teams have equal records and have played each other, the winning team is ranked higher. Should go without saying, really, but there may be instances where strength of schedule could dictate otherwise. In this case strength of schedule will be ignored.

6. Likewise, two teams with the same record, but not having played each other, will be ranked according to how they've fared against common opponents. This seems logical, no?

That's it. Sometimes these rules may come in conflict with one another. They're ranked roughly in order of precedence, though. In most cases, they won't apply, and wins and losses combined with strength of schedule will be the primary factors.

To make this preseason list, I did not apply these rules to the 2007 season, in case you're wondering. Too much graduation, too much of people getting kicked off teams - they're no longer valid after a full offseason. What methodology got applied here, that is simple. I didn't Phil Steele-ize this stuff, agonizingly and painstakingly weigh each variable, compare starting lineups, and pore through stats. Instead, I looked at the various rankings available, weighed them against each other a little bit, and basically just decided for myself whether too much (or not enough) Kool-Aid has been drunk regarding particular teams. And a little bit of a bump upwards in some cases for being the best team in a conference. Maybe Texas is better than Clemson - they probably are - but leading a conference is the difference between watching the selection show to find out who your opponent will be and watching it to find out if you're in at all.

The Poll, then. The very first FOV poll. It's what I will submit on my own behalf unless there's convincing howling that I've gooned this up.

1. Ohio State
2. USC
3. Georgia
4. Florida
5. Oklahoma
6. Clemson
7. Missouri
8. Texas
9. Arizona State
10. LSU
11. West Virginia
12. Wisconsin
13. Kansas
14. Virginia Tech
15. BYU
16. Texas Tech
17. Auburn
18. Tennessee
19. Wake Forest
20. Oregon
21. Illinois
22. Fresno State
23. South Florida
24. Michigan
25. North Carolina

Yes, North Carolina at #25. It even struck me as odd - "what am I doing?" thought I. Then I said, you know what? Having finished their season preview, and having done some of the work on the rest of the ACC, even those that haven't posted, I couldn't convince myself that UNC would lose to those teams I left off the list.

#26, for those interested, would be a fairly interesting looking Cincinnati team.

No comments: