SportsRatings: College Football

First College Football Playoff Rankings of 2019: What to watch for: OSU vs. LSU, Oklahoma vs. Memphis

Here we go. The first College Football Playoff Committee top 25 is about to come out. Although they don't really mean much right now, if history is any guide, these rankings will cause a shakeup in the AP Top 25 in the week following, as the sportswriters tend to follow the committee's lead.

But what does the committee value? Résumé, or performance?

The answer is, both, of course. When Ohio State beat Wisconsin 59-0 in 2014, the Buckeyes might not have had the résumé of Baylor or TCU. But they beat Wisconsin 59-0. The committee might weigh résumé and performance on a case-by-case basic. Here are two cases where we will get a clue as to what they are thinking.

LSU vs. Ohio State

The #1 spot should come down to LSU vs. Ohio State.

This year's Ohio State team is beating teams to almost the magnitude of that 59-0 over Wisconsin. Almost. They beat Wisconsin 38-7, and Cincinnati 42-0. And don't forget Miami Ohio 76-5, Indiana 51-10 or Northwestern 52-3.

LSU can claim a better résumé: they beat AP #10 Florida and #12 Auburn; they also beat Texas who fell out of the poll this week. They also beat 5-3 Georgia Southern. LSU is #1 in our Success ratings, which measures a team's wins and losses without considering victory margin.

Ohio State's résumé is almost as good, with wins over #16 Wisconsin, #17 Cincinnati, and Indiana who is just behind Texas in the "others receiving votes" department. Other victims include 6-3 Florida Atlantic. Ohio State is #2 in Success.

When it comes to team strength, there's no comparison so far: Ohio State is #1 far and away, as seen in our Strength ratings. LSU is just #5.

That's not to say LSU hasn't had its moments. They beat Georgia Southern 55-3, Utah State 42-6, and Mississippi State 36-13.

The committee has a choice here. In general, victory margin shouldn't matter, and the team with the better résumé should be ranked first. But when two undefeated teams have fairly similar résumés, you can't ignore the eye test. Ohio State is blowing out everyone they face; whether or not they could beat Alabama or Clemson, we don't know, but up to this point they clearly look like the dominant team in college football.

Verdict: It's likely the committee puts Ohio State #1, and LSU #2. If they put LSU #1, then they are making a choice that résumé is strictly more important at this point in the season, and choosing not to look at gaudy blowout scores.

Penn State's résumé ought to put them at #3. Alabama and Clemson haven't beaten any current top 25 teams, so I would expect they will follow at #4 and #5.

Even if LSU has to settle for #2, if the Tigers beat Alabama next week, they might elevate to the top spot anyway, as that would make LSU's résumé far and away the best in the country.

Oklahoma vs. Memphis

This is a comparison that not many people are thinking about, but it gets to the nitty-gritty in a few areas. It's another résumé vs. performance comparison. It's also a Power conference vs. Group of Five comparison, and it will tell us something about the value of a name.

Oklahoma sits at #9 in the AP poll, with Memphis at #19.

Here's Oklahoma's current résumé:

Record: 7-1
Loss: at #20 Kansas State (6-2), 48-41
Wins over AP top 25: none
Wins vs. opp w/winning record: Texas (5-3)

Here's Memphis' current résumé:

Record: 8-1
Loss: at Temple (5-3) 30-28
Wins over AP top 25: #23 SMU, #25 Navy
Wins vs. opp w/winning record: SMU (8-1), Navy (7-1), Tulane (6-3)

So you can see, by résumé the Tigers are clearly the more deserving of a higher rank as a 1-loss team. Oklahoma's loss—to a top 25 team—is perhaps more forgivable, but Memphis' 2-point road loss to a winning team isn't that bad. And the Tigers clearly have better wins, being the only team to beat either SMU or Navy, both of whom are ranked. Tulane isn't bad either. Oklahoma counters with a close win over unranked Texas on a neutral field.

This is reflected in our Success ratings where Memphis is #7, and Oklahoma a dismal #21.

But once again, we have to ask, how will the committee use the "eye-test", or margin of victory? 

Memphis has had some close calls against questionable opponents, like their 15-10 win over 3-6 Ole Miss, or their 42-41 escape vs. 2-7 Tulsa. Meanwhile Oklahoma has been blowing out all the cupcakes on their slate, beating 3-6 Houston 49-31, 4-5 UCLA 48-14, 3-5 Texas Tech 55-16, 3-6 Kansas 45-20, and 3-5 West Virginia 52-14. As a result, Oklahoma easily ranks ahead of Memphis in our Strength ratings, #8 to #27.

How the committee ranks these two teams will say a lot about what they are valuing right now. A good résumé? Memphis should clearly be in front. Dominant performances overall? Oklahoma, clearly. A mix of the two? Then it gets close.

I'm pretty sure most people are assuming Oklahoma will rank well ahead of Memphis, as they do in the AP poll. But the AP poll is riddled with bias from pre-season rankings (Oklahoma #4, Memphis unranked), how big-conferences are preferred over small-conferences, and program name (Oklahoma: 7 national titles, Memphis: none).

Do the committee members share these biases? I'm sure they do. Their job is to minimize these biases, but it's clear that some still do exist.

Verdict: If Memphis is ranked ahead of Oklahoma, then the committee has really looked at résumé, and decided that solid wins are more important than impressive wins over lesser competition. 

If Oklahoma is ranked slightly ahead, then they've looked at résumé and performance and struck a balance.

If Oklahoma is ranked far ahead, like they are in the AP poll, then they are emphasizing performance over résumé, and probably falling victim to one of several biases that would lead people to assume that the Sooners should be way ahead of Memphis.

I think observers will be surprised at how high Memphis is ranked, and how low Oklahoma is ranked, in the first Playoff Committee rankings. I don't think Memphis will be ahead of Oklahoma, but it will be close. 

