The NFL season is still more than a month away, and we’re kind of running out of things to talk about. I mean, as long as guys are still driving drunk we’ll never truly run out of things to talk about, but as the season grows closer, we start to run out of worthwhile things to talk about. No matter how many season previews we read, and scouting reports we scour over in a vain attempt to predict the winner in a volatile sport, we can never truly know what will happen until the teams start actually playing games.

Sure, we can make incredibly accurate guesses based on the information we are privy to, and some have even made a career out of making astute predictions based on their football expertise. These people are usually called “football analysts”, “sports handicappers” and “gamblers that don’t live in their car”. Sometimes we use computers to calculate different variables to predict the next Super Bowl champion, but how many season simulations would it take to make an accurate prediction? The Predictalator is hoping the answer to that question is 50,000.

The people over at have simulated the 2012 NFL regular season 50,000 times, and before I get to the results, here’s how it works.

Just as the Predictalator plays each individual NFL game 50,000 times before it’s actually played, it can also play the entire NFL season 50,000 times before it’s actually played. The actual 2012 regular season and playoff schedules have been played 50,000 times, allowing us to compute average records and likelihoods of each team making the playoffs, winning their division and bringing home the Super Bowl trophy.

The Predictalator’s technology is built to handle such a demanding task. Making sure that we have the best possible inputs for players’ statistics, progression over time and age, roles, health and playing time as well as teams’ coaching styles and weather are keys to the accuracy of the output.

In general, we apply strength-of-schedule-adjusted, relevant statistics from every player’s careers (weighted more heavily on the most recent 16 games) to a fairly traditional player development curve that considers age and previous playing time. Not only does this development curve help to set average inputs, it combines with health history to dictate the variance (“boom or bust” potential) of a player’s inputs.

We might be heading into overkill territory when we start simulating each game 50,000 times. I understand the concept of collecting as much data as possible, but in some team’s cases, one injury could derail an entire season. Just ask last year’s Indianapolis Colts. The Detroit Lions can’t just restart the simulation if Calvin Johnson blows his ACL while running his first route of the season. No matter how many games are played hypothetically, games can only really be played once.

Oh wait, did you actually want to know who the Predictalator is hailing as the next Super-Bowl champion? Fans in Philadelphia, get ready to bust out the moderately priced champagne and start planning the parade.

The Philadelphia Eagles turn their “Dream Team” into Super Bowl champions, winning it all league-high 18.0% of the time. In the most likely Super Bowl, Philadelphia defeats the New England Patriots 54.3% of the time and by an average score of 28-26. New England, the AFC’s top-seed, wins the Super Bowl 14.2% of the time. The Green Bay Packers (9.2%), Houston Texans (9.2%) and Pittsburgh Steelers (7.2%) follow the Eagles and Patriots in Super Bowl likelihood.

The Predictalator has spoken, and the Dream Team that also happens to be a dynasty will claim the Lombardi Trophy in February. If I were Michael Vick, I’d start getting my ring finger sized. Hell, I’d be getting them all sized to save some time when his dynasty team inevitably wins at least six more championships.