theknightswhosay

Top 25 after Week 12

In College Football, General LSU, Rankings, Rankings Commentary on November 18, 2018 at 12:52 PM

After relying on my original computer formula for 100% of the list below last week, I just couldn’t do it this week.  I am switching #1 and#2.  Notre Dame has the better schedule for the moment by just a whisker (mostly because the Irish played Michigan),but Southern California (Notre Dame’s next opponent) will have a worse rating than South Carolina (Clemson’s next opponent) next week.  In addition, the Gamecocks (who are already bowl-eligible) will be able to pad their record with a win over Akron on December 1.  The Trojans, assuming they lose, will be done for the season (and ineligible for a bowl).

The Irish looked good, both in uniforms and in playing ability, against Syracuse and became my computer #1. Irish safety Alohi Gilman is pictured intercepting a pass
in Yankee Stadium on Saturday .

One other difference is the last four teams in the list below are the four teams(in order) that appear in the top 20 of my weighted ratings (which better mimic the CFP committee considerations by giving priority to how many of the best teams you play over your average opponent… to the extent the CFP committee is based on wins and losses and an objective evaluation of strength of schedule anyway), but did not appear in the top 20 of my older formulation.  Utah St. was #21 in both, so I thought it made sense to put those teams after the Aggies. If you were curious, the teams omitted as a result of this decision are Cincinnati, Buffalo, and Troy.  It’s three omitted teams instead of four because Missouri would be ranked either way.  The Tigers are just two spots higher this way. 

Syracuse, Auburn, and Northwestern are the three teams in the top 25 of the weighted ratings but not listed below.  Had I followed the weighted ratings exclusively, they would have replaced Texas, Fresno St., and Washington.

A perfectly average SEC team would be ranked #30, so that makes it much easier for the SEC teams to get the extra points awarded in the weighted ratings.  The Big Ten has the second-best average rating, but a perfectly average Big Ten team would only be ranked #48.

Auburn (Alabama’s next opponent) has a very similar rating to South Carolina, so I don’t think the Tide will be able to narrow the gap much if at all.  Alabama may pass up Notre Dame though, another reason not to make the Irish #1 right now.

I normally only use this top 25 blog entry to talk about why the ratings are what they are, but I wanted to make a couple of comments about Alabama.  It’s a coincidence that this demotion comes after their worst first half of the season, but I guess it’s fitting.  I thought LSU should have used the option against Alabama, and I think the Citadel’s performance in the first half supported that idea.  LSU won the 2011 “Game of the Century” by using more option than Alabama expected as well.  Nick Saban said probably none of the Citadel players could play for Alabama (maybe one or two could be decent walk-ons), and it still made them competitive for a while.  That’s not the case with LSU obviously.  Maybe it’s something we can work on in the offseason.  I don’t mean become one of those all-option all-the-time teams that almost never throw the ball (like Georgia Tech and the service academies), but we need ways to spread out defenses like Alabama’s horizontally in order to sustain drives and limit opponents’ possessions.

LSU fell two spots after playing possibly the worst team at the FBS level (they lost to the other candidate, UTEP, but the Owls have faced a stronger schedule).  If Michigan and LSU win next week, the Tigers should move back ahead of the Buckeyes. Washington St. may be ahead for good, although of course the Cougars could lose to Washington or Utah (which clinched the Pac-12 South).  Oklahoma and Georgia could also suffer losses in the next two weeks.  In my weighted ratings, LSU is still #6 behind Michigan. 

RankTeamPrev.
1Clemson2
2Notre Dame3
3Alabama1
4Georgia4
5Michigan5
6Oklahoma6
7Ohio St.8
8Wash St.9
9LSU7
10C. Florida12
11Kentucky18
12Boise St.13
13Texas16
14Penn St.11
15Florida14
16Utah21
17W. Virginia10
18Fresno St.
19Washington20
20Army17
21Utah St.22
22Texas A&M
23Missouri
24S Carolina
25Miss. St.

Out of Top 25: (15) Cincinnati, (19) Buffalo, (23) Duke, (24) UAB, (25) Iowa St.

Advertisements

Week 12: Not Rivalry Week Yet

In Bowls, College Football, General LSU, History, Me, Preview, Rivalry on November 16, 2018 at 7:42 PM

Apart from some remotely possibly upsets of top teams (I mentioned Clemson and Notre Dame in the Rankings blog), I’m not wildly excited about any of the matchups this week.   I still thought of somethings I’d like to talk about. 