Final note: another relative ranking to watch is Oklahoma vs. Baylor. The Bears are undefeated and beat Kansas State, the team Oklahoma lost to. That would logically put Baylor in front, though once again, Baylor hasn't always been impressive.

Best wins of all-time: Where does Ohio State - Wisconsin rank?

Ohio State just blew out highly-touted Wisconsin, 38-7, no doubt one of the best wins of the year. But in 2014, as recalled in this ESPN article, Ohio State won the Big Ten championship by beating the Badgers 59-0. Where does that win rank among the best wins of this century? Or all-time in college football.

We will use our Strength power rating to rank the biggest wins of all-time. The bigger the blowout, and the better the team you beat, the higher the game ranks.

The 2000s

We start with this century and games that most fans will recall or be familiar with. And it turns out, Ohio State's big win over Wisconsin takes the #1 spot.

1.  2014   Ohio State                 14-1   76.85  59-0  N  Wisconsin (11-3)            
2.  2003   Oklahoma                   12-2   73.70  65-13 N  Texas (10-3)                    
3.  2005   Texas                      13-0   72.12  70-3 vs. Colorado (7-6)               
4.  2001   Miami FL                   12-0   71.21  59-0 vs. Syracuse (10-3)             
5.  2014   Alabama                    12-2   70.16  59-0 vs. Texas A&M (8-5)            
6.  2018   Clemson                    15-0   67.44  44-16 N  Alabama (14-1)             
7.  2000   Oklahoma                   13-0   67.37  63-14 N  Texas (9-3)             
8.  2016   Alabama                    14-1   65.85  52-6  N  Southern Cal (10-3)          
9.  2016   Michigan                   10-3   65.61  78-0  @  Rutgers (2-10)        
10. 2015   Clemson                    14-1   65.27  58-0  @  Miami FL (8-5)

The 2014 Buckeyes crushing the 11-3 Badgers ranks as the best win of the 20-year sample from 2000 to 2019 (so far), ranking just ahead of a huge 65-13 blowout of Texas by 2003 Oklahoma in the Red River Rivalry. Note that spot #7 is also Oklahoma laying it on Texas, this time 63-14 by a 2000 Sooners team that would win the national championship. In fact, many of the teams above won the national title, including 2005 Texas who crushed a decent Colorado team 70-3 for the Big Twelve title just before beating USC in Pasadena.

2001 Miami was national champ, too, but their best game was defeating 10-3 Syracuse at home, 59-0. That seems to be a common score, as 2014 Alabama beat Texas A&M 59-0 as well. 2015 Clemson beat Miami 58-0 on the road, one point shy of our target and only good for #10. That wasn't one of the Tigers' national title teams, but the 2018 Tigers' 44-16 beatdown of Alabama made our #6 spot. Alabama finished 2nd to Clemson in 2016 but they played the best game of the year in the opener, beating USC 52-6. And the same year, Michigan walloped Rutgers 78-0, the blowout score making up for the fact that Rutgers was a whipping boy for the Big Ten that year (this hasn't changed).

If we go back farther, do these big wins stand the test of time?

Post-World War Two

It turns out that this century's biggest wins comprise half of the biggest wins since 1946, with the Ohio State-Wisconsin blowout still at #2.

1.  1977  Pittsburgh                 9-2-1  77.43  76-0   @  Temple (5-5-1)                    
2.  2014  Ohio State                 14-1   76.85  59-0   N  Wisconsin (11-3)            
3.  1968  Houston                    6-2-2  75.47  100-6 vs. Tulsa (3-7)                     
4.  1947  Michigan                   10-0-0 74.77  49-0   N  Southern California (7-2-1)     
5.  2003  Oklahoma                   12-2   73.70  65-13  N  Texas (10-3)                    
6.  2005  Texas                      13-0   72.12  70-3   vs. Colorado (7-6)               
7.  2001  Miami FL                   12-0   71.21  59-0  vs. Syracuse (10-3)             
8.  1966  Notre Dame                 9-0-1  70.86  51-0   @  Southern California (7-4-0)        
9.  1946  Michigan                   6-2-1  70.56  58-6   @  Ohio State (4-3-2)                
10. 2014  Alabama                    12-2   70.16  59-0  vs. Texas A&M (8-5)            

The #1 spot now is taken by 1977 Pitt, who beat a middling Temple team 76-0 on the road. The Panthers lost to 11-1 Notre Dame and 11-1 Penn State that year, finishing #8 a year after winning the national title. At #3 we have 1968 Houston's 100-6 win over Tulsa. Sure, the Golden Hurricane were only 3-7 that year, but the Cougars scored 100 points! A bit more "respectable" blowout occured in 1947 in the Rose Bowl, with Michigan beating USC 49-0. In fact, if we credit that as a home game for the Trojans (which it practically is, let's face it), this would be #1 on the list.

1966 Notre Dame gets in the act too with the USC beatdowns, winning 51-0 for 8th place. And 1946 Michigan takes 9th place with a big win over Ohio State. 

Let's go back another big war.

Post-World War One

Going back to 1919, the Ohio State-Wisconsin game is the only 21st century blowout that makes the list, and most of the post-WWII games are gone as well.

