Tip Top 25 in helmets, smaller
                                1936 National Championship
                                                    Home

Fixing the Final 1936 AP College Football Poll

1) Minnesota 7-1
2) Louisiana State 9-1-1
3) Pittsburgh 8-1-1
4) Alabama 8-0-1
5) Washington 7-2-1
6) Santa Clara 8-1
7) Northwestern 7-1
8)
Notre Dame 6-2-1
9) Nebraska 7-2
10) Pennsylvania 7-1
11) Duke 9-1
12) Yale 7-1
13) Dartmouth 7-1-1
14) Duquesne 8-2
15) Fordham 5-1-2
16) Texas Christian 9-2-2
17) Tennessee 6-2-2
18) Arkansas 7-3
      Navy 6-3
20) Marquette 7-2
To the left is the final 1936 AP college football top 20. You can access all of these teams' full schedules here: College Football Data Warehouse (amongst a number of other places). The fixed final AP poll, expanded to 25 teams, follows the article below.

One of the most hotly debated arguments over who should be #1 took place following the 1936 season, which was also the debut season for the official AP poll. So controversy was there from the very beginning. 7-1 Minnesota ended up #1 despite not even winning their own conference, which wouldn't happen again until Alabama repeated the feat 75 years later.

7-1 Northwestern beat Minnesota 6-0 and thereby won the Big 10 outright, but they fell to #7 in the final AP poll after #8 Notre Dame crushed them 26-6 in their finale. #2 LSU had a better record than Minnesota at 9-0-1, but then they were upset 21-14 by #6 Santa Clara in the Sugar Bowl, validating the AP poll's choice of Minnesota over them.

The controversy involved #3 Pittsburgh, who whipped #5 Washington 21-0 in the Rose Bowl, significant because Minnesota only beat Washington 14-7. That came on top of Pitt beating #9 Nebraska 19-6 (Minnesota beat them 7-0) and #8 Notre Dame 26-0 (Again, Dame beat Northwestern 26-6, who beat Minnesota 6-0).

Northwestern's touchdown to beat Minnesota 6-0 in 1936
Northwestern scoring the touchdown that beat Minnesota 6-0. That gave Northwestern the Big 10 championship, but Minnesota finished #1 in the AP poll. So Alabama's #1 finish over SEC champion LSU in the final 2011 season's AP poll was not unprecedented
.

Minnesota vs. Pittsburgh

Football writers, and no doubt fans across the nation, spent the off-season following the Rose Bowl arguing over whether Minnesota or Pittsburgh was the "rightful" mythical national champion (MNC) of 1936. Several post-bowl polls of writers, fans, and even coaches were published, mostly voting for Pittsburgh. A UPI poll went with Pittsburgh at #1, as did the Illustrated Football Annual. However, none of those sources is listed in the NCAA Records Book, and the AP poll went on to become the most influential "authority" on national championships, so today Minnesota is the consensus choice as MNC of 1936.

Slippery Rock #1?

Out of all the arguing, the Slippery Rock mystique was born. According to the school's website, a sportswriter penned a satirical piece proclaiming small Pennsylvania school Slippery Rock the 1936 national champion because they won 14-0 over Westminster, who won 7-6 over West Virginia Wesleyan, who won 2-0 over Duquesne, who won 7-0 over Pittsburgh, who won 26-0 over Notre Dame, who won 26-6 over Northwestern, who won 6-0 over Minnesota. The article was published nationally (though I have never run across it, and the writer is not named), and it is the reason Slippery Rock scores have frequently been included along with the scores of major teams by stadium announcers, radio programs, newspapers, and television networks over the many decades since.

Unfortunately, Slippery Rock finished 6-3 against what would today be a division 2 schedule, so they are not a serious option for #1, or even for the top 25, though they did defeat Indiana 10-0 and California 33-0. But to be more precise, they beat Indiana-Pennsylvania 10-0 and California-Pennsylvania 33-0.

The AP Poll and Bowl Games

There were calls for the Associated Press to conduct a post-bowl poll, but the AP declined to do so, a decision that had a big impact on college football and on how people view past "national champions" today. So why did the AP decide not to count bowl games? Primarily to protect Notre Dame and the Big Ten, who would not play in bowl games. Notre Dame was by far the most popular team in the country, and they along with the neighboring Big 10 teams were the center of the college football world at that time. A post-bowl poll would have been a huge disadvantage to them all.