The Former Rivalry Week

I miss the days where this was THE main rivalry week. 

The Big Ten used to finish up for good, but now they have 3 more weeks including the championship.  Tomorrow it will be exactly 11 years since Ohio St. beat Michigan, probably with no suspicion that they were about to be involved in the craziest ending to a college football season in recent memory.  Although the Buckeyes were ranked only #7 going into that final game, they would enter the bowls as the #1 team in the BCS standings.  Despite its second loss coming in the last regularly-scheduled game, LSU would become the surprise #2 after winning the SEC championship on the same day Numbers 1 and 2 in the BCS (Missouri and West Virginia) both lost. 

LSU LB Ali Highsmith gets to the ball before Ohio St. QB Todd Boeckman can throw it in LSU’s 38-24 championship win in New Orleans on January 7, 2008.

Anyway, I bring that up because the normal time of year for Ohio St. to play Michigan going back to the 1930s was between about November 17 and November 24.  2007 just happened to be the last time the game was on the 17th.  The end of the Big Ten season got pushed closer to the end of November in 2010; and then with the start of the Big Ten Championship game in 2011, the Big Ten season now extends into December.. 

Some Big Ten teams finished conference play even earlier.  For instance, in 2005, Wisconsin played its last Big Ten game on November 12.  There were 11 teams in the Big Ten then, so I guess the Badgers were the odd men out for the rivalry week.  Other end-of-season rivalries in the Big Ten were Minnesota-Iowa, Michigan St.-Penn St., Purdue-Indiana, and Illinois-Northwestern.

Althoughit was often played later (and only became the traditional final regular-seasongame in 1977), Florida played Florida St. on November 17 as recently as 2001.  2001 was also the last time UCLA played USC onNovember 17.  Sometimes there was a latergame for one or both schools, but it was the second-to-last Saturday inNovember going back to the 1970s.

17 Nov 2001: Kevin Arbet tackles Craig Bragg as USC upsets UCLA 27-0 to qualify for a bowl game in Pete Carroll’s first year with the Trojans.

Another big rivalry that used to be the second-to-last Saturday in November was Oklahoma-Nebraska.  It was permanently moved to the last Saturday in November in the early 1990s before it stopped being an annual game in 1998.  Of course Nebraska was a much more important team in those days than they are today.  The date would sometimes vary a week or so, but the rivalry had been played around that time of year since the 1940s.

The Iron Bowl was played between November 17 and November 23 every year from 1993 through 2006.  Those were the first 14 seasons in which I had a meaningful interest in college football on the national level, though I followed LSU for about 5 years before that. 

Anyway, so I think that’s enough explanation of why I always feel like something is missing this week, especially since it became the week for the SEC to take it easy. 

How the SEC Schedule for Mid-November Deteriorated

Although it had been done occasionally a few times before (for instance, South Carolina played Middle Tennessee the week before Clemson in 2006; and LSU played Conference-USA opponents before Arkansas a few times in the 1990s), Alabama led the way with a real commitment to this trend. 

Startingin 2007, the Tide has usually had a bye before the LSU game, so since theycouldn’t have another bye before Auburn, they played UL-Monroe.  The ended up losing to LSU, Auburn, and ULMin 2007; but that didn’t deter Alabama from that strategy.  In 2008, the Tide did the opposite (byebefore Auburn, non-conference game before LSU), and it worked.  Alabama only went a combined 3-3 against LSUand Auburn between 2009 and 2011, but they’re a combined 11-2 in regularly-scheduledgames against the two rivals since.

For itspart, LSU played Tulane the week before the Alabama game in 2008 and 2009,which did not work.  Then LSU went to thebye before Alabama (which worked for two years and hasn’t worked since), but theprecedent was already set.  Sometimes it’sin late October instead of November, but the Tigers have had a late-seasonnon-conference game most of the years since. They did not have one in 2016 only because of rescheduling that resultedfrom the hurricane that hit Florida. 

Auburn has been more consistent.  Except for 2013 when the Plains-Tigers were able to use a second bye before Alabama, Auburn has had a non-conference opponent the week before Alabama every year since 2011.

Georgia originally scheduled its late-season non-conference opponent before Auburn, but in 2014 the Bulldogs changed it to the week before Georgia Tech.  I’m not sure why it wasn’t done that way last year, but Georgia is back to that pattern this year. 

A few of the less significant SEC programs are still playing regular games, but the SEC schedule leaves a lot to be desired…

Ole Miss-Vanderbilt Headlines This Week’s SEC Schedule

Anyway, so we are now at the stage where the big SEC rivalry game this week is Ole Miss-Vanderbilt.  I’ll explain why.