1.  1944 Army                       9-0-0  95.86  59-0   N  Notre Dame (8-2)                 
2.  1945 Army                       9-0-0  94.13  61-0   @  Pennsylvania (6-2)              
3.  1920 Auburn                     7-2-0  83.26  77-0   N  Washington & Lee (5-3)           
4.  1920 California                 9-0-0  80.19  127-0 vs. St. Mary's (n/a)                  
5.  1931 Southern California        10-1-0 78.79  60-0  vs. Georgia (8-2)                   
6.  1977 Pittsburgh                 9-2-1  77.43  76-0   @  Temple (5-5-1)                    
7.  1920 Centre                     8-2-0  77.22  120-0 vs. Samford (3-5-1)                    
8.  2014 Ohio State                 14-1   76.85  59-0   N  Wisconsin (11-3)                 
9.  1942 Georgia                    11-1-0 76.08  75-0   N  Florida (3-7)                   
10. 1968 Houston                    6-2-2  75.47  100-6 vs. Tulsa (3-7)    

Army's teams in the mid-1940s were monsters, and fittingly had the biggest wins of any team of the last 100 year. In 1944 Army beat Notre Dame by a familiar score: 59-0, a year before the 1945 squad topped Penn 61-0. The next biggest wins were both from 1920, just after the war ended and Auburn and California both had huge wins.  USC beat Georgia 60-0 to earn the #5 spot. 

Notice that all but one game is a shutout.

1900-Present

Army's wins aren't topped until we go back almost to the turn of the century, where Big Ten teams (err, rather, Western Conference teams) Minnesota and Michigan had monster blowouts at home. 

1.  1904  Minnesota                  13-0-0  99.44 1904  146-0 vs. Grinnell (1-6-1)                   
2.  1901  Michigan                   11-0-0  98.80 1901  128-0 vs. Buffalo (n/a)                     
3.  1944  Army                       9-0-0   95.86 1944  59-0  N  Notre Dame (8-2)                 
4.  1945  Army                       9-0-0   94.13 1945  61-0  @  Pennsylvania (6-2)               
5.  1902  Michigan                   11-0-0  90.19 1902  107-0 vs. Iowa (5-4)                       
6.  1904  Michigan                   10-0-0  89.76 1904  130-0 vs. West Virginia (n/a)               
7.  1917  Georgia Tech               9-0-0   86.56 1917  98-0 vs. Carlisle (3-6)                   
8.  1904  Wisconsin                  5-3-0   85.82 1904  82-0 vs. Drake (5-4)                     
9.  1916  Rice                       6-1-2   85.48 1916  146-3 vs. Southern Methodist (0-8-2)        
10. 1920  Auburn                     7-2-0   83.26 1920  77-0  N  Washington & Lee (5-3)  

The farther back we go, the bigger the scores get and the less importance there is on opponent quality. If you're winning by 120+ points, it's going to rank high, thus 1916 Rice beating winless SMU 146-3 sits at #9. But Army's wins over quality opponents still hold down the #3 and #4 positions. Michigan also takes the #5 an #6 spots, and another western team, Wisconsin, hold the #8 spot.

Note 1917 Georgia Tech's placement at #7 with a 98-0 win over Carlisle. Where is the famous 222-0 win over Cumberland in 1916? It turns out that Cumberland is rated so poorly, the formula doesn't rank that win high enough to make the top 10. Sometimes it doesn't matter how much you win by.

All-time College Football history

We knew that one we got back to the 1880s we'd be in the territory of Yale and the rest of the Ivy League, and sure enough Yale takes the top two spots, Princeton has #3 and #8, and Harvard #4 and #5, with only Minnesota, Michigan, and Army staying on the list from the later decades.

1.  1886  Yale                       9-0-1  123.78   136-0 vs. Wesleyan (2-6)                   
2.  1887  Yale                       9-0-0  104.15   106-0  @  Wesleyan (4-5)                   
3.  1889  Princeton                  10-0-0 103.02   98-0  @  Wesleyan (5-7-1)                   
4.  1891  Harvard                    13-1-0 101.88   124-0 vs. Wesleyan (2-7)                   
5.  1887  Harvard                    10-1-0 100.45   110-0 vs. Wesleyan (4-5)                  
6.  1904  Minnesota                  13-0-0  99.44   146-0 vs. Grinnell (1-6-1)                   
7.  1901  Michigan                   11-0-0  98.80   128-0 vs. Buffalo (n/a)                   
8.  1887  Princeton                  7-2-0   98.56   95-0  @  Pennsylvania (6-7)              
9.  1944  Army                       9-0-0   95.86   59-0  N  Notre Dame (8-2)                
10. 1945  Army                       9-0-0   94.13   61-0  @  Pennsylvania (6-2) 

The takeaway here is that if you want to have a huge win in the late 19th century, play Wesleyan. 1886 Wesleyan lost to Yale three times: 75-0, 62-0, and 136-0 (they also, somehow, beat 10-2 Lafayette 26-0). Next year's team lost to Yale three more times (they did score, losing 75-4 in their final try) and also lost to Harvard 110-0 to make the wrong end of the #2 and #5 entries. Wesleyan might look like a whipping boy, but their records aren't terrible considering they played the best teams in college football multiple times each every season!

The Modern Era: The Last 50 Years

While we don't like to exclude any of college football's history, as we go back in time the game changes a lot and there is less parity and familiarity with the teams and games. This one is probably the best representative of college football's modern period, encompassing the last 50 years, a good mix of familiar games and history, the modern era.

1.  1977  Pittsburgh                 9-2-1  77.43  76-0   @  Temple (5-5-1)                    
2.  2014  Ohio State                 14-1   76.85  59-0   N  Wisconsin (11-3)            
3.  2003  Oklahoma                   12-2   73.70  65-13  N  Texas (10-3)                    
4.  2005  Texas                      13-0   72.12  70-3  vs. Colorado (7-6)               
5.  2001  Miami FL                   12-0   71.21  59-0  vs. Syracuse (10-3)             
6.  2014  Alabama                    12-2   70.16  59-0  vs. Texas A&M (8-5)            
7.  1973  Oklahoma                   10-0-1 68.35  52-13  N  Texas (8-3)                      
8.  1984  Clemson                    7-4-0  68.33  55-0   @  Virginia (8-2-2)                  
9.  1995  Nebraska                   12-0-0 67.90  62-24  N  Florida (12-1)                    
10. 1971  Nebraska                   13-0-0 67.74  38-6   N  Alabama (11-1)                    

The 1970-2019 list is topped by Pitt, of course (since they topped the post-WWII list), followed by five games from the 21st century. Then we have yet another Oklahoma destruction of Texas, this time in 1973, followed at #8 incongruously by an unspectacular mid-80s Clemson team crushing Virginia in an early-season matchup.

