Analytics software has Super Bowl prediction
If a computer simulation is right, it might not be a very super night for Philadelphia Eagles fans in a week’s time. At least there was the NFC Championship Game…and cheesesteaks.
Oh, and Super Bowl commercials. Always have that to look forward to.
PredictionMachine.com ran its usual simulations of the Super Bowl – 50,000 times in all – to come up with the most likely scenario. Using the Predictalator software designed by Paul Bessire, the analytics software came up with the determination that the New England Patriots win 60.7 percent of the time from that pool of 50,000 simulations.
The scoreline most typical from these 50,000 simulations is the Patriots winning 27.6 – 23.0.
Lest you scoff at the idea of a computer analytics program picking the game, the track record for PredictionMachine is very solid. Last weekend, PredictionMachine accurately picked both winners in the Championship games. Beyond recent history, all-time, the software is 57-29 in the NFL playoffs.
The founder and creator of the site, Bessire has used his background and a Masters in Qualitative Analysis from Cincinnati to help create a program that accurately provides a simulation of sporting events. Individual players are given grades as is the likelihood of certain conditions and play choices. All of which makes PredictionMachine’s infrastructure incredibly solid.
Some other interesting tidbits from the Predictalator as it pertains to some of the Super Bowl’s internals:
· Nick Foles 20-33 215 yards, 1.5 TDs, 0.8 INT
· Tom Brady 22-35 285 yards, 2.2 TDs, 0.8 INT
Carson Wentz instead of Foles:
· Patriots 27.3 - Eagles 26.9 (Patriots win 51.4%)
· Carson Wentz 21-34 248 yards, 2.1 TDs, 0.7 INT
Jimmy Garoppolo instead of Brady:
· Patriots 23.6 - Eagles 23.1 (Patriots win 52.5%)
· Jimmy Garoppolo 20-33 240 yards, 1.7 TDs, 0.9 INT
· Patriots 27.1 - Eagles 23.2 (Patriots win 58.0%)
· Tom Brady 22-35 280 yards, 2.0 TDs, 0.8 INT