Nonconference wins for each P5 conference

Submitted by mfan_in_ohio on November 5th, 2018 at 3:39 PM

The current AP and Coaches polls have 6 SEC teams (Alabama, Georgia, LSU, Kentucky, Mississippi State, and Florida) in the top 25, with both Auburn and Texas A&M receiving votes, despite that conference having an absurdly weak nonconference schedule.  Meanwhile, the "down" ACC has the same number of ranked teams (4) as the B1G.  Now that we are 75% of the way through the year, it's a good time to look back at the nonconference results and see which conferences deserve the level of respect they get.  So here are all of the nonconference wins by each conference over any team that is ranked in one of the polls: 

Big Ten (5):

Texas (Maryland), Utah State (MSU), Boston College (Purdue), Fresno State (Minnesota), Iowa State (Iowa)

Not a bad list, although the Big Ten might lead the nation in embarrassing losses too, as there are division contenders with losses to Akron and Eastern Michigan.  Also there is Rutger.

ACC (0): 

None, as Clemson's win over Texas A&M no longer applies.  Syracuse could add one with a win over Notre Dame though.  Fingers crossed.

Big 12 (2):

Boise State (Oklahoma St.), NC State (West Virginia)

Considering that it plays a 9 game round robin schedule, the Big 12 as a group plays a relatively decent nonconference schedule that also included wins over teams such as USC and Tennessee.  Also that once-embarrassing Oklahoma win over Army in OT might make this list soon.

PAC 12: (1)

Michigan State (Fightin' Herms)

The MSU win is a lone bright spot, as the PAC 12 lost in almost all of its opportunities against good teams.  UCLA played three opponents that are currently ranked, and unfortunately UCLA sucks.  Notre Dame beat Stanford and will likely beat USC, who also lost to Texas.  BYU is also a popular opponent that is not up to its usual strength.  At least Colorado beat Nebraska by 5, so they are equally as good as Ohio State. 

(Ok, the OSU game was in Columbus, so that makes Colorado better.) 

SEC (1):

Washington (Auburn).  

Pretty embarrassing, especially given that they have 4 nonconference games per team.  Granted, a bunch of those games haven't been played yet, but the only ranked nonconference opponent left on anyone's schedule is Clemson vs. South Carolina.  Good luck with that.  



It's kind of an amazing result: the Big Ten has more wins over ranked nonconference opponents than the other four P5 conferences combined.  While the polls don't really mean anything anymore, the Big Ten's big lead in this department also existed last week for the CFP rankings; Utah State wasn't in their top 25, but Virginia (lost at Indiana) was.  Caveat: Of course, it's hard to have too many wins over ranked opponents, as losses tend to make those opponents unranked.  

Unfortunately, there isn't a lot of looking backward by voters or committee members, as LSU is still getting the benefit of what was thought to be a "big win" over Miami, just as Virginia Tech was still ranked after losing to Old Dominion because their win over Florida State was overvalued for weeks.  Meanwhile, CFP voters look at things like wins over P5 teams with winning records, which is ridiculous when many of those teams are playing teams like Furman, Liberty, and the Citadel to get those winning records.  For as much grief as Washington State gets for scheduling home games against Eastern Washington and San Jose St., and playing @Wyoming, the 9th conference game makes their schedule at least as tough as any in the SEC, where virtually every team has three cupcakes.  At least Wazzu played a road game.

It's hard to look at this and think that teams should ever schedule difficult nonconference opponents.  In fact, conferences have an incentive to game the system (much like basketball with RPI) and work together to all schedule mediocre-to-bad teams, and run up the nonconference records, just to have more teams in the conference with winning records.  It's unfortunate, because those are some of the most entertaining games, but I fear that that's where we may be headed.



November 5th, 2018 at 4:08 PM ^

So with the WVU-NC State game having been cancelled (I assumed it had and that WVU won), that's 5 top 25 wins for the B1G to 3 total for the rest of the power 5. 