Two much more notable wins that conclude the list are by Nebraska: their 38-6 win over Alabama in 1971, and their 62-24 beatdown of Florida in 1995, both wins over formerly undefeated teams that earned the Cornhuskers the national title each time.

Maryland falls in classic trap game vs. Temple

After two games, no one was looking much better than Maryland. They'd beat Howard 79-0, then proved their worth by beating AP#21 Syracuse 63-20. How good could they be, potentially? It looked like we'd have to wait until they faced Penn State to find out.

But the Terrapins faced Temple on Saturday. At first blush, I figured Maryland would win by 28 or so. At least four touchdowns, but just four touchdowns to be conservative. My power ratings, which aren't as swayed by two games, figured a 15 point advantage for the Terps. ESPN's FPI gave over a 10-point spread. Then I looked at the line for the first time:

Temple +6.5

What? Maryland is only favored by 6 1/2 points? What a steal! That's the gimme of the century, right? At least, that's what my brain said for about 3 seconds, before the rest of my pattern recognition kicked in.

That smells like a trap, it said. It looks like the line has been purposefully set low to ensnare casual bettors, to get all the action on the Maryland side...when the smart money knows Temple is going to keep it close.

How could the above be true? Who could "know" the game is going to be close? Sure, people like Temple's defense, but it wasn't all that great last year. No real "football analysis" necessarily pointed toward a close game. Several "experts" declared that Maryland would cover easily, so there wasn't general agreement that something was up.

But certain patterns over the years have become clear, if not explicitly codified. It seems that every few years there is an explosive team that runs up amazing scores for 2 or 3 games...then hits an inexplicable "wall" against an unexpected team, and just can't get their offense going. Sometimes it's because a team can't deal with heavy quarterback pressure, sometimes their "quicker" offense is outmanned by sheer size of the opposing defense, and other times, it just happens. I thought, hmm, I could see it happening against Temple.

Not only that, but after two home games, Maryland was on the road for the first time this season. Sometimes everything works at home and then on the road, nothing works. Add in the fact that last season, Temple beat Maryland 35-14 in College Park, and you might figure the swing between then and now can't be too great.

Given this odd intuition, I quickly changed my tune and realized that Temple was the better pick. And especially the under, which was 66 or so. If Temple was going to keep the game close, then it wasn't going to be due to an offensive shootout. No, this would be a boring game where neither team could generate offense. I figured Maryland would still win, maybe 23-20, maybe in overtime.

As it turned out, the score was 7-2 Temple at the half. Maryland's offense didn't score for over 30 minutes! And Temple won outright, 17-14, with the Terrapin O scoring just 13 points, the defense adding 2 safeties.

There are some who believe that the mob controls betting, and by proxy, college football via the referees in certain games. These theorists would say that the action was directed toward Maryland and the game was controlled so that Temple would win; normally this happens by key holding calls.

Maryland was penalized 9 times for 88 yards, which is fairly high, but Temple drew 7 for 65 yards. I didn't see enough of the game to know whether any were crucial to keeping the score in check on Maryland's side. I don't believe this theory enough to take it seriously, even though at time it really looks bizarrely convincing...

In any case, there are some things that show a tight game can be accounted for using only stats. CBS Sportsline's computer had Maryland winning just 20-19, eerily close to my seat-of-the-pants pick. Given that, it's not hard to see why some complex system used by the oddsmakers (since their money depends on it, they should have the best systems) pointed to a low spread. They might have even bumped it up a bit to keep the action more even, rather than bumping it down to make the action lopsided.

In my opinion, this oddball game was just "one of those things" that happens, and I've seen it happen over and over again. 

What's next for Maryland? Well, if another one of "those things" happens, then next week won't be kind to Maryland. Oftentimes when a team thinks they're doing very well, and then they are "exposed" as not being the next great thing, team morale collapses and there is major hangover to the next week. Maryland was already getting testy and fighting with Temple players as reality set in that they weren't going to steamroll the Owls.

So watch out next week vs. Penn State. I'm thinking that the Nittany Lions are going to blow out the Terps. They'll watch game film from Temple and do whatever Temple's defense was doing. And they have a better offense than Temple. 

The saving grace for Maryland might be playing on their home turf. It might turn out that Maryland is a great team at home and a relatively lousy one on the road. But either way, the Penn State game is the real pivot point for Maryland's season. A win could mean they can pick up where they left off and forget the Temple game. A blowout loss and the whole thing could fall apart, much like things did last year after they beat Texas in the opener and struggled thereafter, beating Bowling Green in game two and then...falling to Temple in game 3. They alternated big wins and big losses before losing their last four games (almost beating Ohio State though!) and finishing 5-7.

Maryland has again shown that they can be very good, and also not-so-good. Can they find any consistency, or will they continue to be the most up-and-down team in the country? Next weekend will say a lot.

150 Years of College Football: Top 15 all-time (best team from each decade)

ESPN recently featured its poll of the top 150 teams of all time, honoring the 150th anniversary of college football's first game in 1869.

(Fun fact: 2019 is the 150th year of college football, as well as the 150th anniversary. Normally the 150th anniversary means you're on year 151. But since there were no games in 1871, 2019 is year 150).

