The word "prediction" seems a little serious for what amounts to educated guesswork, but here goes:
Philadelphia at St. Louis, 1 p.m. ET: The Eagles are ahead of the Rams in their development, but not enough to takes this game for granted. The Rams' defensive line matches up well. Steve Spagnuolo's knowledge of the Eagles will help. I'm tempted to pick St. Louis here, but I don't see the Rams scoring enough points in the end. Philadelphia averaged 36 points per game on the road last season with Michael Vick starting at quarterback, never scoring fewer than 26. The Rams allowed 17 or fewer in six of eight home games, but Atlanta (34 points) and Kansas City (27) fared better in the Edward Jones Dome late in the year. My best guess: Eagles 24, Rams 20.
Carolina at Arizona, 4:15 p.m. ET: The pressure falls squarely on the Cardinals to win their home opener against the NFL's worst team from last season. Without Kevin Kolb at quarterback, the Cardinals managed only 12 points against the Panthers during a horrid 2010 defeat at Carolina. They acquired Kolb to avoid suffering through similar performances. Kolb, the Eagles' starter entering last season, says he learned a lesson about putting too much pressure on himself heading into an opener. Time to show it. My best guess: Cardinals 21, Panthers 13.
Seattle at San Francisco, 4: 15 p.m. ET: Both offenses figure to have their problems adapting to new schemes and, in Seattle's case, new personnel early in the season. The Seahawks look like a team that should stop the run effectively, helping them put the game in Alex Smith's hands. But as the 2010 opener between these teams showed, there's no use in projecting exactly how so many offseason changes might play out in Week 1. The 49ers should be favored because they're at home and Seattle has more question marks on its offensive line. Simple as that. My best guess: 49ers 20, Seahawks 16.
At least one and probably two NFC West teams will emerge from Week 1 undefeated. We'll take it. And if you haven't made your predictions yet, please do so here.