Posts Tagged ‘Rankings’

In Defense of My Top Three

In Rankings Commentary on October 19, 2010 at 9:55 PM

I just posted the conferences & divisions chart for this week. It can also be found on the “ratings site” tab.

For background of where I’m coming from, please see the end of my last post and the comments here. I should not have checked those comments again before bed last night, but what’s done is done.

I remain convinced that if we had a 7-week season followed by bowls, LSU needs to playing Auburn for the national championship. If we had an 8-week season followed by bowls, the winner of LSU-Auburn would deserve a spot against the winner of Oklahoma-Missouri.

To be clear, my formula in no way (except for when home advantage comes into play) factors in margin of victory (MoV), how well a team was playing when a game was played, and the like. It matters who you beat. How good who you beat is depends on that team’s record and their strength of schedule (which is determined by opponents’ and opponents’ opponents records). I think it’s difficult enough to evaluate a group of 120 teams who typically play only 12 games apiece when you look at wins and losses alone to try to claim to be able to fairly consider the context of each game and how that led to the final score (which is the only reasonable way to use MoV)

The outcome I suggest wouldn’t happen if we left it up the pollsters, who would give us Oregon vs. Boise St. (Boise). The BCS might give us one of the two in the second scenario but neither team in the first scenario.

I’ve been pretty harsh on the ESPN crew, but there is one exchange I would like to highlight. Craig James said at one point that it matters “who you play and who you beat.” Kirk Herbstreit’s response was, “Why are Boise and TCU even in the conversation then?”

Craig was missing a qualifying clause: “When comparing undefeated teams….”

In that case, Kirk’s rhetorical point is well taken. They shouldn’t be!

But we consider them in case we’re comparing them down the line to, let’s say, an Ohio St. team who will have only had one impressive win (over Iowa) if they finish the season undefeated, the way things look right now. I’d be hard-pressed to say that Ohio St. team belongs ahead of an undefeated Boise St. team, especially if Virginia Tech continues to do well and Miami continues to struggle in the ACC. I don’t know how good Utah (awful schedule so far) and Air Force (losses to Oklahoma and San Diego St.) really are, but if they’re both undefeated otherwise from now until the end of the season, especially if we add in an improved Oregon St. team (which also might affect Boise St.), maybe they’d be better if they’re undefeated. So I’m not inflexible here, and I won’t say, “no non-AQs, no way, nohow.”

But they don’t just jump to the front of the line because of last year’s team. We’re halfway through the season, even factoring in bowl games and conference championships. Last year is a memory. It deserves no place in the rankings right now.

To get to the nuts and bolts, instead of using my system, which is not biased, but which will be attacked as such, I’ll use Anderson & Hester, which prides itself on rewarding regional dominance. So if anything, it’s an advantage to Oregon and Boise St.

These are Oregon’s opponents:
#12 Stanford
#67 Arizona St.
#75 Tennessee
#96 Washington St.
#109 New Mexico
I-AA/FCS Portland St.

Tennessee, incidentally, is a common opponent with LSU. But Tennessee is LSU’s fifth toughest opponent thus far. I know that’s the one that LSU came closest to losing to, but as I keep reminding people, Alabama beat a mediocre Tennessee team by 2 last year as well.

So Oregon has played one top-66 opponent. Alabama had three in the first five weeks, losing to the fourth. You could make the argument that maybe they shouldn’t even go ahead of Alabama at least until they’ve had two more wins than losses against such teams. My rankings aren’t that harsh toward undefeated teams, but I would have trouble saying that would be wrong.

I just don’t think having one win against a quality team, even if the margin was a big one, qualifies a team to be #1 after 7 weeks. Let’s say a scheduling happenstance has Wisconsin playing Michigan St. later in the year and they go into last week undefeated. You could certainly argue that Ohio St. was better than Stanford, so would the Badgers be a deserving #1 in that case? They also defeated ASU, Oregon’s second-best win right now.