Arkansas has played Mississippi St. annually since 1992, but the Bulldogs have won 5 of 6 in the series, and the Hogs are only 2-8 on the season.  Arkansas could back into a single-digit game like they did against LSU last week, but I hardly expect high drama.  So that’s not a game to watch. 

Missouri and Tennessee (the CBS game of the week) have slightly better combined records than Vanderbilt and Ole Miss, but that’s only been a rivalry (of sorts) since Missouri joined the SEC in 2012.  It hasn’t been a very interesting one either.  Missouri ended both 2015 and 2016 really badly and lost to the Vols in the process.  The Tigers won the other games.  The only game of the six decided by fewer than 8 points was in 2012 (when each team would finish 5-7).

Ole Miss and Vanderbilt, however, is a competitive longstanding rivalry between fairly evenly-matched teams. Since 2005, the only SEC team against which the Commodores have a winning record is Ole Miss (7-6).  Vanderbilt won 5 of 6 in the series from 2007 to 2012, but Ole Miss responded by winning the next 3.  The two programs have exchanged home wins over the past two years.  The Commodores have won 4 of the last 6 games played against the Rebels in Vanderbilt Stadium. 

The three touchdowns by Vanderbilt RB Ralph Webb (#7) were the difference in Nashville two years ago. The Commodores had ended a 3-game series winning streak by the Rebels.

As for this year’s respective teams, both are near .500 and have identical 1-5 conference records.  Nonetheless,Vanderbilt could still guarantee a bowl game by finishing the season with home wins over the Rebels and the Volunteers, their two biggest historical rivals.  The Rebels are still on probation and ineligible for a bowl, but I’m sure there is motivation to avoid a losing record and potentially finish with a winning record (which they could do by beating Vanderbilt and winning the Egg Bowl over Mississippi St.).

LSU and Rice Renew a Rivalry Few Missed

One other rivalry I’d like to mention is LSU-Rice.  It was before my time, but this used to be an annual series.  Other than in-state (former/sporadic) rival Tulane, LSU has played Rice more than any other team that is currently outside of the SEC. LSU and Rice played each other every year between 1932 and 1952 and every year but one between 1955 and 1983.  The only meetings between 1983 and this season were in 1987 and 1995.

Rice has only beaten the Tigers once since 1966.  However, despite LSU winning a national championship in 1958, it was a competitive series between 1955 and 1966.  Rice had a 5-4-2 record against LSU during that span. 

The most notable Rice win was in 1961.  The Owls denied the Tigers a chance at second national championship in four seasons.  After losing the opener to Rice 16-3, LSU would win the next 10 games including the Orange Bowl.  Rice would finish 7-4 and lose in the Bluebonnet Bowl, the Owls’ last bowl appearance until 2006.

Rice made 5 bowl games from 2006 to 2014, winning 3 of them, their only wins in bowl games since 1953 (they also lost the 1957 Cotton Bowl and the 1960 Sugar Bowl). 

The Owls have returned to their prior form since that 2014 bowl win though.  After falling just one win short of qualifying for a bowl for the fourth consecutive year in 2015 (with a 5-7 record), Rice has only won 5 games since the start of the 2016 season.  Two of those wins were over FCS opponent Prairie View A&M, including in the opener this year, which was Rice’s only victory in its last 21 contests.  Two of the other wins since 2016 were over UTEP, which finally ended a 20-game losing streak two weeks ago against Rice. The fifth win was over UNC-Charlotte, which only began playing in the FBS in the past few years.  

Top 25 after Week 11

In College Football, Rankings, Rankings Commentary on November 11, 2018 at 4:59 PM

Alabama did end up becoming #1 based on the computer data.  This is both in the formula that I’ve been using for about 10 years and in the “weighted ratings,” which I created last year to give higher ratings for playing very good opponents. 

For instance, I think Mississippi St. is one of the 25 toughest teams to beat, but they’re not in the top 25 below since that system averages every week together equally.    So when Mississippi St. lost to Alabama and beat Auburn, they got 0.27 for those two weeks.  That’s fewer points than Alabama-Birmingham got for beating Charlotte and Texas-San Antonio, for example.  I don’t think for a second Alabama-Birmingham would beat Auburn or Alabama (we’ll see how they do against Texas A&M), but since they have a lot more mediocre wins than losses they appear higher in the original rankings than Mississippi St. does.  The Bulldogs have 4 losses to the Blazers’ 1, so it’s hard for them to overcome the loss disparity AND get enough points in the 5 FBS wins to get a higher rating than UAB (with 8 FBS wins).

