For the past 20 Thursdays, I have posted AccuScore reports on games involving AFC East clubs.

The videos generate interesting responses. Some critics condemn AccuScore for getting a prediction wrong, but forecasting the winner and loser aren't what make AccuScore a worthwhile tool in previewing a game.

Many readers focus too much on the predicted result and not the calculations that foretell how an underdog can win.

Did AccuScore predict the New York Jets would beat the Indianapolis Colts or New England Patriots? No, but they were absolutely correct in breaking down winning formulas for exploiting or shutting down each opponent.

So if you're the kind of Jets fan who's prone to getting offended, don't get too upset with the surface statement from AccuScore's 10,000 simulations of Sunday's game against the Pittsburgh Steelers for the AFC Championship.

The Steelers won 63 percent by an average of 4.4 points.

OK, deep breath.

Now consider some additional information from AccuScore.

The basis for Pittsburgh winning 63 percent is its run defense, which held LaDainian Tomlinson and Shonn Greene to 3 yards a carry in the simulations. When the Jets' backfield tandem combined for 100 yards, they won a whopping 74 percent of the time. They broke triple digits only 15 percent of the time in the sims, though.

In games when the Jets limited Ben Roethlisberger to no more than one touchdown and sacked him four times, they won 52 percent of the time.

Mark Sanchez completed 54 percent of his passes for 200 yards and a touchdown in AccuScore's sims, but when he completed at least 60 percent with a touchdown and no more than one interception, the game turned into a pick 'em proposition.