Old guard Eastern teams (Yale, Harvard, Princeton) also wouldn't play in bowl games. Bowl games were seen by all these schools as bad for college football, a primary example of the overcommercialization and overemphasis of the sport. There was a decades-long war between the academic and athletic sides of major football schools, with the academic side fighting to limit or ban intersectional games in general, and to ban recruiting and scholarships. Had the AP poll begun in 1919 rather than 1936, they probably would have conducted post-bowl polls, because the Big 10, Notre Dame, and old guard Eastern teams still played in bowl games then. Harvard played in the Rose Bowl following the 1919 season, Ohio State following 1920, and Notre Dame following 1924. But the climate was very different in 1936
.

So beyond protecting its most beloved teams, the AP had a "righteous" reason to discount bowl results, since influential college football puritans in the Great Lakes and Eastern regions viewed post-season play as a contemptuous vulgarity.


Much like people have been clamoring for a playoff for decades, people clamored for a post-bowl AP poll for decades. Following the 1947 season's bowl games, the AP actually conducted a poll, and #2 Michigan passed up #1 Notre Dame in it, but the AP declared that the post-season poll did not count. Following the 1965 season's bowls, the AP conducted its first post-bowl poll that "counted," but then they went back to ending the poll before the bowls the next 2 seasons. Finally, in 1968 the AP permanently went to post-bowl polls.

People who defend the AP poll's old pre-bowl "champions" like to say that back then, bowl games didn't "count"-- they were only "exhibition" games. However, this is simply not true. The NCAA has always counted bowl games in teams' and coaches' all-time records, and while the AP poll didn't count bowl games, other writers and organizations did all along.

A "regular season" national champion means nothing more to me than a "post-October" national champion. My only interest lies with a ranking that accounts for all games played. And the AP poll didn't just ignore bowl games, it also frequently ignored regular season games that took place after Thanksgiving weekend. That had an impact on the 1936 rankings, because #16 Texas Christian defeated #6 Santa Clara the week after the final AP poll was published, and that was not a post-season game.

1936 Post-Bowl #1

Pittsburgh likely would have passed up Minnesota for #1 in a post-bowl AP poll for the 1936 season. But who should have been #1? As indicated in the intro, Pittsburgh clearly outperformed Minnesota against 3 high-profile common/comparable opponents, but on the other hand, Minnesota had 1 loss and Pittsburgh had a loss and a tie.

But does that really mean Minnesota had the better record? Pitt played a tougher schedule, and in fact, against the AP poll's own ranked teams, Minnesota was 2-1 (.667) and Pittsburgh was 3-1-1 (.700). Looks like Pitt had the better record. Furthermore, Pitt also beat 5-3 Ohio State, who will make the fixed and expanded AP top 25. OSU's only losses came to #3 Pitt by 6 points, #7 Northwestern by 1, and #8 Notre Dame by 5. They beat 5-2-1 Indiana 7-0, 4-3-1 Illinois 13-0, and 5-3-1 NYU 60-0. Clearly OSU was a top 20 caliber team, and so Pitt played 6 such teams, twice as many as Minnesota did. But they did not take twice as many losses.

Minnesota's average margin of victory was 21.4 points, Pittsburgh's 19.0, which is an unimpressive difference given how much stronger Pitt's schedule was. Pitt outscored winning teams by an average of 13.2 points, Minnesota by 10.3, and Pitt outscored AP-ranked opponents by an average of 13.2 points, Minnesota by 2.7.

So Pittsburgh #1? Well, there is a case to be made for Minnesota. Minnesota's loss came at Northwestern, a 7-1 team ranked #7. Pittsburgh's loss came at home to Duquesne, 8-2 and #14, and Pitt was tied at Fordham, 5-1-2 and #15. Duquesne lost to a pair of unrated teams, 8-2 West Virginia Wesleyan and 7-3 Detroit, and they would not have been rated at all had they not beaten Pitt. Fordham was tied by unrated 5-4-1 Georgia, and they lost to unrated 5-3-1 NYU, and they would not have been rated had they not tied Pitt. So Pitt's loss and tie were far worse than any Minnesota result. For this reason, I think you could go with either team at #1.

Therefore, if the AP poll voted Minnesota #1 after the Rose Bowl, I would not change the ranking. However, in this case, there were heavy indications that a post-bowl poll would have voted Pittsburgh #1, and a post-bowl ranking is what I am interested in.

So Pittsburgh #1? Well, I am uncomfortable with making that change based on a hypothetical AP poll-- we just cannot know with absolute certainty how that particular group of sportswriters would have voted after the Rose Bowl. So I am going with the compromise here-- they're both #1.

Move Pittsburgh up to #1 in a tie with Minnesota. Louisiana State drops to #3.