That’s not to say it will be easy for Oregon to defeat Arizona or even some of the more inconsistent Pac-10 teams such as USC, Oregon St., Washington, UCLA, but why don’t we wait until they play some more of those instead of just one?

Boise can’t even look at their schedule in the future and say that. “Now, wait a minute, you’re not giving Oregon credit for it’s future schedule, why do you want to punish Boise.” I don’t, because I’m consistent, unlike the biased or misguided masses who have Oregon and Boise #1 and #2. For Boise, they ignore the future schedule, but it’s justification for Oregon.

Anyway, since I’m fair, unlike those people, let’s look only at Boise St.’s schedule so far.:
#36 Virginia Tech
#37 Oregon St.
#57 Toledo
#73 Wyoming
#100 San Jose St.
#109 New Mexico St.

The best team they beat is #36. That’s OK, especially considering they also beat another team right behind them. But since people want to act like James Rodgers was going to win the Heisman, and that’s why we shouldn’t consider Oregon St. without him, why don’t we look further into Virginia Tech? They started so badly they then went and played James Madison at home and lost (Boise St.-Va. Tech was a neutral crowd, although Boise had a significant traveling disadvantage). Also, let’s further consider the fact that Boise was lucky to escape with a win against Virginia Tech. I can understand putting them near Oregon though. Maybe it’s tougher to beat two top-40 teams that one top-20 team. Maybe it’s better to put someone tested at all against the top 20, on the other hand.

There is a third win just inside the top 60, enough for a decent top-25 resume, but how does that compare to the schedules SO FAR of Oklahoma, Auburn, and LSU?

#15 Florida St.
#17 Texas
#31 Air Force

Let’s stop right there. That’s already three opponents better than Boise St.’s best opponent. Iowa St. is also inside the top 50, ahead of Toledo. I don’t think either win is impressive, but even if we cancel out those, that’s still three opponents better than Boise St. best two opponents. And Florida St. and Texas aren’t even close. I don’t see any logical basis for not putting Oklahoma ahead of Boise St.

To go back to Oregon for a second, when we add in Cincinnati, that’s a total of 5 Oklahoma opponents better than Oregon’s second-best opponent.

#20 Mississippi St.
#21 North Carolina
#25 West Virginia
#26 Florida

That’s four opponents better than Oregon’s second-best and Boise’s best. Call me crazy, but I think 4 wins against #20 to #30, one at a neutral site, one on the road, is more impressive than 1 win against #10 to #20 at home. As mentioned, Tennessee is in the 70s in A&H, Vanderbilt is #82. Last week, LSU beat a I-AA team.

#20 Mississippi St.
#22 South Carolina
#28 Arkansas
#50 Kentucky

That’s four better than Oregon’s second-best or Boise’s third-best. Also three better than Boise’s best.

I could take either side of the Boise vs. Oregon debate, but either of them vs. LSU, Auburn, or Oklahoma, I’m sorry, I don’t see an argument there. Since we have opinion polls involved (I’d be happy if we didn’t, assuming it’s a transparent, relatively easy-to-follow rating system), I don’t have a problem with pollsters considering margin of victory if it’s a close call, but there should be a close call first, and there just isn’t.

There is also the strategy of attacking the results of teams that are beaten by the big three (in my view and that of many objective sources). Tennessee almost lost to UAB, for instance. Colorado beat Georgia, who easily beat Tennessee. I think this does more harm to Oregon since Tennessee is higher on their list, but just as an example.

So let’s look at Stanford then and see how #12 their worst result looks. Beat USC with a field goal at the end of the game, just like the Washington Huskies did the week before. Washington, by the way, lost to 2-5 BYU. USC at #27 in A&H is Stanford’s best win, by the way, followed by Notre Dame at #34. Auburn and Oklahoma each have three wins higher than that, and LSU has four.

I’m not saying Stanford is a bad team by any stretch, but the point is you can over-analyze pretty much any team, especially as the season continues, to make them look bad. Also, if that’s all you have to hang your hat on, you don’t deserve to be #1 in the country. It’s good, but it’s not enough.