ESPN's list isn't bad, but suffers from recency bias. They polled 150 (of course) players, coaches, writers, etc. to form their list. Clearly these people are experts in college football over the last 50 years. Some may go back a little farther, maybe 75 years, but that's half of the history of college football that isn't going to be in their memories.

As a result, of course there are far more teams from recent decades at the top of the list. Their top ten features five teams from 2000-2018, one from the 1990s, three from the 1970s, and one from the 1950s.

This isn't necessarily a bad thing; it's obvious that players today are better than those from the pre-WW2 era. Many earlier teams are included farther down in ESPN's list, but in all, there are 82 teams from the last 50 years and just 10 from the first 50 years. If you're trying to survey all of college football's history, there should be more of an effort to include earlier eras. 

That's what we've done. In order to sample the entirety of the college football experience, we've made a list of the best teams, but allow only one team per decade. This ensures no eras will be forgotten, and also conveniently limits the amount of writing we have to do but limiting the list to 15 :)

We've used our power rating, combining Strength (points-based) and Success (win-based) to find the top team from each decade, then ranked those #1 to #15. Here are our results.

1. 1945 Army (9-0-0)  Rating: 93.41  ESPN rank: #58. ESPN top team from 1940s: 1949 Oklahoma (#37)
   This all-time classic Army squad ranked a clear #1 (by nearly 20 points) over #2 Alabama in Strength, and edged the 10-0-0 Tide in Success as well. They had both our #1 ranked Offense and Defense as well. It might be easiest to convey how good this team is by looking at other teams from the same year. 6-2 Penn shut out four of their first six opponents, but lost to Navy 14-7; then they hosted Army and lost 61-0. That Navy team was 7-0-1 with a tie vs. Notre Dame when they played Army and lost 32-13. The Notre Dame team that tied Navy? They lost to Army 48-0. Army's full season is well-documented on this page which also includes much info about the squad.
   
   ESPN's poll really dropped the ball by ranking this team #58. For at least a quarter century this team was considered the clear all-time best, and it still shows up as the best team of all time in many books and websites. You can argue that a 1945 team couldn't compete with today's teams, but if that's the case then ESPN shouldn't have teams from the 1950s (or even 1970s?) in their top ten. And the 1949 Oklahoma team was arguably only the 2nd best team of that year, let alone best of the decade.
    
2. 1971 Nebraska (13-0-0) Rating: 87.27  ESPN rank: #1. ESPN top team from 1970s: 1971 Nebraska (#1)
   ESPN redeems themselves somewhat by ranking this Cornhuskers squad at #1. In terms of dominance and strength of schedule, there's hardly a better example. Nebraska was 10-0, outscoring opponents by an average of 39-6 when they met Big Eight foe Oklahoma, 9-0 and winning 45-16. In an all-time classic, Nebraska beat Oklahoma—in Norman—35-31. This set up a meeting with 11-0 Alabama in the Orange Bowl which was almost like a 4-team playoff long before the playoff era, as 11-0 Alabama had defeated 10-1 Auburn.
   
   The Cornhuskers beat the Tide in dominant fashion, 38-6, and Oklahoma 'completed the square' by beating Auburn 40-22. Thus Nebraska had wins over two 1-loss teams, a rarity before the playoff era, and their win over Colorado gave them wins over the final AP #2, #3, and #4 teams. Naturally they were unanimously #1. The team finished with our #1 defense and #2 offense (Oklahoma was #1 on offense). 
    
3. 1889 Princeton (10-0-0)  Rating: 86.02  ESPN rank: Not ranked  ESPN top team from 1880s: 1888 Yale (#94)
   The football scene at this time was dominated by Harvard, Yale, and Princeton. Princeton's schedule in 1889 is pretty non-descript apart from their games with these 2 teams, though they did hand Lehigh 2 of their 3 losses. Harvard was odd man out in the national title race, losing to both Yale and Princeton, by scores of 6-0 and 41-15. Which brings us to Yale, who was 16-0 upon meeting 9-0 Princeton.
   
   Yale had an incredible thirteen shutouts under their belt. They beat 7-2 Cornell twice—both of Cornell's losses—by a total of 126-6. Still, our power rating would have predicted a 14-point Princeton win. Princeton won 10-0, which I'm sure still beat the spread vs. a team with 13 shutouts. Princeton had our top-rated offense and defense, and according to Wikipedia their captain was Edgar Allan Poe, 2nd cousin of the author, and that's gotta be worth a rating on ESPN's top 150 alone! It's hard to argue too much with the 1888 Yale team, though, as they didn't give up a single point all year. Still, our power rating rates 1889 Princeton as the tougher team and having the harder schedule.
    
4. 2018 Clemson (15-0)  Rating: 82.09  ESPN rank: #5  ESPN top team from 2010s: 2018 Clemson (#5)
   Another team on which we can agree! There are two things which should put this team high on anyone's list. First, all through the 2018 season, Alabama was looking like an all-time top-ten team, maybe the best ever. When Clemson throttled the Tide 44-16, didn't that make the Tigers an all-time great? Second, they won 15 games in the modern era including two dominant wins over 1-loss teams (Alabama and Notre Dame, whom they beat 30-3), both historic accomplishments. 
   
   It's true that the playoff era is going to give teams a higher rating in almost any system, from humans or computers. More wins, especially wins over playoff teams, juice a team's résumé. Particularly in ELO systems (which our Success rating is not), two late-season, huge wins count for a lot. But in this case it's very deserved. Clemson's season prior to the playoff was impressive but not amazing despite their 13-0 finish, and even their two big wins at the end didn't put them #1 in our FBS-only Strength model, though they had the #1 defense. They were a clear #1 in Success, and there was no doubt in the voters minds, of course, as they finished a unanimous #1.
    
