Template talk:College Football National Champion pre-AP Poll navbox

Which selectors

So, the template currently identifies only some of the national champions recognized by the NCAA record book. A few examples include:

1935 - the template lists Minnesota, Princeton, and SMU, but not LSU or TCU which were named co-champions by Williamson
1933 - the template lists Michigan and Princeton, but not Ohio State (Dunkel) or USC (Williamson)
1932 - the template lists Michigan and USC, but not Colgate which was named co-champion by Parke Davis
1931 - the template lists USC and Pitt, but not Purdue which has precisely the same claim as Pitt (Parke Davis co-champion)
1929 - the template lists Notre Dame and Pitt, but not USC which was picked as NC by Berryman, Houlgate and Sagarin
1928 - the template lists Georgia Tech and USC, but not Detroit which was named as co-champion by Parke Davis
1927 - the template lists Illinois, Georgia, and Texas A&M, but not Notre Dame (Houlgate) or Yale (CFRA)

We need to figure out which ones are included. The options appear to be:

List all of the national champions recognized by the NCAA, an option that avoids WP:OR issues.
Establish an objective and reasonable methodology to determine which to include (e.g., minimum 2 selectors or only specified selectors [an option that raises WP:OR issues since the NCAA recognizes all as major)

Cbl62 (talk) 18:17, 13 January 2016 (UTC)[reply]

Indeed. Berryman and Williamson both seem eccentrics. Also think of things like 1) one cannot lose their last game and/or 2) against another claiming a championship; as a way to clear up potential oddities. Davis selections are probably oversights. Cake (talk) 18:37, 13 January 2016 (UTC)[reply]

More discrepancies

A few more instances in which the template currently identifies only some of the national champions recognized by the NCAA record book:

