The premise is simple: I'll tell you what players you can confidently start the season with, and what players should cause your stomach to wobble if you have to start them.
2006 stats at NFL.com and pro-football-reference.com
Matt Hasselbeck (QB): People don't get too excited about Hasselbeck, but in the last four seasons he's averaged over 200 yards a game and around 1.5 TDs a game. He's not one of the top fantasy QBs (Manning, Bulger, Brady, Palmer, Brees, McNabb) , but he's among the best of the rest of the bunch. If you don't prioritize getting one of the top fantasy QBs (which is reasonable--to get one of them, you have to pass on a good starting RB or WR), then a QB like Hasselbeck is who you should target: steady, consistent, predictable, but not flashy.
Shaun Alexander (RB): From 2001 to 2003, he was one of the underrated stud fantasy producers. In 2004 and 2005, he emerged as one of the clear top fantasy producers. After a decline in 2006, should we consider Alexander on the decline? I'm not sure. I'd be nervous to draft him early (or expensive), but at the same time, he's a clear feature back with a history of production, so you have to be happy with him.
Here's what I do like though: draft Alexander and Hasselbeck and start both of them. Seattle is usually a pretty good offense, and if you start their starting QB and their feature RB, you are going to get a lot of their points.
Josh Brown (K): Seattle's kicker is one of the legitimate starting kickers.
In a deep league, or in a league starting three WRs, you could start Deion Branch or D. J. Hackett. I've never been a fan of Deion Branch as a fantasy WR. Pretend you don't know that he had two spectacular Super Bowl performances, don't look at the name, and just look a the numbers he's put up season by season in his career. Do the numbers you see point to a good fantasy WR? They don't even point to a mediocre fantasy WR. I'd rather start D. J. Hackett: we have enough evidence available to show that Branch is a poor fantasy producer, but Hackett only has two seasons and seems to have the ubiquitously described "upside." Normally, I prefer proven production to potential; however, when the proven production is sub-par*, I'd rather go with potential.
*I don't really get the term "sub-par" in common usage. We use it to refer to something that is below average, but in golf, you want to be sup-par. Par is average, but shouldn't something bad be considered "above par," and something good be considered "below par"? But I like the term "sub-par": it's the most concise phrase to describe something below average. I'm just always self-conscious when I use it (self-conscious enough to use an asterisk, anyway).