FiveThirtyFake

FiveThirtyFake uses statistical analysis--hard numbers--to tell soft stories.

Pre-Game: Preliminary polling of the players as they exited their locker rooms show many were motivated by dislike of losing.

Start of game: Our model forecasts that Clemson will probably win the game. There's a 91.2% chance. By no means does that mean Pitt has no path to victory lights, however. 91.2% is barely beyond a toss-up, which is in the 70's if you pick your cherries right.

13:08, 1Q. Clemson's chances have fallen to 84.4% as Pitt scores first. Probably could be an exciting game.

2Q. The referees are not rigged. There is some "not heard mentality" going on, however.

HT:  28-27 Clemson and the "Tigers" have a 81.1% chance of winning, slightly better than toss-up and with roughly equal chances that there will be a Clemson blowout or a Pitt Blewitt.

9:00 left in the 3Q  Clemson's chances are up to 90.8% Clemson probably finished Pitt off in that third debate.

Early in the 4Q  Clemson has an eight-point lead and a 90% chance of victory according to our model. Clemson will probably win.

Mid-4Q: The referees are not rigged.

5:00 4Q: The referees are not rigged.

1:11 4Q Clemson clinging to a two point lead in the polls, 42-40. No need to worry too, too much, though if you're a "Tiger" fan, there is only a minute and change remaining. Through 20,000 game simulations on our model the team with a two-point advantage with a minute+ remaining won 99.9% of the time. But it is by no means over.

And now with :41 to play all of a sudden PITT--off a 21-yard pass play to the Clemson 34--PITT has a 51.6% chance of winning. It's a toss-up! How exciting is this! In the span of 20 seconds Clemson's chances have fallen from 99.9% to 48.4%. Hard numbers don't lie!

At 1:02 am after the polls have been closed for five hours FiveThirtyFake is proud to announce in an empty Death Valley with the lights cut out and the scoreboard black our model has once again correctly forecasted the shocking 43-42 Pitt upset.

Be sure to tune in to our podcast with Harry, Micah and Clare as we dissect the numbers and explain why our model gave Pitt a better chance of winning than did others, e.g. those at The New York Times.