The Dagger interview: Jon Wilner, Associated Press voter
In the last two editions of Poll Dancing, we've had some fun looking at the top 25 ballot of Jon Wilner. The writer for the San Jose Mercury News votes in both the football and college basketball polls for the Associated Press and has drawn much derision around the Internet for his unusual rankings.
For those unaware of Wilner's outlier-happy ballots, this week he ranked Wake Forest, Clemson and North Carolina well below their actual AP rankings and had Michigan State, Marquette and Tennessee higher than any of the other 71 voters. Obviously, there's always going to be an extreme in a panel of 72 people, but Wilner's tendency to break away from the norm has earned him the "worst voter" (dis)honor a number of times from Pollspeak.com and has made him a frequent target on college message boards.
However, for those who dismiss Wilner as an out-of-touch sportswriter "without access to the Internet" (as one recent post on an ACC board read), you may be surprised to know that Wilner does his research, and plenty of it. He also defends his ballots on his blog at mercurynews.com. And earlier this week, Wilner graciously accepted our email interview request.
While we still don't agree with many of his theories, at least we know that he's thinking about it, which is probably more than any of the coaches voting in the ESPN/USA Today poll can say.
The Dagger: How do you prepare for the weekly rankings?
Jon Wilner: I keep all my ballots on the same legal pad (there's more chance of my computer hardware crashing than me losing that legal pad). The rankings are due to the AP office first thing Monday morning, so I sit down on Sunday night and list the wins and losses for all the teams in my top-25 and then another 10-15 teams that I've got on my "consider" list. Then I go through and fill out the ballot. The football ballot is agonizing: It often takes me two or three hours; the basketball ballot doesn't take as long -- maybe 45 minutes or an hour. They're two different animals: Basketball has so many teams and so many games that it's far more fluid than football. There are times in basketball where Team A beats Team B, Team B beats team C and Team C beats Team A -- all within seven or 10 days. At some point, you have to make a subjective call; otherwise you get bogged down in comparative scores and can't complete the ballot.
TD: What is your rankings philosophy?
Wilner: I try to adhere to the same philosophy in basketball as in football: rank teams based on results, with an emphasis on strong schedules, quality wins and even quality losses. But that's tougher in basketball than football because teams can go weeks and weeks without playing quality opponents. (Another problem is defining "quality opponents." I tend to use the AP rankings as my guide, since I am voting in the AP rankings.)
TD: Your reluctance to rank Clemson higher than No. 23 stems from the team's historically poor finishes. Why is it fair to judge this Clemson team based off past versions?
Wilner: My view of that issue is completely different. As you alluded to, the AP encourages its voters to base their rankings on performance. What's a more performance-oriented system than using victories over ranked teams as your guide? And Clemson has yet to play/beat a ranked team. Ideally, I'd apply the wins-over-ranked-teams standard to everyone. But the problem there is that because of soft early schedules, it would take until Christmas to find 25 teams that have actually beaten ranked opponents. So that's where the subjectivity comes into it. You simply can't hold every team to the exact same standard every week, or you could not fill out a ballot. I've tried.
TD: Undefeated and No. 2 Wake Forest moved from No. 19 to No. 8 in your poll. You called their pre-UNC games "epically creampuff". The win against Carolina helped the Demon Deacons on your ballot, but not enough to move the team ahead of two-loss squads like Michigan State and UCLA? What was the rational for this decision?
Wilner: The AP also encourages voters to make drastic changes in the ballots from week to week if necessary. It's not uncommon for me to move teams 8-10-12 spots (in football or basketball). Wake's schedule was awful, so I had them way down the ballot. Then they beat the Heels, so I moved them up. I'm of the belief that strength of schedule and quality wins/quality losses are as important, if not more important than, won/loss record. I have no problem with a two-loss team being ahead of a one-loss (or no-loss) team if there's a difference in SOS and quality wins/quality losses. And in my opinion, Michigan State and UCLA have played much tougher schedules than Wake. Yes, Wake beat UNC which beat MSU, but that was six weeks ago. I always struggle with how long to give results vs common opponent precedence over everything else -- you can only do it for so long before it paralyzes your ballot. (That, too, has happened to me.) I digressed a little there, sorry. Back to the emphasis on SOS: I also have no problem moving teams up after they lose, if they lose to good teams. My thinking is, if you're No. 20 and you lose to No. 10, that doesn't mean you aren't the 20th-best team. In my football poll in the fall of 2007, I had two loss-LSU ahead of no-loss Ohio State because of the difference in the team's schedules. My system usually earns me the "worst voter of the week" award on Pollspeak.
TD: What are your opinions on each voter's ballot being made public?
Wilner: Absolutely, all ballots should be made public. The more transparency, the better. I post my football and basketball ballots on my blog every week. For football, I include each team's AP rank, so readers can easily compare my rankings to the overall poll.
