No matter what you say or what you think, this year's Super Bowl is basically an even match-up. The New England Patriots and the Seattle Seahawks were each the best team in their conference this season, and —"Deflategate" accusations notwithstanding—either team could easily take home the Lombardi Trophy.
With two strong teams, there isn't a clear favorite or underdog, and there won't be an "upset" regardless of who wins. That makes predicting the winner a harder task than usual. That, of course, hasn't stopped professional sports data sites from offering their analyses and forecasts handicapping the outcome.
Ed Feng, who runs The Power Rank, is one of those sports number crunchers. He believes the Seahawks should be favored by 1.3 points, which is barely any spread at all. Feng's methodology is figured by calculating a number of different rankings using margin of victory in past games, yards per play and other statistics.
These rankings combine to give a team a final overall ranking, which leads to a rating number and how many points by which it would beat the average team. Comparing rating numbers between two teams translates to a predicted margin of victory, which is how Feng calculates Seattle to be a favorite by a razor-thin margin.
Read MorePatriots president on Tom Brady's sickness
But Feng is an enlightened sports prognosticator, aware that an aggregation of predictions will be better than any one person's forecast. There are even scientific papers and books written about this topic. So for an even better set of betting predictions, Feng put himself up against eight of his colleagues and competitors in the sports data industry.
Seahawks, but barely 'a coin toss'
The beauty and complexity of this approach is that Feng's competitors look at the game in a variety of ways. It's not clear which is the "best" method, though it is clear they all end up near the same result. None of the picks stand out wildly from the group.
The chart shows that the range of forecasts is very tight. The ensemble forecast puts Seattle ahead by less than half a point, 0.46 points to be exact. That gives the Seahawks a 51.3 percent chance of victory, "essentially a coin toss," says Feng.
Aggregating the other sports prediction methods
- Advanced Football Analytics—Brian Burke uses yards per play and offensive turnover rates in ranking teams.
- Numberfire—Rankings based on an expected points analysis of every play.
- Football Outsiders—Aaron Schatz uses the idea of success rate on every play in his DVOA rankings.
- Inpredictable—Mike Beuoy takes market data and uses regression to rank teams.
- Prediction Machine—Paul Bessire simulates the game 50,000 to come up with a prediction.
- Sagarin Pure Points—Jeff Sagarin, who developed his ranking algorithm back in the '80s, uses margin of victory in games.
- Massey Peabody—Cade Massey and Rufus Peabody use football play by play data and weight recent games more.
- Line—the point spread from the markets, a pick right now.
Notice the chart above doesn't specify which provider gave which estimate, as a way of protecting their intellectual property. Some of these sites sell their data to their clients, who can use that information to make more informed bets.
"Last year, the line and computers were on opposite sides of the game," Feng says. While the line favored Denver, the main computer models, such as Feng's, all favored Seattle, which won 43-8. "There's more consensus this year."
So what it comes down to right now: Nobody really knows who is going to win. Anybody who tells you otherwise is just hoping, praying, stretching and guessing. All the numbers point to a close game. The only upset will be if we see a blowout.
That's why they play the game.
Disclosure: CNBC's sister company NBC Sports broadcasts the Super Bowl.