Santa Clara and Louisiana State

8-1 Santa Clara (#6) defeated 9-1-1 Louisiana State (now #3) 21-14 in the Sugar Bowl, which they would do again the next season. Obviously Santa Clara needs to be rated ahead of LSU, but LSU should remain ranked ahead of #4 Alabama (8-0-1).

8-0-1 Alabama had a better record than 9-1-1 LSU, but Alabama did not play a team ranked nearly so highly as Santa Clara. Alabama played just one ranked opponent, tying #17 Tennessee (6-2-2) at home. LSU defeated #18 Arkansas (7-3). Alabama defeated 3 major winning teams, while LSU defeated 5. And LSU vastly outperformed Alabama. All 9 of their wins were by more than a touchdown, whereas Alabama beat Mississippi State 7-0 (LSU beat them 12-0), Loyola-New Orleans 13-6, Georgia Tech 20-16, and Vanderbilt 14-6 (LSU beat them 19-0). LSU was tied at 2-6-1 Texas in their 2nd game, a much worse upset than the tie Alabama took, but everything else points to SEC champion LSU.

Move Santa Clara up to #3, dropping Louisiana State, Alabama, and Washington 1 spot each.

Northwestern

#7 Northwestern (7-1) was a victim of Lastgamitis, their last game an embarrassing 26-6 rout at #8 Notre Dame (6-2-1). But they're also the team that beat #1 Minnesota, giving them the Big 10 title outright. That win balances out their upset loss, leaving them with a better relevant record than everyone ranked ahead of them except Minnesota, with whom they were even, but had the head-to-head edge. So should Northwestern be ranked #1?

I don't think so. Their victory over Minnesota was a close home win, dependent on a roughness penalty that gave them the ball at the Minnesota 1 yard line to set up the game's only touchdown. Their loss at Notre Dame, on the other hand, was not at all close. Minnesota performed far better against common opponents, and Pittsburgh played a far tougher schedule (and they stomped Notre Dame, who stomped Northwestern).

But should 7-1 Northwestern be ranked behind 8-1 Santa Clara? Both teams took an upset loss (NW to #8 Notre Dame and SC to #16 TCU), but Northwestern defeated #1 Minnesota, who is ranked higher than Santa Clara. And Santa Clara actually played 2 games after the final AP poll. They lost their regular season finale to #16 TCU at home the week after the final poll, then "upset" LSU in the Sugar Bowl, so Northwestern may well have been ranked ahead of both Santa Clara and LSU in a post-bowl poll anyway. They certainly would have been ranked well ahead of Santa Clara had the AP at least waited until the end of the regular season to publish a final ranking.

So we'll move Northwestern ahead of both. As for 8-0-1 Alabama, they didn't beat a ranked team, barely got by some of the unrated teams they did play, and tied #17 Tennessee, so Northwestern being rated ahead of them is a no-brainer. And as established above, Alabama should remain ranked behind LSU.

Move Northwestern up to #3. Santa Clara, Louisiana State, Alabama, and Washington all drop a slot. This higher ranking for Northwestern bolsters Minnesota's one argument against Pittsburgh (their loss being better than Pitt's loss and tie).

Washington

7-2-1 Washington, originally ranked #5, did not beat a rated team, and they were tied by 2-5-2 Stanford. They were overrated, and after losing 21-0 to Pittsburgh in the Rose Bowl, they certainly would have dropped in a post-bowl anyway. We'll move them back behind 7-2 Nebraska (#9). Like Washington, Nebraska lost only to Pittsburgh and Minnesota, but unlike Washington, they did not take an upset loss or tie.

Drop Washington back to #9, moving Notre Dame and Nebraska up 1 spot each.

Yale and Pennsylvania

7-1 Yale (#12) defeated 7-1 Pennsylvania (#10) 7-0, and there isn't much more to say than that. We'll leave Penn ahead of 9-1 Duke (#11), because with Yale moved up where they belong, Penn did not take an upset loss, while Duke did (to #17 Tennessee).

Move Yale up to #10, dropping Pennsylvania and Duke 1 spot each.

Mississippi State

Unranked 7-3-1 Mississippi State was treated shabbily by the AP poll. Their 3 losses came to LSU (now #5), Alabama (now #6), and Duquesne (#14), and they tied #16 TCU. They had a better relevant record than all the teams ranked behind TCU, and they should be ranked ahead of those teams.

MSU also had a better relevant record than did TCU, but the AP poll can keep TCU ranked higher. MSU didn't actually beat anyone of much value, so that tie with TCU was their only accomplishment. TCU played an incredibly difficult schedule (11 winning teams!), they defeated 3 ranked opponents (#18 Arkansas, #20 Marquette, and now-#4 Santa Clara), and they had a fantastic finish (6-0-1 over their last 7 games, and the big wins over Santa Clara and Marquette in their last 2).

