Tip Top 25 in helmets, smaller
                                                    Home

College Football Researchers Association

The College Football Researchers Association was founded in 1982 by Anthony Kusher and Bob Kirlin. They selected their mythical national champion (MNC) each season based on a top 10 vote of their membership, which was published in their monthly newsletter. The CFRA wasn't around long, and their last champion was named just 10 years later, in 1992. But in addition to selecting contemporary champions, they retroactively selected champions for 1919-1981, based on a poll of members conducted by Harry Frye.

The College Football Researchers Association was resurrected in the Fall of 2009, in name at least, and the website of the current incarnation can be found here. I will discuss the new CFRA at the end of this article, but here's a spoiler: it is quite lame.

The College Football Data Warehouse uses the CFRA's selections, in addition to those of Helms and the National Championship Foundation, for national championships prior to the AP poll's 1936 launch. You have to wonder why they don't also use them for national championships after 1935, when their criteria remain the same.

The answer, of course, is that even more than the National Championship Foundation's picks, too many of the CFRA's modern day "champions" are simply ridiculous. That begs the question, why then should we trust their picks for seasons prior to 1936?

CFRA Selections 1919-1972

The CFRA's selections for 1919-1935 are adequate-- little better or worse, overall, than those of the NCF and Helms. But the CFRA was greatly hampered, particularly compared to the NCF, by their lack of ties, a side effect of their selection process (which was similar to the AP poll's). Their selection of Alabama as the lone "champion" of 1926 looks quite poor compared to Helms and the NCF's splitting of the title between 9-0-1 Alabama and 10-0-1 Stanford, since the two teams tied in the Rose Bowl. And the CFRA is the only non-computer organization that proclaims 10-0 Alabama the lone "champion" of 1930 over 10-0 Notre Dame. A shared title would have been far better.

Like the NCF and Helms, the CFRA's selections for 1936-1967 are much better than those of the AP poll, since the CFRA counted bowl games. However, while the CFRA did a decent job selecting retroactive champions through 1972, they did an awful job selecting champions for seasons they had actually witnessed, and in fact they are easily the worst selector, aside from computer rankings, listed in the NCAA Records Book for 1973-1992, after which they mercifully stopped bothering.

CFRA Selections 1973-1992

The CFRA's selections for this 20 year period resemble the selections of a computer, and it looks like they voted for the team they thought was the best, regardless of wins and losses. That, of course, is a power rating, and while interesting, it is a bad way to crown a national champion, and they certainly were not taking that approach in their selections for 1919-1972.

For the 1973 "title," the CFRA retroactively selected 10-0-1 Oklahoma over consensus choice 11-0 Notre Dame. Now, an unusual choice is not necessarily a bad one. As I discussed when fixing the 1973 AP poll, the National Championship Foundation split that year's title 3 ways amongst Notre Dame, 10-0-1 Ohio State, and 10-0-1 Michigan (OSU and Michigan tied each other), but while this choice was unusual, it was also astute and logically valid, whereas the CFRA's selection of Oklahoma was not. Oklahoma may well have been the best team in the country in 1973, but the fact is that they were tied by Southern Cal, whom both Notre Dame and Ohio State defeated by more than a touchdown.

For 1976, the CFRA selected 11-1 Southern Cal over 12-0 Pittsburgh, and as with their selection of Oklahoma '73, they were the only non-computer system to do so. Southern Cal lost to 6-5 Missouri 46-25. They defeated Notre Dame at home 17-13, whom Pitt beat 31-10 on the road.

For 1977, the CFRA split the title between Alabama and Notre Dame, and for 1978 they named Alabama the lone "champion." You can read my opinions on those poor choices here (1977) and here (1978).

But before you come to the conclusion that the CFRA just had a crush on Alabama, take a look at 1979: every other organization, human and computer, unanimously selected 12-0 Alabama, but the CFRA selected 11-0-1 Southern Cal instead. So they weren't biased so much as they were incompetent.

And for 1980, the CFRA made it a record-breaking (and mind-boggling) five bad picks in a row, choosing 11-1 Pittsburgh the "champion" over 12-0 Georgia, again the only non-computer system to do so.

The CFRA crowned Oklahoma in 1986 (Oklahoma lost to Miami, who lost to 12-0 Penn State), and they made the same weak choices the AP poll made in 1984 and 1991. The CFRA's cousin, the NCF, was far from perfect, but their selections 1973-1992 were vastly superior to the CFRA's, and they did better a job for each of the poorly selected seasons I just listed for the CFRA.

Conclusion: The CFRA is not considered an authoritative selector outside the College Football Data Warehouse.They are about as good a selector as the NCF and Helms for 1919-1972, but they were so awful 1973-1992 that in the end, I think they are probably the worst selection organization listed in the NCAA Records Book (other than computer rankings). Their competition for that dubious distinction is Parke Davis.

The Current "College Football Researchers Association"

In the Fall of 2009, a blog emerged with the title "Official Site of the College Football Researchers Association Poll." From the About page of the site: "Last September, with the coordination of myself and others and with the permission and involvement of several past members, the College Football Researchers Association was reorganized, and a group of both new and original CFRA voters was assembled to recreate this influential poll."

None of these people are named, and the person writing the blog only refers to himself as "me" and "myself," but judging from his e-mail address, his name is B. Matthews. As for the site itself, there is little or no actual college football "research" going on over there, nor even much interest in it. It appears to be nothing more than a fan poll that has co-opted the CFRA name since its old picks are listed in the NCAA Records Book. From the blog's first post: "The CFRA poll will be voted on each week by me and other college football fans deemed worthy enough to participate. ... I've always thought it would be cool to come up with a rankings system that complied [sic] the votes of friends and other fans."

So perhaps the venture would be better titled "My College Football Friends and Fans Poll."

In any case, he and his friends have been voting in a weekly poll since September 2009, when the poll launched with 6 voters. That number has since exploded to 17 voters. In addition, they have gone back and voted for "CFRA champions" for 1993-2008, including the rather dubious selection of Michigan as a stand-alone "champion" of 1997. If you check out the "CFRA Almanac" portion of the site, you will find "CFRA national champions" selected for every season from 1869 to the present. However, since the NCAA Records Book only lists the original CFRA's selections for 1919-1992, those are the only ones I will be listing in my articles about each year's national championship.

Speaking of 1869, check out the first sentences ever written on this blog: "Way back in college football's inaugural season of 1869 two polls were formed that enabled a group of fans to vote on a national champion. One of those polls, the College Football Researchers Association Poll... emerged as the most preeminent poll in college football until the Associated Press rankings took its place in 1936." So when he co-opted the name, he apparently didn't even know that the CFRA was formed in 1982 (not 1869), nor that they did not exist prior to the AP poll's arrival in 1936, nor that they have never been considered in any way "preeminent."

Conclusion: quite lame. The current CFRA bears no resemblance whatsoever to the old one, is not in fact a "college football researchers association" at all, its website is poorly written, and many of the statistics listed on their "Champion Stats" page are not even close to correct.

But if you're interested in voting in their poll, they're still taking members. Or if you want to run your own organization that's listed in the NCAA Records Book, the Helms Athletic Foundation and National Championship Foundation are also both defunct, so you could co-opt one of  their names for your own blog, and voila!-- you're "listed" in the NCAA Records Book.

Home