AP Voters continue to try to get it right; continue to fail miserably
The new CFB polls came out on Sunday, and after a week fraught with upsets, they were a predictable mess. An even bigger mess than the last time I ranted about the polls. This is where we stand:
|AP||USA Today||Jeff Sagarin|
|#1||Florida (55)||Florida (58)||Florida|
|#2||Texas (1)||Texas (1)||Boise State|
|#3||Alabama (4)||Alabama||Virginia Tech|
|#5||Boise State||Boise State||Iowa|
|#6||Virginia Tech||Virginia Tech||USC|
|#9||Ohio State||Ohio State||Texas|
|#13||Iowa||Penn State||Ohio State|
|#15||Penn State||Houston||Miami (FL)|
|#17||Miami (FL)||Iowa||Georgia Tech|
I want to make it perfectly clear up front that I’m not saying the Sagarin Computer Rankings are gospel. Honestly, how many of us thinks that Iowa is a top-5 team, that Boise State has played the 35th-toughest schedule, that Virginia Tech belongs ahead of an Alabama team that crushed them in week one, that the Pac 10 deserves two top-ten teams, or that Arizona deserves to be ranked? These things have a tendency to average themselves out with time (so the first BCS Poll of the season won’t be totally screwy), but for now, it can serve merely as an interesting comparison to the other polls as one that is devoid of prejudice or bias.
The first thing that the human polls have to explain that the computers do not is how you justify having a team jump way up or fall way down in one week. Sometimes it’s obvious. For example, it makes sense that Miami would fall eight spots after getting dismantled in its first game against an opponent that doesn’t struggle to stay in the top-25 on a week-by-week basis.
Others are harder to explain. Here’s one for the AP voters – can you explain why you only gave Iowa 39 points in week 3 (putting them not only out of the top 25, but also behind six other unranked teams), but then felt that one game made them a top-15 team? One game made the difference between not-even-close-to-ranked and top-15? The logic of that escapes me.
You could argue that they were just making up for mistakes in past weeks. They were underrated, and they compensated this week, you might say. That’s a fair argument. But, then it has to go the same way for the teams you thought were good but actually are not. California played three games against weak teams and then got positively steamrolled at Oregon. They fell 18 spots (and appropriately so), but what justifies them being in the top 25, other than that they were once ranked high?
The team that exemplifies the benefit of early high ranking even more than that is Ole Miss. They are 2-1 and played poorly against South Carolina last week. Nothing logically justifies them even sniffing the top-25, and yet they are 21. They are the lowest-ranked team on Sagarin’s rankings (47th!) to be in the top 25. This is one where you must be compelled to admit Sagarin is much closer to the truth than the human polls, as no logic can justify the Rebels currently being ahead of teams that have played better and have better records.
You also might notice that USC, just one week removed from losing to an unranked team, is already back up to No. 7. That’s some kind of recovery. They moved up five spots just for beating Washington State at home. That’s much better than Oklahoma State who, two weeks removed from a loss to No. 12 Houston, is somehow sitting way down at No. 14. It’s good to have name recognition.
Speaking of name recognition, there is nobody in the country benefiting from it more right now than LSU. The Tigers have been thoroughly unimpressive, playing three out of four games very close against opponents that are not the toughest in the world, yet they come in at No. 4. I’m beginning to think voters feel morally obligated to have three SEC teams in the top 5. They are ahead of Virginia Tech who, even with one loss, has wins over Nebraska and Miami to its credit. That’s more than LSU can boast. They are also well ahead of another slew of undefeated teams (Cincinnati, TCU, Houston and Iowa) all of whom, I assure you, have played better to date than LSU. This one really kills me because they play at Georgia this week. That means that one of those teams are going to come away looking better this week than they really are because they’ll have defeated an opponent that is ranked far too high.
And you’d better believe the AP loves wins against overrated opponents. It’s what keeps Florida State and Georgia Tech hopping in and out of the rankings constantly. Georgia Tech just leapfrogged into the top 25 with a win over North Carolina, who’d been quite unimpressive at No. 22. For some reason, teams such as these keep making brief cameos in the top 25, while Missouri (which is undefeated and has played more consistently than others) has now been overlooked three weeks in a row. They appeared in the top 25 in week two, then fell out after winning and haven’t been able to crack the rankings again. They haven’t played a great schedule, granted, but the bigger problem is that they don’t have wins against overrated opponents to their credit. The pollsters should realize that consistency deserves consideration over teams that keep coming out with one flashy showing (Florida State, Georgia Tech, Iowa, et. al.)
Missouri is, in fact, the only underrated Big 12 team right now. Every other one is ranked higher than the unbiased computers have them, especially Oklahoma State (29th in Sagarin’s rankings) and Kansas (35th).
The AP voters continue to get their weekly polls very wrong. They continue to reward teams based on name recognition or based on beating other teams with name recognition that they expect to be good, but are not. It is time to start rewarding teams for what actually matters – on the field performance. We can only hope that they, and their somewhat less egregious counterparts in the USA Today/Coaches Poll, get their heads screwed on better by the time the BCS Polls come out.