analyzing playoff predictions: afc

The AFC field is set:

Colts
Chargers
Patriots
Bengals
Jets
Ravens

- - -

After week 1, I had 4/6 right. I thought the Steelers were clearly in, and I projected the Jaguars as likely. The most interesting story is that I only thought 2 teams had no shot of making the playoffs after week 1: the Browns and the Bengals. If you remember week one, the Bengals lost 7-12 to the Broncos on the flukiest play ever. I figured that if you can only score 7 against the Broncos and then lose in a heart wrenching way that your season was over. Clearly that was not the case.

In week 2, I marked the Colts and Ravens as clearly in, and the Jets, Pats and Chargers as likely. That's 5/6. I also had the Steelers as likely. Bengals were still clearly out in my book.

Still 5/6 in week 3, Bengals moved up to likely out. (Similar to Bluth stock being upgraded to don't buy.)

By week 5, the only team clearly in was the Colts. I was down to 4/6 as the Broncos seemed like they had to get in.

At week 7, I thought that the Steelers were clearly in along with the Colts and Pats. Finally saw the Bengals as a playoff team. Had 4/6 missing the Chargers and Ravens.

Week 9: 4/6. The Jets and Ravens had moved down the list to the bottom of likely out. Steelers were still clearly in.

Week 11: First time had all the clearly in correct: Colts, Pats, Chargers. 4/6 with Jags and Steelers over Ravens/Jets.

Week 13: Still 4/6 without Ravens/Jets.

At week 14, 5/6 with the Broncos in, and the Jets at the bottom of the wild-card fight.

With only one week to go, I was only 5/6, backing the Steelers instead of the Jets.

- - -

The situation didn't get progressively clearer as it did in the NFC. Instead, at every point 4 playoff teams were apparent (although not the same 4 all year) and even early on, 5/6 were known.

Doing a little cumulative math:

After 2 weeks, we know 8/12 NFL playoff teams.

After 7 weeks, we know 9/12.

After 11 week, we know 10/12.

This surprises me a little. This tells me that right out of the gate, we know more than we think we do. Conventional wisdom is that two weeks is too early to judge your season, but not really. After 2 weeks, I identified 9 teams as out of the playoff race, and only 1 of those made it. I picked the 12 I thought would make it, and 8 of them did. And I picked 11 that I thought likely wouldn't make it, but had a shot, 3 of those teams did.

So of those groupings, this was my error rate: 1/9 (11%), 4/12 (33%), 3/11 (27%).

Another way of looking at, is that after 2 weeks, I could be 67% confident of a team making the playoffs (8/12), but I could be 80% confident of a team not making the playoffs (16/20). With 80 percent certainty I could have told 20 teams to stop caring about the NFL after just two games.

I won't chart each team's status but basically the Bengals were the one surprise team this year, the Jets and Ravens were the winners of the wild-card fight, and the Broncos and Steelers collapsed pretty hard.

I think the Jets are really a 7-7 team that doesn't belong there. They got to play the Colts and Bengals in their last two games, teams that had already clinched their playoff spots. If I were the Colts or Chargers I would much rather play the Jets than the Titans, Steelers or even the Texans. But now the Jets are in, and have momentum and have just of good of shot of advancing.

- - -

I won't ever do this again for the whole season, because it's fairly tedious and you don't know the point until the season's over. But it does tell me that two weeks is not too early to make snap judgments, and that mid-way through the season, you pretty much know how a team will finish. (The Titans were an exception this year, but they still couldn't make the playoffs, so what's the point?)