 |
| Patrick Willis: the most important player on the BQB's Super Bowl pick |
Also at long last, the culmination of our offseason work: our 2012 season predictions. We took all the relevant info we looked at during the season, put it together, and voila -- projected records for our 32 teams. Here's how we did it:
1. Expected wins: We started with a team's expected wins from 2011, as determined by Pythagorean expectation.
2. New coaches get a bump: Since teams with new coaches typically see a bump of 1.5 wins the following year, we added 1.5 wins to the 7 teams who changed coaches this year.
3. Draft: We rated every team's draft, normalized those rankings to a 0 to 100 scale, and added or subtracted wins based on how they drafted (up to a cap of +/- 2, so the best team gained 2 wins and the worst team lost 2 wins).
4. Offseason: We assessed offseason movement (non-rookies) for each team (AFC and NFC), normalized those, and added or subtracted wins based on where they fell (the cap here was +/- 4)
5. Preseason: The preseason barely matters, but it seems to matter a little when teams go undefeated or winless; if a team went undefeated, we added a win, if it went winless, we subtracted one.
6. Schedule: We took the result of the first 5 steps, used it to run teams head-to-head (home advantage = 1 game), then took the weighted average of:
- The original projected wins # (75%) and
- Number of wins when you run the schedule (25%)
We then normalized the records so that, in total, teams would go 256-256 (since every game must be won or lost).
Disclaimer: We realize this isn't the be-all and end-all of projections, and that they're fairly quantitative and not at all subjective. Having said that, we got some reasonable results in spots and, perhaps most notably, have 6 repeat playoff teams and 6 new playoff teams, about the average over the past 10 years.
After the jump, our ratings, predictions, and week 1 picks!
