The week for Super Bowl prognostication is upon us, and that includes the digital world, too.

With Super Bowl LIII between the New England Patriots and Los Angeles Rams now less than a week away, everyone is starting to form their opinion of how the game will play out. And has been the case for years now, EA Sports is using its “Madden NFL” video game series to predict the winner in a computer simulation.

The pick is in, and the virtual landscape is calling for an upset, with the “Madden” simulation predicting a 27-24 win for the Rams.

If you want to beat Tom Brady, you’ve got to be able to hit him and hit him hard, which is exactly what Aaron Donald did in this virtual Super Bowl. The Rams star pass-rusher collected four sacks en route to Super Bowl MVP honors.

Coincidentally, the Super Bowl sim played out similar to the AFC Championship, only the Patriots weren’t so fortunate this time around. After jumping out for a 17-3 halftime lead, the Patriots apparently had no answer for the high-powered Rams offense in the second half. Unlike the Kansas City Chiefs before them, LA was able to complete the comeback with a late touchdown from Todd Gurley giving them the Lombardi Trophy.

So, what does it all mean? Absolutely nothing, obviously; it’s a video game. But if you’re a Patriots fan looking for some good mojo, you can probably take solace in the fact that the “Madden” simulation predicted a Patriots win last year only for Nick Foles and the Philadelphia Eagles to prove that wrong by winning Super Bowl LII.

Since the Super Bowl simulations began nearly a decade ago, the predictions were correct five of nine times. However,  the “Madden” simulation has correctly predicted the Patriots’ Super Bowl outcomes in three of four trips since the simulations began.

So, who knows?