Football: Did the fans vote differently?

This year, in addition to the CIS-endorsed and UFRC-run Top 10 poll, our own Mike Radoslav organized a Fans Top 10 polling panel, which he described as a group of "fans, ex-players and a few members of the media representing each conference." Assuming there was little overlap in membership in the Fans and UFRC groups, this was an interesting experiment, if nothing else, because it allowed for additional voices in the weekly top 10 discussion.

I was curious, however, to see how different their opinions were from the media members' poll (see here). Maybe the fans would be more emotional, overreacting to wins and losses, but rational in their detachedness from any one team (other than their own); while others would naturally favour their home team/conference over another, but bring greater inside knowledge to the process.

Or not. Turns out, there's not much difference at all.

First, let's look at the top overall vote-getters from the Fans. These are the eight teams with the most votes across all weeks of the season. Including anyone else means we have teams like Laurier, who had one or two good weeks in the voting, but were otherwise not a factor, and therefore not very interesting from a voting point of view. (Click for bigger versions of all charts.)

(Percentage of all votes, or vote share, is just the number of points for that team divided by all voting points from that week. Using vote share helps us compare two polls with a different number of voters. The thin red lines represent a theoretical "consensus #1" team and the same cutoff for the consensus top five.)

Looking at the more noticeable drops in each team's vote share, you can see the (likely) effects of Western's loss to McMaster, Laval's loss to Montreal, Saskatchewan's loss to Alberta, and Montreal's two late-season losses. St. F-X also climbed steadily until their loss to Bishop's (at home, even!).

Interestingly, it seems to take several wins for fans to “forgive” a loss (like Saskatchewan and SMU) and losses near the end of the season are very damaging to the perception of a team. Anyone who's surprised by that has never met a sports fan.

What about the UFRC results? If you've been following along, you might have already guessed the answer:

If you think I just copied the chart from above and changed "Fans" to "UFRC"...well, you'd be wrong. But I don't blame you: the results really are that similar. And if one was being critical, one could say these polls capture nothing more than wins and losses...which are already captured pretty well in the win-loss record.

But maybe early-season polls can shed some insight on the differences between the fans and the media. Let's add up the vote totals through Sept. 9 (the "early share"), and see if we learn anything. Here are the top five early-season vote-getters, along with their vote share in the final vote (Nov. 3):

FANS
Laval: 18.2% early share to 18.2% final share (no change)
Western: 15.6% to 9.9% (-5.7)
Saskatchewan: 12.8% to 14.3% (+1.5)
Calgary: 10.6% to 15.2% (+4.6)
Saint Mary's: 10.3% to 7.3% (-3.0)

Western down, SMU down, Calgary up 4.6.

Okay, and the UFRC? Well, they had the same top five teams in their first few votes, so not much is different:

UFRC
Laval: 18.2% to 18.4% (+0.2)
Western: 16.6% to 10.0% (-6.5)
Saskatchewan: 12.9% to 13.6% (+0.8)
Calgary: 11.6% to 16.3% (+4.6)
Saint Mary's: 11.5% to 9.8% (-1.7)

Western down, SMU down, Calgary up 4.6. Wait, that sounds familiar.

(Between you and me, I'm running out of ways to differentiate these polls.)

Let's do one more thing: take the eight teams from above and chart the difference between the Fans' vote share and the UFRC's vote share.

...Actually, let's not bother with the chart, because the only noticeable differences in vote share (and I'm using the word "noticeable" loosely here) are St. F-X in Week 5 (the Fans had them 2.5 points lower) and Saint Mary's in Week 11 (Fans also lower by 2.5). Neither difference was that significant, though, and having just four teams in the conference might explain variations in AUS voting. But these differences really are just too small to matter.

So, to recap: the Fans voting and the UFRC voting start out the same, follow the same trends, reward wins and losses in the same way, have the same opinions on each team as the season progresses, and give the same result in the end. (This is hardly surprising to Mike, who said, "Having compiled the lists, they always seemed pretty similar.")

Which means either, “the fans know just as much as the people paid to cover the league,” or, more generously, “nobody knows anything more than anyone else.” Let's go with the second one.