Early college football rankings are great hype material and provide good discussion points, but are they worth the consequences?
Computers can't see everything that's happening on the field, so there is a valid place for human opinion in the process that determines the best teams in the country. However, computers are completely unbiased, so they also have earned their seat at the table.
Ultimately, there are issues with the human polls, and most of the issues occur before the season ever kicks off. Here are the reasons that college football rankings should come out no sooner than Week 4.
Enjoy, and feel free to add some reasons of your own (or completely disagree) in the comments.
When preseason polls come out, there are many factors involved in the selection process:
- Teams' success during the previous year.
- Players lost to the NFL or graduation during the offseason.
- Head coach's success.
- Projected player talent for the replacements coming up on the roster.
There is one major point that cannot be considered before the first game kicks off: actual football.
While it is great to see something like No. 3 Oregon vs. No. 4 LSU appear on the opening-day schedule, did the rankings make the game any bigger? Of course not. It was the most anticipated opening game in recent memory, and the rankings had little to do with it.
Everybody knows what teams are supposed to be good each season. For instance, nobody needs a preseason list to tell them that Ohio State, Oregon and the winner of the Texas A&M vs. Alabama game are going to be teams to watch for the national championship.
What do early polls tell fans? Nothing they didn't already know.
The early polls create bias that is carried throughout the football season. For instance, Boise State has historically been highly successful on the big stage, but should the Broncos have been ranked No. 18 at the end of 2012?
When a team is put into the preseason poll, there is a subconscious tendency to bring it back as soon as it demonstrates the slightest ability to win a game.
Boise State faced one ranked team last season, and that was Michigan State. Michigan State was only ranked in the early polls. As of Week 4 of the 2012 season, Michigan State was already down at No. 21, even with the win over preseason-No. 24 Boise State.
Boise State was a good team with a great defense, but its most significant win last season was by two points over the unranked Washington Huskies.
Was Boise State really the No. 18 team last season, or does the nation think that even the unranked Baylor Bears could have beaten the Broncos?
The bias against teams exists, too. A lot of teams are underrated at the beginning of the season until they demonstrate the ability to sustain success. In a perfect world, last year's results would have nothing to do with the current season's.
There are a few games that matter during opening week. In 2012, the Michigan State vs. Boise State, South Carolina vs. Vanderbilt and Michigan vs. Alabama games were the three big ones.
Others mattered but were not quite as highly anticipated: Northwestern vs. Syracuse and Notre Dame vs. Navy (in Ireland) are two good examples.
However, all of those games, with the exception of South Carolina vs. Vanderbilt, were non-conference games that didn't truly tell anybody anything about what the teams would do throughout the season.
Plus, Week 1 is simply the first of 12 to 14 games for every team. The Week 1 version of any given team can be infinitely different from the team you see on rivalry weekend.
This is because most teams don't start conference play until Week 3 or 4. Even if there's a conference matchup to open the season, the teams still play only the allotted eight or nine conference games, so the difference gets made up later in the season.
The first few weeks are largely FBS vs. FCS tune-up games, like Kent State vs. Towson. Basically, nothing happens until conference play starts in earnest, which is about a month after the preseason polls are released.
Rocking the boat used to be frowned upon, but rocking the boat has produced a lot of forward movement in college football.
The BCS system, though despised greatly, was a step forward from the previous situation. The only way a No. 1 vs. No. 2 postseason game happened prior to the BCS was if the teams happened to land in the same bowl against each other.
So, rocking the boat has worked, but not enough. The boat needs to be rocked harder. Enter the Week 4 polls.
If the polls came out in Week 4, then the voters would have a great idea of which 25 teams are actually the best. Even if they don't yet have a great idea, they'd have a better idea than before any games are played.
Instead of having four polls released that tell fans which teams the voters think will be the best, which the fans basically already know, the first poll will come out and reflect who has actually performed on the field.
In general, if there is one improvement to be made in college football, it's that on-field results need to be given the highest priority.
That was the theoretical improvement that the BCS provided in the championship game, and it's also the improvement that's coming with the new playoff system in 2014.
The same progress needs to be made in the polling systems, because the human eye will always be a necessary part of the selection process. What the gurus need to do is find ways to eliminate the human error that is inherent in that aspect of the formula.
With the absence of conference play in the dawn of each season, the results of the first few weeks of college football are mostly questionable.
With the voters trying to make themselves look more correct than they were about preseason polls, teams can slip into midseason rankings when they don't necessarily belong.
The hype associated with a match involving highly ranked teams will always have hype with or without the numbers beside the names, so the preseason rankings are basically useless for generating viewers.
The preseason polls don't have any actual information to offer fans beyond what teams they should expect to be good, which they already know.
Waiting to release the first poll until Week 4 or after Week 4's results would eliminate all the issues except inherent human bias that has nothing to do with preseason polls. There is no way to eliminate that particular bent, though.
Even if it didn't create a perfect poll, it could only improve the current situation. In fact, there may even be a case to follow the BCS's example (it did get one thing right) and wait for Week 8, when most teams have played at least a third of their conference slates.
The bottom line is that polls need to wait until Week 4 at a minimum. Anything before then is almost pure speculation as opposed to informed opinion. It would be a nice improvement to go along with the four-team playoff.