Opportunity forgoes that knocking nonsense and makes itself at home

by Joel on Oct 14, 2007 9:45 PM EDT in Tennessee Volunteer Football Comment 4 comments

So the first BCS standings are out, and the Vols are 21st. Not bad, considering. Here are the top ten:

		Harris Poll			USA Today			Computer Rankings								BCS
	TEAM	RK	PTS	%	RK	PTS	%	A&H	RB	CM	KM	JS	PW	%	COMP AVG	BCS AVG	PRVS
1	Ohio State	1	2845	.9982	1	1495	.9967	23	25	22	19	19	19	.830	T-5	.9416	NR
2	South Florida	3	2508	.8800	3	1320	.8800	25	24	25	25	25	25	1.000	1	.9200	NR
3	Boston College	2	2650	.9298	2	1383	.9220	20	21	19	23	21	20	.820	7	.8906	NR
4	LSU	5	2303	.8081	5	1173	.7820	22	22	23	24	24	24	.930	2	.8400	NR
5	Oklahoma	4	2503	.8782	4	1288	.8587	12	14	15	12	14	15	.550	11	.7623	NR
6	South Carolina	6	2009	.7049	8	997	.6647	21	20	20	22	23	23	.860	3	.7432	NR
7	Kentucky	11	1759	.6172	13	874	.5827	18	23	21	21	22	21	.850	4	.6833	NR
8	Arizona State	12	1697	.5954	12	936	.6240	24	17	24	20	17	22	.830	T-5	.6831	NR
9	West Virginia	8	1926	.6758	7	1007	.6713	10	19	11	17	18	18	.640	10	.6624	NR
10	Oregon	7	1974	.6926	6	1077	.7180	17	8	18	14	8	11	.500	13	.6369	NR

So what do you see? I see one, two, possibly three future opponents of the Volunteers, currently ranked Nos. seven, six, and four. That's opportunity, folks. We'll need to continue to get better, especially on defense, if we want to make the most of it, but the table is set.

My question to you, what if, if, Tennessee beats South Carolina and Kentucky, wins the SEC East, and beats LSU (assuming they win the West) in the SEC Championship game? Where would that put us?

0 recs | Comment 4 comments

Story-email Email Printer Print

Comments

Winning the SEC championship,

even with winning out, would put us in whichever BCS bowl game gets the SEC champ. (Orange? I forget.) There are still too many zero- or one-loss teams with a reasonable chance of ending with zero or one losses to expect a championship berth. Even if we did see a year where everybody had at least 2 losses, I'm not sure the pollsters would ever put Tennessee over teams that have been in the top 10 for a while, no matter how impressive the wins are.

If Tennessee actually made the championship game, it would be the most bizarre year of the BCS era, bar none. Additionally, if Tennessee were to actually win, we'd hear no end of how the championship meant less because everybody stunk it up, or some variant of that. Not that I'd mind having to defend a championship, but it'd get old fast.

In short, winning out through Athens gets us the BCS bowl slot.

by Hooper on Oct 15, 2007 9:18 AM EDT reply actions 0 recs

Here's what I see....

Objective computer data ranks three SEC teams LSU, South Carolina, and Kentucky as the #2, #3, and #4 teams, respectively, in the country. Yet the subjective human polls' average ranking for these same teams is #5, #7, and #12.

Meanwhile, the human polls rank on average Ohio State @ #1 and Boston College @ #2 with commensurate #5 and #7 computer rankings.

To carry this further, Tennessee is T-17 in the computer rankings, yet the human polls average ranking for Tennessee is #21.

Auburn is #12 in CPU rank; #19 in human polls.

I get steamed every time I see these type of results for the SEC, and expecially for Tennessee.

Does this slight to the SEC bother anyone other than me?

by volinal on Oct 15, 2007 12:26 PM EDT reply actions 0 recs

Human v. Computer

It's not necessarily much of a slight. The SEC has no undefeated teams, and human voters are quite likely to place "undefeated" over "one-loss beast". The same applies, to a lesser degree, to one-loss versus two-loss teams. In the human polls, you can really see the effect at the top, where undefeated takes 1-3.

The other effect in the human polls is that lower-ranked teams generally have to wait for higher-ranked teams to lose before they can move up, while computers have no such inertia. Given that Auburn racked up early losses, they (like Tennessee) have a harder road to climb the polls.

These are not factors that are SEC-specific other than that the top SEC teams tend to face more teams with realistic chances of beating them than most top teams in other conferences face. Taking Tennessee as an example, they face Cal, Florida, Alabama, South Carolina, Auburn, Arkansas, Kentucky, and possibly LSU before bowl season even begins. The only non-SEC team who routinely faces such a schedule is Notre Dame (which is perhaps the only praiseworthy note they've had all season). With a schedule like that, Tennessee was much more likely to have an early loss or two than teams like USC or Ohio State.

The CPU polls have also been known to produce screwy results from time to time. Since they are based only on statistically-measurable quantities, there are occasional teams that run outside the scope of their expected data ranges (outliers, if you will). That's how you get computers kicking USC out of a championship game that both human polls would have had them play. There are biases to be found in the human polls, but there are biases in the CPU polls as well. Fans from other conferences would look at those same numbers and declare that the CPU polls are biased in favor of the SEC and against their favorite conferences! the truth most likely lies somewhere in the middle.

Besides, teams evolve (and devolve) significantly during a season. Tennessee has found a workable defense in the last few weeks, for example. At this point in the season, it really matters very little how things look in the polls. After about week 10, then things really begin to matter. Human poll inertia can be moved this far out, but is very difficult to overcome (other than losses) in the last two or three weeks.

by Hooper on Oct 15, 2007 10:03 PM EDT up reply actions 0 recs

A Lot of Football Left to Play

So many variables, if this team wins out the regular season and wins in Atlanta, who knows...it would be highly improbable, but not impossible, to end up in the Sugar Bowl.

Tennessee, and many SEC schools, not Georgia and LSU - their media darlings - are hindered by a persistent and consistent bias in the coaches and harris polls...less the AP historically, but hey they don't count.

I must disagree with hooper though, it's pretty consisttent week-end week out, those teams that will be lower in the human than the computer, and those teams that will be haigher in the computer than the human.

While statistical outliers do come into play, especially during the season when teams haven't necessarily played the same number of games, or the same number of home games, these outliers tend to work themselves out in the long range. I don't believe the human prejudices do.

Also, to say that pollsters are less likely to move someone up because teams in front of them haven't lost is while accurate, kinda the whole point of the criticism - pollsters should vote on who is the best team, based on work that season, not simply reward teams for not losing to easy opponent and refuse to move a 6-2 team over a 5-1 team on just that criteria.

by General Neylands Legacy on Oct 17, 2007 3:17 PM EDT reply actions 0 recs