Very often in college football, especially early in the season, we see a difference in results between the human polls and the computer rankings. After all, they generally do two different things.
There are two different types of major computer rankings. The first judges what a team has accomplished. It looks at wins and losses and strength of schedule to create an overall picture of a team’s accomplishments. Those rankings are still somewhat useless (and often hilarious), because teams haven’t played enough games for a real SOS measurement to be accurate.
The second type are power ratings, often called “advanced metrics” or “analytic rankings.” These look at every drive, every play, that a team has made, and uses it to look forward and predict who is more likely to win future matchups. Who’s the better team on a neutral field? These rankings try to tell you.
Human polls, on the other hand, are some kind of subjective synthesis of the two. Voters take into account how good a team looks/seems, but generally use wins and losses to override that. At the end of the day, the record trumps the potential in human polls, as it should.
This year, so far, the advanced metrics computers like Ohio State a lot more than the human voters do. Just how wide is the difference? Let’s look at the nitty gritty.
Next… Voters are severely underrating Ohio State
Ohio State’s Computer Ratings
I’m not going to go through every major computer rating out there, but I’ll touch on some of the most well-known.
First, let’s start with Sagarin. Jeff Saragarin’s ratings have been widely-respected and quoted for decades. His are among the best at predicting future games, and his predictor ratings have Ohio State at the top, a whole five points better than No. 2 Alabama. All of his other ratings also have the Buckeyes at No. 1, though not quite by as wide a margin.
What about ESPN’s ratings? Buckeye fans generally don’t like ESPN and FPI–and I’ve written at length about FPI’s issues–but FPI is a pretty accurate predictive ranking. It currently has the Buckeyes at No. 2, just a hair behind Alabama. The same is true of Bill Connelly’s SP+ ratings.
Now, these metrics are still using a bit of their preseason expectations, but that doesn’t matter here. In fact, that makes the point even stronger. Advanced metrics didn’t like Ohio State at the start of the season. Even with that preseason bias against them, the Buckeyes have already risen to the top.
It is very clear from the advanced metrics that Ohio State is, at absolute lowest, in a class–along with Alabama–above the rest of the college football world so far this season. Some analysts have even begun to recognize this.
Still, far too many voters haven’t. Ohio State sits at No. 4 in the AP Poll and No. 5 (barely) in the Amway Coaches Poll.
What’s holding the Buckeyes down? Some voters explain their ballots. Most, however, do not. So we’re left making educated guesses as to what’s keeping Ohio State down.
The most obvious reason, and the one that is surely compelling most voters, is that the Buckeyes started lower on their ballots. Voters are still hesitant to move teams down until they lose. Alabama, Clemson, and Georgia haven’t lost any games yet, so they haven’t fallen behind Ohio State. With the crowded field up top (and throw in LSU, which picked up an assumed big win over Texas) with those three plus Oklahoma, Ohio State just has limited room to rise.
Also, there is likely still a cloud hanging over the Buckeyes. Whether consciously acknowledging it or not, every voter is aware that Ohio State has suffered a blowout loss in three consecutive seasons. Is that stopping voters from putting the Buckeyes up with the upper elites? On some level, likely yes. Should it? Of course not.
Neither the pollsters nor the computers had Ohio State among the top teams when the season started. The pollsters were going based off recent results, while the computers were unsure of Ryan Day and Justin Fields.
After five games, the computers have learned their lesson. When will the voters follow?
.png?w=600)