1926: the template lists Lafayette (Davis only), but not Navy (Boand and Houlgate) or Michigan (Sagarin)
1925: the template lists Dartmouth (Dickinson and Davis), but not Michigan (Sagarin) Michigan lost to Northwestern, always saw Alabama's first one have Dartmouth as a footnote, not Michigan. This is a good example of our challenge. Yes, Michigan was a one-loss team, albeit one that gave up only 3 points all year and was rated by Yost as the greatest team he ever saw. Sagarin ranks them as national champ, and NCAA considers Sagarin's rankings as one of the major/official rankings for 1925. Isn't it WP:OR for us to reject the NCAA guide and deem Michigan unworthy based on subjective opinions? Also, it would be inconsistent to exclude Sagarin's 1925 pick, but then include 1923 Cornell which was also selected only by Sagarin. I am aware that 1925 Michigan is a particularly strong team – Friedman to Oosterbaan. In this case I am not dogmatic about leaving them out, but there are sources which leave them out, and there seemed good reason to do so, Yost's comment aside (I wager 2001 UF is Spurrier's best; doesn't mean they beat '01 Miami). Bama won the Rose Bowl, and Dartmouth was the power in the East.
1923: the template lists California (Houlgate only) and Cornell (Sagarin only), but not Yale (Berryman) Might be the only case where Berryman looks both innovative and sane.
1922: the template lists Cornell (Helms and Davis), but not Iowa (Billingsley) or Vanderbilt (Berryman)
1921: the template lists Lafayette (Boand and Davis), but not Iowa (Billingsley and Davis), Vanderbilt (Berryman), or Wash & Jeff (Boand) W&J was an oversight. Back and forth on those Iowa teams, but ultimately Iowa does not claim them. Quite biased, and still I would not include 21/22 Vandy on this template. They had to fight in the papers for their southern claims. If '22 Vandy, why not '22 Michigan, oh because '22 Iowa...who doesn't even claim it, as do neither of these schools. In '21 Vandy is not considered stronger than Georgia Tech or Centre without even getting to the East. Regarding Iowa, I don't think that the university's decision to "claim" a title should matter; all that should matter is the selections of those deemed official by the NCAA. Here, Iowa was named NC by two of the major selectors, and so I can't see a sound basis for not including Iowa in the template.I have to disagree. With rare exception, whether the school claims the title or not is the quickest way to ascertain contemporary opinion, rather than whether they happen to fit a formula applied at a later date. 1922 is one of the years I track best. Iowa is Western champs those years for sure, but Princeton, Cornell, and California are in an elite tier for 1922. I'd bet any would have been favored in a contest. Iowa/Michigan/Vandy/Army (2 ties), probably in that order, are your next tier for 1922, with Iowa and Vandy making it difficult for retroactive selectors. The feats of Iowa were in '21 beating ND and in '22 beating Yale. The first especially is a big deal, but, for example, '21 and '22 Cornell thrashed the entire East. Iowa was not considered in the same league with the Wonder teams of Cal or Princeton's team of destiny. Michigan's 1922 success only hinders their claim.
1920: the template lists Princeton (Boand and Davis), but not Notre Dame (Billingsley and Davis), Harvard (Boand), or Georgia (Berryman) Harvard an oversight - Georgia not. 1920 Georgia Tech arguably as strong to contemporaries, without even leaving the state. UGA doesn't claim a title until 1942, and only 1927 is slippery. 1920 Georgia is easy to leave out.
1919: the template omits Notre Dame (NCF and Davis), Texas A&M (Billingsley and NCF), and Centre (Sagarin) Centre did not overtake Harvard until 1921 remember. They were not considered national champions in the same league as Harvard and Illinois. I worried the same for aTm. ND does not claim a title until 1924, as I understand, but the inclusion of 1919 and 1920 would not bother me. Regarding undefeated Centre: The template should either include or exclude Sagarin's picks. Currently, it includes 1923 Cornell, but not 1925 Michigan or 1919 Centre. This type of inconsistency is based on subjectivity and WP:OR. Regarding Texas A&M: two of the major selectors named them NC, so I see no objective basis for excluding. Regarding Notre Dame: a university's subjective "claim" should be irrelevant; ND was picked by two major selectors in both 1919 and 1920, and that's what should count.
1916: the template omits Georgia Tech (Billingsley) and Army (Davis) 1916 Georgia Tech was a power but its claim to national champion is a phantom. PS Didn't someone list Colgate in '16? The NCAA guide does not list Colgate. Why is Ga. Tech's claim phantom? Picked by Billingsley, undefeated team, outscored opponents 422 to 20, not seeing much wrong with that. A great year no doubt but they were still a southern team. They also suffered a tie to W&L. Most importantly, any research into 1917 will show you Tech is not claiming itself as repeat southern and national champion, but the first-ever. It's part of the Pitt-GT lore then. Pitt was '16 NC, and GT was '17 NC once Pitt turned them down (also much better w/ common opponents), then in '18 Pitt reasserts itself. Odd about Colgate.
1915: template includes Pitt (Davis only), but not Minnesota (Billingsley) or Oklahoma (Billingsley MOV)
1914: template includes Illinois (Billingsley and Davis) but omits Army (Helms, Houlgate, NCF, Davis) and Texas (Billingsley MOV) Army was an oversight
1913: template omits Auburn (Billingsley MOV) Auburn is tricky in that they did go undefeated in the schedule given, and it is perhaps their greatest team, so I do not blame them for claiming it. However, again, any contemporary account will not declare them national champion. Southern champion, to be sure (PS Auburn I understand wishes to claim national titles 1913 (OK..) 1914 (not positive they would beat Tennessee), 1910 (lol no, Vandy would crush))
1912: template omits Penn State (NCF)
1911: template omits Minnesota (Billingsley) and Penn State (NCF)
1910: template omits Pitt (NCF)
1908: template omits Harvard (Billingsley) and LSU (NCF) 1908 LSU's page will give you the explanation there.
1907: template includes Penn (no selectors) 1907 season article says Penn claims it despite that. Can also see the 1907 Carlisle article, which so far I've written. Since no NCAA-sanctioned selector recognizes Penn as NC, that should be the end of the story. A university athletic department can claim anything they want, but that doesn't make it so.

Note: I tend to agree that a claim based solely on Billingsley MOV (an alternate method used by Billingsley) probably should not be included in the template. Cbl62 (talk) 22:42, 13 January 2016 (UTC)[reply]

I don't have an exhaustive memory of every champion, nor was I wise enough to always refer to the category, such that some are left off unintentionally for others in future to add and provide further refinement. However, I effectively leave off Billingsley, Berryman, and Sagarin, at least by themselves. I try to have some knowledge of southern football during this stretch. 1917 Georgia Tech is the first southern national champion, period, and the second in 1925 Alabama. With that in mind, I put my comments above in red. Otherwise it was just a wall of text here. Pardon I went at it chronologically, so you might start at the bottom. Cake (talk) 03:18, 14 January 2016 (UTC)[reply]

I have now tracked the template to the NCAA record book for the years 1869-1919 so that it includes every NCAA-recognized claimant to the national championship with the exception of those whose sole claim is under the Billingsley MOV (i.e., alternate) system. Cbl62 (talk) 02:46, 14 January 2016 (UTC)[reply]

The championships hard to justify:

1935 LSU = 2 L, including to TCU; Williamson co-champion

1935 TCU = 1 L, to SMU, Williamson co-champion

1933 Ohio St = 1 L, to Michigan, Dunkel champion

1933 USC = 1 L 1 T, Williamson champion

1927 Yale = 1 L, to Georgia, Football Research champion

1927 Notre Dame = 1 L 1 T, loss to Army which loss to Yale, Houlgate champion

1927 Texas A&M = 1 T, unclaimed until 2012, Sagarin champion

1927 Georgia = 1 L to next year's champ Georgia Tech in the rain (ahem, very similar to UF's loss to Tenn the next year), Berryman, Boland, Poling champion