So why don’t I just use that system for everything?  The best example is the final rating of last year.  It exacerbated the differences between Alabama’s and Georgia’s respective schedules and gave Georgia the higher rating.  Georgia played Auburn twice, and the second time they did it, Alabama was idle because only two teams can be in the SEC Championship game.  So playing the extra quality opponent, along with the overall schedule, helped Georgia overcome having the extra loss.  I think Georgia and Alabama were close enough in my original formula, so I wouldn’t want to try some kind of average either. 

If Georgia Tech had won a couple more games, for instance, Alabama should have still been #1 after beating Georgia.  But I agreed with the top 4 it picked before the bowl games last year, so it may continue to be useful for that purpose among the top teams.  Also, the SEC is good; but I’m not sure 9 SEC teams in the top 25 (the result of the weighted formula right now) is appropriate, so I’m not using it at all for the top 25 at this time.

I think at the end of last season my weighted ratings did a better job with teams lower in the top 25, at least if your primary concern is most difficult teams to play, which is probably closer to the CFP committee’s thinking.  It’s more difficult in both my weighted ratings and in the CFP for teams in lesser conferences to rise in the rankings.    After last season, I used the original rankings for the top 10 and the weighted ratings for 11-25.  I may follow something similar at the end of this season; but since it’s new, I’m not sure if that will be the optimal solution every year.

For now, the entire top 25 is exactly as dictated by the original ratings. 

The Clemson offense celebrates a touchdown in Chestnut Hill, MA, on Saturday.  The Tigers were never threatened and won 27-7, their closest win since September.  Clemson hosts (23) Duke next week before completing the regular season against South Carolina.

I still do not expect Alabama to be first in the original ratings after next week because FCS opponents do not help ratings very much (look how far Army fell); and both Clemson and Notre Dame are playing fairly good opponents (Duke and Syracuse, respectively).  These results tell me that I made the right call last week in keeping Alabama #1 and Clemson #2 for continuity between two weeks ago and now.  Also, Clemson has clinched a berth in the ACC title game, which will help the Tigers to finish ahead of Notre Dame assuming they keep winning.

The two new Mountain West teams in the top 25 is a little strange, but this might help explain how the formula can react to obscure results.  The main reason is that both Boise St. and Utah St. were in the top 30 to begin with and both won, but that’s not the whole story.  BYU’s win over Massachusetts helped their value as an opponent not because the Minutemen are very good, but they did have good opponents themselves.  This contributed to the increase in points for both Boise St. and Utah St. as well as the Mountain West in general (BYU also played Hawaii).  Of course it also helped Boise St. a lot to beat Fresno St.  Utah St.’s win over San Jose St. didn’t count for much, but UNLV’s upset win over San Diego St. (which Utah St. does not play this year) helped the Aggies too.  Also, Boise St. and Utah St. helped one another because both are in the Mountain Division and both beat teams in the Western Division.

Utah (which has not played BYU yet) got back on track with a win over Oregon.  The Utes also have a helpful non-conference win over Northern Illinois, which is now 7-3.  It also helped that two of the three teams who beat Utah won on Saturday (and the other was idle).

Cincinnati benefited from losses by six teams between 15th and 26th (the Bearcats were 27th last week).  Three additional teams (Buffalo, Duke, amd UAB) in that range got very few points, so Cincinnati would have moved up significantly even if they’d played a worse team than then-7-2 South Florida.  Since South Florida has now lost 3 in a row and might well lose 5 in a row, the Bearcats will have to beat Central Florida to keep going up.  So I wouldn’t expect two American Conference teams to be in the top 15 for very long.

Anyway, it’s important not to just look at last week’s results and think that’s the whole story of why a team moves from Point A to Point B.

RankTeamPrev.
1Alabama1
2Clemson2
3Notre Dame3
4Georgia4
5Michigan5
6Oklahoma7
7LSU6
8Ohio St.10
9Wash St.8
10W. Virginia11
11Penn St.15
12C. Florida13
13Boise St.
14Florida17
15Cincinnati
16Texas21
17Army12
18Kentucky9
19Buffalo18
20Washington16
21Utah
22Utah St.
23Duke22
24UAB23
25Iowa St.

Out of Top 25: (14) Michigan St., (19) Fresno St., (20) NC State, (24) Iowa, (25) Boston College