Superbowl XLVI Projection, Recap

Congratulations to the New York Giants.

Let's review their odds of being the Super Bowl winner from each post this season:

Week 10-11.15%
Week 21-11.89%
Week 32-13.64%
Week 43-13.81%
Week 53-21.66%
Week 64-22.17%
Week 74-21.68%
Week 85-22.11%
Week 96-22.49%
Week 106-31.477%
Week 11, Thu6-31.479%
Week 11, Sun6-40.728%
Week 116-40.715%
Week 12, Thu6-40.604%
Week 12, Sun6-40.497%
Week 126-50.187%
Week 13, Sun6-60.082%
Week 136-60.082%
Week 14, Thu6-60.080%
Week 14, Sun7-60.166%
Week 147-60.166%
Week 15, Thu7-60.153%
Week 15, Sun7-70.030%
Week 157-70.027%
Week 16, Thu7-70.027%
Week 16, Sat8-70.111%
Week 16, Sun8-70.103%
Week 168-70.098%
Week 179-70.370%

It's rather amazing that I had them at under a 1 in 1000 chance after week 16, and under 1 in 250 even after the season ended and they were in the playoffs. That is, it's amazing if you trust in my system, which I sort of do. Week 1's prediction of Baltimore over San Francisco came so close to reality, and I would have looked like a genius, or at least extraordinarily lucky. But, of course, these are only projections, and teams can always beat the odds. As they say, anything can happen on any given Sunday.

Now that the NFL season is over I may have to track down another project to combine sports and stats, as I sort of miss seeing my new projections each week and then having them demolished the next. My options include March Madness. Now it would be a Herculean effort to try to predict the field of 68, since I'd have to simulate the season, seed the teams in each conference, figure out how many enter each conference's postseason tournament, and figure out which conferences get automatic bids for the regular season and/or conference tournaments. Of course the seeding process is legendarily very manual, as apparently the committee tries to keep teams close to home and even adjusts their seeds by up to 1 position to do so. There are also rules against rematches from the season AND the previous tournament in the first two rounds, as well as a rule against intraconference match-ups until the 4th round, unless the conference has 9 teams in, and probably rules about full moons and infield flies as well. I'd probably just go with a 1-68 seeding, and after handing the first 4 games, just pit 1 vs 64, 2 vs 63, etc, to get a "close enough" result.

There's the NBA, but Saint Louis doesn't have a team, and I haven't really cared since the days of Jordan in Chicago, so, maybe down the road, but not the very next item.

For those of you hockey fans out there, the Blues are making quite the resurgence, so the NHL might be next on my list.

Finally, MLB would seem to be a natural fit, and the tiebreaker scenarios are much easier than the NFL (play another game!). If I can get the time to put the full 2,430 game season into a file for my simulations, I just might do this. Be warned, as the season goes into September, I'll be tempted to update daily as the races progress.

I hope some of you enjoy the slight mathiness I throw out there from time to time, at least when it shows your team with a chance, or when your team has a chance to show me they're better than my math

No comments:

Post a Comment