5. 1890 Harvard (11-0-0)  Rating: 80.04  ESPN rank: not rated. ESPN top team from 1890s: 1892 Yale (#105)
   It's hard to explain how good these teams are with little film, and no living first-hand witness to their greatness. But we can try to explain why our system rates them so highly. And it's easy to see by looking at their dominant scores. Harvard finished 11-0, with 9 shutouts, and a national-title-winning 12-6 win over 13-1 Yale. Basically you had 1890s version of an undefeated Clemson vs. Alabama, and Harvard was Clemson in this scenario. As for what makes them the best of the 1890s decade, look at their dominance in outscoring their opponents 555-12. How do they compare to ESPN's choice, 1892 Yale? Again it's hard to argue with a team that doesn't allow a score (Yale had 13 shutout wins), but our Strength power rating likes 1890 Harvard a lot more. Six of one, half a dozen of the other—it's hard to compare these late 19th century Ivy League teams, but we find our power ratings' pronouncements interesting nonetheless.
    
6. 1905 Chicago (11-0-0)  Rating: 77.36  ESPN rank: #148. ESPN top team from 1900s decade: 1901 Michigan (#92)
   It's nice to see this team slip into ESPN's list, albeit just barely. The early 1900s era was one in which it was very hard to compare teams as they didn't necessarily play outside their region much, leading to a big debate as to whether the midwest teams (particularly Michigan) could beat the dominant Ivy League teams. Princeton didn't lose to a "midwest" (non-eastern) team until West Virginia in 1919; Harvard in 1921 to Kentucky's Centre College; and Yale fell to Iowa in 1922. Before that, the midwest and Ivy were mostly separate football worlds, and the national champ was declared by sportswriters based on their best guess, with Michigan being the midwest's premiere team. In 1905, Michigan was 12-0 and hadn't ceded a single point when they met 10-0 Chicago, who also hadn't given up a single point all year. If you predicted a defensive battle then you would be correct as Chicago won 2-0, the result of a "safety touchdown" which was worth 2 points then as it is now (without the "touchdown" aspect in the name). Just out of curiosity, we ran this game through our power rating, prior to the date, and it predicted a 17-7 Chicago win. Not sure where it was expecting to find all that offense!
    
7. 1995 Nebraska (12-0-0)  Rating: 74.76  ESPN rank: #4  ESPN top team from 1990s: 1995 Nebraska (#4) 
   We pretty much agree with the very top teams in ESPN's list, from a modern standpoint. The best team of the 90s was clearly this Nebraska squad, which dominated all its opponents including beating 12-0 Florida 62-24 in the national championship game. They were only #2 in defense but a clear #1 on offense, and rated 12 points better than the #2 team in Strength. In terms of Success, or wins over other good teams, they also beat 10-2 Colorado, 10-2 Kansas, and 10-2 Kansas State (who beat each other round-robin for their other loss), giving the Huskers four wins over the final AP top 25 (#2, #5, #7, and #9).
    
8. 1920 California (9-0-0)  Rating: 72.59  ESPN rank: #139  ESPN top team from 1920s: 1924 Notre Dame (#107)
   ESPN says it pretty well: "The greatest of the Bears' 'Wonder Teams' of the post-World War I era shut out seven opponents, averaged 56 points, gave up only 14 points all season, and blanked Ohio State 28-0 in the Rose Bowl."  Remarkably, despite all those shutouts Cal's defense only ranks #13 in our power rating, since there were so many other teams shutting all their foes out, too. But their offense ranked #1, and they were far and away #1 in Strength. As far as strength of schedule goes, their first six games were non-descript, though beating St. Mary's 127-0 when Stanford only managed a 41-0 win helped their Strength, as did beating Nevada 79-7 when Nevada was 2-0 otherwise. But what made their season was the win over Ohio State, as the Buckeyes had won every other game and given up no more than 7 points previously.  
    
9. 1917 Georgia Tech (9-0-0)  Rating: 71.67  ESPN rank: not rated  ESPN top team from 1910s: 1919 Notre Dame
   Georgia Tech didn't even place first in Success in 1917—that went to 10-0 Pitt—but they were a close 2nd and their Strength was far and away #1, thanks to an offense that averaged 54 1/2 points per game. This wasn't even the team that scored a record 222 against Cumberland—that was the previous year's squad. The Yellow Jackets (this season pre-dates the Ramblin' 'Reck nickname even) were declared national champs by the powers that be at the time, and they were coached by John Heisman at the time. ESPN's pick of 1919 Notre Dame as best team of the decade is weak; their schedule included only one winning team that played a full schedule that year, and is best remembered for Knute Rockne being on the team.
    
10. 1988 Notre Dame (12-0-0)  Rating: 69.85  ESPN rank: #18  ESPN top team from 1980s: 1987 Miami (#11) 
    Now here's a real Notre Dame team. The Fighting Irish weren't expected to be great that year, but won their way through a tough-as-nails schedule to become national champs. They weren't the best team—both Miami and Florida State were better teams that year—but they played the best when it counted. And it particularly counted when they beat Miami 31-30 in South Bend, the game that decided both the national title and the honor of placement in this top ten. Notre Dame barely survived a trip to Michigan, 19-17, in game one, and had to travel to USC late in the season, winning 27-10, before earning a berth in the national championship game against West Virginia. Winning 34-21, that gave the Irish wins over the #2, #4, #5, and #7 teams in the final AP poll. Miami beat the #3, #4, and #10 teams and were within a 2-point conversion of beating Notre Dame; had they converted, Miami would be on this list at #7. Was the 1987 Miami team more deserving of "team of the 80s" than 1988 Notre Dame? Maybe, as the Hurricanes beat 11-1 Florida State and 11-1 Oklahoma that year. But paradoxically, the Canes couldn't get a win over themselves, and it was Notre Dame's win over Miami in 1988 that gave them uber-points in our Success rating.
     
