Seahawks: A Tale of Two Super Bowl Linear Trend Failures

12 seconds into the game, I opted to do some simple back of envelope extrapolation. The score was 2 to 0 and the game was 1/300th complete (barring overtime).<br /><br /><b>Expected Final Score: 600 to 0</b><br /><br />25 seconds into the game, I saw the error of my folly. My model was breaking down! We were clearly losing momentum!<br /><br />At halftime, I revised my estimate using a much larger data set. The score was 22 to 0 and the game was 50% complete.<br /><br /><b>Expected Final Score: 44 to 0</b><br /><br />Imagine my surprise when Denver scored a touchdown. Nobody could have expected them to score, on a percentage basis, infinitely more points in the second half than in the first half! Black swan event!<br /><br /><b>Final Score: 43 to 8</b><br /><br />Two linear trend failures and we still won? We sure dodged a bullet up here in Seattle. Nail biter to the very end! Whew!<br /><br /><div style="color: red;"><span style="font-size: large;"><b>Go Hawks! :)</b></span></div>

Seahawks: A Tale of Two Super Bowl Linear Trend Failures

12 seconds into the game, I opted to do some simple back of envelope extrapolation. The score was 2 to 0 and the game was 1/300th complete (...

Read more »