Bring Mississippi State in at #17. Tennessee, Arkansas, Navy, and Marquette all drop a slot.

North Carolina and Tennessee

The AP poll sure didn't have much respect for North Carolina back in the days of yore. UNC came into the fixed AP poll in 1943, 1939, and 1938, and they moved up 5 places when I fixed the 1937 poll. Now here they are underrated again in 1936. The issue this time is pretty simple-- 8-2 North Carolina (unranked) defeated 6-2-2 Tennessee (originally #17) 14-6.

UNC should be ranked ahead of Tennessee, and we'll put them roughly in the middle of where they are now-- behind Arkansas and Navy, and ahead of Marquette. Marquette was not impressive, and they are lucky to be rated at all.

Bring North Carolina in at #20 and drop Tennessee to #21. Arkansas and Navy each move up a spot, and Marquette drops a slot.

Princeton and Navy

4-2-2 Princeton (unranked) defeated 6-3 Navy (#18) 7-0, and they should be ranked ahead of them. Princeton was tied by unranked 3-4-1 Harvard, but they made up for that by tying #13 Dartmouth in their finale. Navy's losses came to Yale (now #10), Penn (now #11), and Princeton, and they defeated Notre Dame (now #7). Both teams had better relevant records than did 7-3 Arkansas (tied with Navy at #18), who took an upset loss and did not defeat a ranked team. Similarly, North Carolina, currently ranked right behind Arkansas and Navy, took an upset loss. So we'll move both Princeton and Navy ahead of Arkansas.

Bring Princeton in at #18 and drop Navy to #19. Arkansas drops to #20, and North Carolina, Tennessee, and Marquette all drop a slot
.

Ohio State

As previously noted, unranked 5-3 Ohio State lost to Pittsburgh (now #1) by 6 points, at Northwestern (now #3) by 1, and at Notre Dame (now #7) by 5. Those are the results of a team that could be ranked in the bottom of the top 10. Nebraska (now #8) lost to Pitt by 13, and Washington (now #9) lost to Pitt by 21, so I would rank OSU ahead of both those teams myself.

On the other hand, Ohio State did not beat a rated team, so I could accept placing them behind Navy (6-3, currently #19), who defeated Notre Dame and did not take any upsets. Arkansas, currently ranked right behind Navy, did lose to an unranked team, and they didn't defeat a rated team either, so OSU should be ranked ahead of them.

Bring Ohio State in at #20. Arkansas and all the teams ranked behind them drop a spot. We're at 24 teams now, 1 slot left to fill.

Auburn

The last slot goes to 7-2-2 Auburn. They were tied by unranked 7-2-1 Villanova in the Bacardi Bowl in Havana, and they were also tied by unranked 6-3-1 Tulane in their opener, but they defeated Tennessee (now #23), 6-3 Detroit (who beat #14 Duquesne), and 5-4-1 Georgia (who tied #15 Fordham). Tulane would be #26 (Auburn's relevant record was half a game better than Tulane's for the season) and Georgia #27. The SEC, as usual, was the top conference this season, going 40-13-5 against nonconference opponents.

Bring Auburn in at #25.

Fixed AP Top 25

Well, after 2 and a half years of toil, I've now fixed every final AP poll. But no rest for the weary-- next I'll be "fixing" hypothetical AP poll top 25s for every season 1901-1935.

No team fell out of this fixed and expanded AP top 25. Discounting their games against each other, the 5 new teams took 1 loss and 3 ties against unrated teams, and they collected 3 wins and 2 ties against teams the AP poll had rated.


1) Minnesota 7-1
    Pittsburgh 8-1-1
-0.5
+1.5
3) Northwestern 7-1 +4
4) Santa Clara 8-1 +2
5) Louisiana State 9-1-1 -3
6) Alabama 8-0-1 -2
7) Notre Dame 6-2-1 +1
8) Nebraska 7-2 +1
9) Washington 7-2-1 -4
10) Yale 7-1 +2
11) Pennsylvania 7-1 -1
12) Duke 9-1 -1
13) Dartmouth 7-1-1 --
14) Duquesne 8-2 --
15) Fordham 5-1-2 --
16) Texas Christian 9-2-2 --
17) Mississippi State 7-3-1 IN
18) Princeton 4-2-2 IN
19) Navy 6-3 -0.5
20) Ohio State 5-3 IN
21) Arkansas 7-3 -2.5
22) North Carolina 8-2 IN
23) Tennessee 6-2-2 -6
24) Marquette 7-2 -4
25) Auburn 7-2-2 IN

Fixed AP Polls
1936 National Championship
Home