11. 2004 USC (13-0)  Rating: 68.65   ESPN rank: #2  ESPN top team from 2000s decade: 2004 USC
    ESPN puts them a bit higher but we agree this was the team of the decade of the 2000s. The Trojans dominated a fairly non-descript schedule, surviving a scare vs. 10-2 Cal in October, and matching up with 12-0 Oklahoma who was #2 at the time. They blasted the Sooners 55-19, making the Trojans the clear #1—though undefeated Auburn picked up a few first-place votes, too. Our Strength power rating puts Auburn at #7, though in Success they finish a close #2 to USC. Few people recall that Utah also finished undefeated, and we rate them #4 in Strength (#4 in Success). 2004 would have been a great year for a 4-team playoff, but alas it was not to be. We think the correct choice got made despite the limitations.
     
12. 1931 USC (10-1-0)  Rating: 68.60  ESPN rank: not rated  ESPN top team from 1930s: 1939 Texas A&M (#81) 
    Here's a team we don't blame ESPN for not recognizing, as 1-loss teams are no doubt going to get lost in the shuffle. But our power rating has a soft spot for the Trojans of 1931. The reason? After suffering that one loss in their opener (13-7 to 8-2 St. Mary's) the Trojans rattled off 10 straight wins, culminating in a Rose Bowl win over 11-0 Tulane, the only other team that could really stake a claim to being the best in the nation. In all, 9 of the 10 teams they beat finished with winning records, including 8-2 Georgia (a 60-0 win), 6-2-1 Notre Dame, and 8-2 California. Interestingly, ESPN's '30s pick, 1939 Texas A&M, also beat Tulane to win their national title, but just 14-13. The Aggies played a weaker schedule, and were underwhelming offensively. But they were undefeated, can't take that away from them.
     
13. 1952 Michigan State (9-0-0)  Rating: 57.41  ESPN rank: #115  ESPN top team from 1950s: 1956 Oklahoma (#9) 
    Here, ESPN makes the more acceptable pick for the 1950s. We have Michigan State from 1952, even though the Spartans finished #1 neither in Strength or Success that year. They were #2 in both, and edged out Georgia Tech and Oklahoma for our overall #1 rating. The Sooners finished 8-1-1 that year, meaning they weren't going to be national champs, while Georgia Tech was 12-0-0 but only finished #5 in Strength. The Spartans finished #1 in the AP and coaches poll. The real comparison should be with the great Bud Wilkinson teams of the 1950s at Oklahoma. In our power ratings, the 1956 team gets a lot of credit for being the best team, far and away #1 in Strength that year. But they had a very easy schedule, finishing just #8 in Success despite their 10-0 record. The problem is that they didn't go to a bowl game, having won the Orange Bowl the previous year and the Big 8 not sending teams in consecutive years. That left Oklahoma with just one win over a winning team, which to a computer is not a very dominant season. Obviously humans can consider other factors, like how the Sooners hadn't lost a game in 3 years; but in isolation, to our power rating the 1956 Sooners are a dominant team that didn't get the chance to prove their greatness. 
     
14. 1966 Notre Dame (9-0-1)  Rating: 52.60  ESPN rank: #125  ESPN top team from 1960s: 1968 Ohio State (#13)
    The 1960s were a weak era in college football, and not many teams are recalled as all-time greats. That said, the 1966 season was a great one due to three undefeated teams: Alabama, Notre Dame, and Michigan State. Alabama finished 11-0 while Notre Dame and Michigan State fought to a 10-10 tie. Just how are you supposed to rank those teams? Well, according to our Strength power rating, Notre Dame is far and away the best team, Alabama is #2, and #4 Michigan State was lucky to come away with a tie on their home field. In terms of Success, Alabama's undefeated season narrowly trumps the Irish, but not by enough to put the Tide on top. Who's to say that MSU and Notre Dame aren't the two best teams, and Alabama lucky not to play either of them? It could be debated forever, but Notre Dame's wins over 9-2 Purdue, 8-2 Army, and 51-0 shellacking of USC tipped our system to the Irish. ESPN's pick of 1968 Ohio State isn't a bad one for the 1960s, as there aren't really any all-time great teams in the decade. Ranking them at #13 of all-time seems a bit excessive, though, as they beat up on three teams that were 1-9 or 0-10, beat just 4 teams with winning records, and were relatively weak on defense.
     
15. 1878 Princeton (6-0-0)  Rating: 10.73  ESPN rank: not ranked  ESPN top team from 1870s: none
    Finally, we come to the representative of college football's first decade, 1869-1879, and 11-year span with 10 seasons (sans 1871). There are 2 reasons why this team's rating is so low compared to the others. In terms of Success, there were fewer teams to compete against, and fewer games were played. Fewer wins, against teams with fewer wins, equals a lower Success rating, even for an undefeated team. In terms of Strength, winning margins were much lower in this era simply because a touchdown was worth only a single point. So scores of 5-0 were the norm rather than a curiosity. This meant that of 7 relevant teams in 1878, Princeton was #1 with a Strength of 2.37, and at 6-0-0 had a Success of 8.36, for an overall score of 10.73. So much for the ability to compare across eras, but we don't think it's a bad thing that the first decade, where football was more or less still rugby and played only in the Ivy League, is downgraded a bit. The only important game of this season was when Princeton and Yale, both 5-0, met and Princeton won 1-0. Backchecking our power rating, we expected a more exciting game, with Princeton winning 2-1. We got the spread right if not the over/under.

There you have it. Make of it what you will, and if you disagree...well, you can always make your own power rating, I guess :)

 

Week 2: Are Maryland and Wisconsin the best teams in the Big Ten? Or the nation?

The early season prognostication for the Big Ten put Michigan and Ohio State at the top of the heap. This is pretty typical of the last several years, showing how little things have changed in the Big Ten since the 1970s despite other teams occasionally rising to the top (and the league expanding from 10 to 14 teams). 