November 5th, 2018 at 5:21 PM ^

I think this would tell the story better to list the overall record against P5 along with any important wins or losses.  That would flush out wins against teams that are just outside the Top-25 or penalize people for losing to EMU.


November 5th, 2018 at 8:05 PM ^

The point of the diary is to show that voters, whether it is the traditional polls or the CFP committee, are valuing very highly the teams in the SEC despite the fact that their conference has very few wins of note outside the conference.  As a result, the preseason polls that valued teams like Miami very highly are still impacting polls now.  LSU received big bumps from early wins over Miami and Auburn, and are still ranked in the top 5 by some AP voters because of a perceived strength of schedule.  

The problem with going by W/L against other P5 conferences is that it ignores games against legitimately good G5 schools but counts Rutger like it's an actual college football team.  I wasn't really looking at the overall strength of each conference (the B1G as a whole is certainly pulled down significantly by the likes of Rutger and Illinois). I don't really care that Kansas beat Rutger because both teams suck anyway.  Similarly, I don't think that record against .500 or better P5 teams is a good measure because that metric is meaningless when you can go 6-6 and have 3 of those wins against teams like Tennessee Tech and Alabama State.   So I chose to look at which conferences had actually recorded wins against the top 25.  I thought about expanding it to teams that had received votes, but the results were about the same, in that the B1G has by far the most good wins.  

Overall, what this says to me is that the SEC is overvalued, as was the ACC in the first CFP rankings.  Mississippi St. is at 18 (AP) and 15 (Coaches) as a 6-3 SEC team, but they scored 7 or fewer in each of their three losses, and four of their 6 wins are against Stephen F. Austin, La. Tech, La.-Lafayette, and Kansas State.  Kentucky is at 12 after 4 straight games scoring 17 or fewer points and is coming off a pretty BS win at Mizzou (needed a questionable penalty to get an extra play) and got stomped by Georgia.  Also, teams like Sparty and Iowa should be higher. Hopefully they are both in the CFP top 25.  


November 6th, 2018 at 4:42 PM ^

Ok, and let's look at losses to teams not in the top 25.

  • Big Ten: ASU, Temple, Kansas, Buffalo, Duke, Akron, BYU, EMU, Missouri, USF, Colorado, Troy
  • SEC: Colorado State, North Texas (Both from Arkansas, apparently their Rutger)
  • Big 12: Maryland, Iowa, Duke, Ole Miss, Nicholls
  • Pac 12: Auburn, Nevada, BYU, Houston, SDSU
  • ACC: Purdue, Indiana, Old Dominion, USF, Cal, ECU

So, the Big Ten has twice as many bad losses as the next worst conference, the ACC, which is hovering in that 5-6 range that seems to be normal. The only SEC team to have a bad loss OOC is Arkansas.


November 7th, 2018 at 1:39 PM ^

Interesting that the Big Ten has more of both of these categories, despite playing less OOC games than the SEC and ACC (they both have 8 game conference schedules, right?) Though it makes sense that we would likely have more than the the Big 12 and the Pac 12 as we have more teams, and thus more games.


November 7th, 2018 at 2:21 PM ^

We're arbitrarily ranking teams and using an arbitrary cutoff to decide what is a good vs. bad team. We are only looking at two of four categories: wins vs. good teams and losses vs. bad teams. The other two are wins vs. bad teams and losses vs. good teams, which are basically expected outcomes by a borderline top 25 team. If unranked team A beats unranked team B, we are ignoring that game for team A's record in this analysis, but including it in team B's record. While the Big Ten has more unexpected wins and unexpected losses, the other conferences just consolidated expected wins and losses. None of this really makes sense as a useful piece of analysis, which is what I was trying to point out.


November 7th, 2018 at 7:12 PM ^

Yeah, I agree with this premise. You should add to that wins against solid non-ranked Power 5 teams. Like TCU (Ohio), Virginia (Indiana), Pitt (Penn State)