For five weeks we've been updating our Pre-Season rankings based on each team's weekly performance, a method that falls somewhere between a Bayesian and ELO approach. We use this approach because it takes some time before an unbiased system—like our Strength power rating—starts to produce good results. After 5 weeks, the Strength power rating is starting to come together, but might still have some shortcomings. The updated pre-season ratings are nearing the end of their usefulness, and next week (week 6) we start to combine the two before transitioning to the unbiased Strength ratings in week 7. To show where a team's ranking might be heading, we include the Strength ranking in the table of this week's update.
Who is #1? Baylor has been at #1 for 3 weeks now; FSU started at #1, held it during week 1, then ceded the spot to Oregon in week 2. The Bears beat Iowa State 49-28 which was good enough to hold off a huge charge by UCLA, who jumps from #10 to #2 after beating Arizona State 62-27. In Strength, Baylor is #3 but UCLA is only #11, suggesting they're getting too much of a boost from our pre-season #3 ranking. The Bruins disappointed in each of their first three games before ASU, and even in that game the defense was questionable, so we'd guess #11 is more accurate than #2.
The teams from #3 to #6 all underperformed and fell, though they all won, allowing Oregon to move up to #4 without playing and Mississippi to move up 3 spots even though their 24-3 win over Memphis met expectations almost exactly. SEC teams Texas A&M, Auburn, and Georgia failed to impress but all remain in the top 10 both in these rankings and in Strength, and Auburn is #1 there.
Florida State is a different matter. They've dropped in rating every update, this week falling to #6, and in Strength—which carries no pre-season bias—they are a pedestrian #27. This begs the question, how good (or not) are the Seminoles? It's pretty certain that they're not playing like a #1 team, but are they really outside the top 25? Note that one of their games—the overtime win over Clemson—was played without Jameis Winston. In Strength they're getting hit for their 37-12 win over The Citadel, a result that an average FBS team should achieve. In this case, the updated rankings are probably a lot closer to the truth than Strength, which needs a few more weeks to be accurate with some teams.
Michigan State is another interesting case. We ranked them #15 in the pre-season and in each update they've hovered near that spot, this week moving up to #12. In the unbiased Strength ratings they're not far off at #17. But our pre-season rankings for their offense and defense were #36 and #4, and they're known as a defense-oriented team. While their offense has been much improved this year, the Strength ratings (for all Divisions) put the Spartans at #3 on offense—and only #67 on defense! These numbers don't match most observers' realities. The Strength ratings often take some time before things "fall into place" and this is one of those cases. Giving up 46 points to Oregon—in one of only 4 games sampled so far—has the computer thinking Michigan State has an "average" defense. That should correct in the next few weeks. Perhaps MSU's defense isn't quite as good as most people think, but it's got to be better than #67.
Pac-12 teams USC and Stanford (#11 and #18) could be mis-rated, too, as they are only #25 and #30 in Strength. Stanford's game next week against Notre Dame should help answer that question. On the other hand Wisconsin's ranking (#24) nearly matches their Strength ranking (#22) after the Badgers fall 12 spots due to their weak 27-10 win over South Florida. Right now the rankings are off by an average of 9 places per team, with a median of 7 places and a mode of 2.
SportsRatings pre-season rankings updated for week 5
|73||72||North Carolina St||4-1||+1.42||55|
|74||74||San Diego St||2-2||+0.03||94|
|86||92||San Jose St||1-3||-0.65||90|
|127||127||New Mexico St||2-3||-0.74||120|
The team with the biggest ranking discrepancy is Houston, who is #37 or #75 depending on which rating you look at! The Cougars started at #16 in our pre-season rankings and the weekly adjustments haven't captured their (lack of) quality nearly enough. Indiana is another case study in overrating; they were #18 pre-season, but they've fallen to #47, which our Strength rating says is still 14 spots too high. The Hoosiers clearly have the potential to play very well—hence the upset of Missouri—but they are too inconsistent to be a good team week to week. Oregon State is #43 and 3-1, so in this case it might be their Strength rating (#78) that's off. In general though, the Pac-12 doesn't rank very well in Strength and might not be meeting our pre-season expectations.
Georgia Southern and Louisiana Tech are two teams we didn't expect much from this year, and both have exceeded those expectations nearly every week. The Eagles started at #121 and moved to #116, #104, #91, and now #81—but the Strength power rating has them at #54! Louisiana Tech has moved from #114 to #85. Strength has the Bulldogs at #48 despite their earlier loss to Northwestern State.