But early results present two challengers: Wisconsin and Maryland.

The Badgers are not a big surprise, but winning their first two games 110-0 is a surprise for a team that stereotypically does not put up a lot of points. And two shutouts in a row is a rarity for any team, against any competition.

Maryland is a big shock; the best expected of the Terps was a mid-league finish, at worst, the bottom. But after crushing hapless Howard 79-0, the Terrapins proved they're for real by routing highly-touted Syracuse 63-20. Where did all that offense come from?

Here's a rundown of the teams that played the best, in absolute terms, in week 2, according to our Bayesian power rating, followed by a Top Ten ranking formed only by this year's performances (no prior ratings consulted):

1. Oregon 77, Nevada 6 (rating: 89.2)
   The Ducks' 77-point explosion begs the question: how good is Auburn's defense for holding them to three touchdowns? Nevada, after all, beat Purdue the week before 34-31, so Oregon's defensive performance is pretty on-point, too. This win signals that Oregon might still be the best in the Pac-12, something that would be bad for the Pac-12's playoff chances and good for the SEC's odds of getting two teams in the "final four."
2. Ohio State 42, Cincinnati 0 (85.3)
   ESPN had Ohio State on "upset watch" due to Cincy's relatively high rating in the FPI, but the Buckeyes made a mockery of that by dominating the Bearcats start to finish in a shutout win. Last week they let Florida Atlantic score 21 points but the defense proved themselves this time.
3. Clemson 24, Texas A&M 10 (83.5)
   The week's best victory in résumé terms is our third-ranked game performance, as Clemson didn't give the Aggies a chance to think about the upset. If not for a late touchdown allowed they would have finished #2 on this list for a second consecutive week. 
4. Missouri 38, West Virginia 7 (81.6)
   Here's one that's probably overrated by last year's "priors". West Virginia finished #9 in last year's ratings and is still clearly overrated despite being downgraded for last week's scare vs. James Madison. Missouri's loss last week to Wyoming confirms it. But a 31-point win over the Mountaineers is still impressive, even if it is most likely out of place on this list.
5. Alabama 62, New Mexico State 10 (81.2)
   The Crimson Tide keep doing as expected, meaning they beat up both on decent teams, like Duke last week, and poor teams, like New Mexico State.
6. Maryland 63, Syracuse 20 (81.1)
   The Terps have their second entry in the top-ten performances list in two weeks. Last week we wondered what would happen once they faced a good team. The result? 63 more points put up and a 43-point win over a borderline top 25 team.
7. Oklahoma 70, South Dakota 14 (78.6)
   The Sooners' win over Houston didn't make the list last week because they allowed the Cougars 31 points. This time they held South Dakota to 14 and also hit the 70 mark, a solid performance against a good FCS team.
8. Wisconsin 61, Central Michigan 0 (77.8)
   The Badgers make the top ten list again. Last week's shutout was on the road, this one was at home, and neither opponent is very impressive. But winning two games 110-0 is a feat that should make everyone take notice.
9. Utah St 62, Stony Brook 7 (76.7)
   Stony Brook is a good FCS team and Utah State demolished them. Last week's loss to Wake Forest pretty much negates that chance that the Aggies will be a top 25 team this year (at least for several months) but it shows that maybe they've rebounded after the loss.
10. Central Florida 48, Florida Atlantic 14 (74.4)
    A few years ago, week one results suggested that UCF might be among the best teams in the country. After finishing 13-0 and beating Auburn, that early sentiment was proven correct. This year UCF has started strong yet again; their week one win over Florida A&M (62-0) just missed making the top ten, and this one just makes it. It points to the possibility of a third straight undefeated regular season, something thought impossible without McKenzie Milton at QB. They get their biggest test next week when they host Stanford.

And now, the composite ranking of the teams that have played the best so far (their two-game average), without "diluting" it with pre-season rankings:

1. Clemson (2-0, 86.2). Not surprisingly, the defending champs are #1
2. Alabama (2-0, 85.9). Even less surprisingly, last year's runner-up is right there with them, essentially tied for the top spot.
3. LSU (2-0, 78.8). A big gap to #3, but LSU's win over Texas and runaway win over Georgia Southern make them a threat to 'Bama.
4. Maryland (2-0, 77.5) The biggest surprise of the year so far, Maryland could be a real player in the Big Ten—and the nation.
5. Wisconsin (2-0, 77.5) Is it too soon to call for Badgers vs. Terps in conference championship?
6. Ohio State (2-0, 77.1) The Buckeyes aim to remind Maryland and Wisconsin who is in charge; they face them back-to-back in a few months!
7. Penn State (2-0, 77.1) The FOURTH Big Ten team in the top ten. And that's without pre-season favorite Michigan. A big year in store for B1G?
8. Georgia (2-0, 76.1) The third SEC team ranked. The Bulldogs hope to face the 'Bama/LSU winner in the SEC championship. 
9. Oregon (1-1, 74.9) They nearly beat Auburn, and killed New Mexico State (Auburn's tepid 24-6 win over Tulane kept the Tigers out of the top ten).
10. Oklahoma (2-0, 73.6) The Sooners are looking good...now they just have to be sure to beat Texas (they're 8-7 against the Longhorns since 2005!)

So there you have it. The race in the Big Ten (really, the Big Ten?!) looks the most interesting so far, while the "race" in the ACC is easily the most boring. The SEC is having a typical year where LSU challenges in the West and Georgia looks headed for the championship game. The Pac-12 is struggling again, and the Big Twelve's hopes as usual ride on Oklahoma.

So let's hear some praise for Maryland and Wisconsin. They might be the only teams preventing 2019 from becoming a carbon copy of the 2018 season!