1926 Michigan = 1 L, to Navy, Sagarin co-chammpion

1925 Michigan = 1 L, Sagarin co-champion

1922 Vanderbilt = 1 T, Berryman champion

1921 Vanderbilt = 1 T, Berryman champion

1920 Georgia = 1 T, Berryman champion

1919 Texas A&M, played nobody outside the Southwest, unclaimed until 2012, Billingsley MOV champion, NCF co-champion

1919 Centre, played nobody stronger than WV, Sagarin co-champion

1916 Georgia Tech, 1 T, Billingsley champion

1915 Oklahoma, played nobody out of southwest and missouri valley, Billingsley MOV champion

1915 Minnesota, 1 T to Illinois, Billingsley champion

1914 Texas, only beat Wabash outside of southwest and missouri valley, Billingsley MOV champion

1913 Auburn, played nobody outside of the south, Billingsley MOV champion

1908 LSU, even their Southern title was vacated by some as season was clouded by professionalism, played nobody out of south, southwest and missouri valley, think historians just wish Doc Fenton's season isn't forgotten. Cake (talk) 05:33, 15 January 2016 (UTC)[reply]

I don't disagree that some of these claims are harder to justify than others. This is the dilemma we face in the pre-AP poll era. There is no single selector who speaks with the same gravitas as the AP starting in 1936 and the UP starting in 1950. We are left with several selectors who the NCAA deem to be major selectors for these years. The NCAA is the most authoritative body we have. For us to supplant its judgment with our own in deciding which selectors should be recognized, or to omit selected years' picks because we find them undeserving, would IMO run afoul of WP:OR. Cbl62 (talk) 18:16, 29 January 2016 (UTC)[reply]

I'm stuck in a conundrum in that I don't dispute this, but I do discount it a bit. You might see the yearbook of the 1917 Ga Tech team to see what I mean about the 1916 team's claim being a phantom. Cake (talk) 18:46, 29 January 2016 (UTC)[reply]

@Cbl62: I think it would look much cleaner if we tried to provide consensus national champions by after the turn of the century requiring at least 2 selectors and perhaps 3 for especially contentious years of 4 or more champions. I might change the box accordingly. If it's too ugly to face you can change it back. Cake (talk) 08:22, 3 February 2016 (UTC)[reply]

I have no problem with eliminating teams that were recognized as NC by only one of the NCAA-recognized selectors. This has the benefit of eliminating those whose claim to the NC is based solely on a single, oddball selector. Listing only those teams with at least two selectors is a relatively objective and reasonable standard. However, I do see a problem with using different standards for different years -- i.e., requiring only 2 selectors in some years, but increasing the standard to 3 or 4 selectors in other years. That type of fluctuating standard, allowing us to do what you or I may think is "right" in years with multiple contenders, is far too subjective and over the line into WP:OR. Cbl62 (talk) 14:47, 3 February 2016 (UTC)[reply]

I do admit I'm trying both to apply a consistent standard and have its output look 'right'. I think the idea is also to avoid having say 5 champions with two selectors each when there's still a clear-cut champion or champions, such as in 1927 or 1919 or 1922. I am glad you are sympathetic to a quantitative as opposed to qualitative standard, for it gives a proper output IMO with e. g. 1924 or 1925. It seems like being selected by just Parke H. Davis is a lot more important before 1900. Cake (talk) 21:27, 3 February 2016 (UTC)[reply]

What leads you to believe that Davis's selections were more important before 1900 than for the years from 1900 forward. Like others, his selections were retroactive for all years prior to 1933, and he tended to be biased in favor of the Ivy League (he played for Princeton), particularly for years prior to 1920 (he selected at least one Ivy League school as the national champion or co-champion every year prior to 1920). If we are going to have a consistent standard, we'd have to have a strong basis for treating Davis's retroactive selections as more significant than others. Cbl62 (talk) 00:08, 4 February 2016 (UTC)[reply]

He seems to account for what seem like necessary co-champions such as e. g. 1869 Rutgers, but it's probably a worthy sacrifice for consistency. Berryman is often so odd that I wonder if I can exclude him. In application, it would e. g. leave off 1927 Georgia (not claimed by the school, lost to GaTech in last game. Even your Wolverines were ranked higher by Dickinson), which I think helps the credibility of the navbox, frankly. Cake (talk) 02:19, 4 February 2016 (UTC)[reply]

Year breaks

Not sure what the rationale is for the irregular year breaks in the template. Why not use simple decade breaks? Cbl62 (talk) 22:55, 13 January 2016 (UTC)[reply]

Dickinson = 1924. 1906 = forward pass. The rest is Parke H. Davis. If decades are more useful, go for it.

I do think decade breaks are easier to follow. As you have no objection, I will modify accordingly. Cbl62 (talk) 01:57, 14 January 2016 (UTC)[reply]