Why a playoff?
Though I'm a supporter of the revolution, I understand that most leagues continue to use a head-t0-head format for a whole host of reasons.
I will never understand, however, why fantasy football leagues use a playoff.
In real life, a playoff makes sense. You play a regular season, and at the end of that you make the best teams "play off" against each other to determine who is the best.
But that's not what happens in fantasy football. There's no fundamental difference for your team between a regular season fantasy matchup and a playoff fantasy matchup.
Essentially, a fantasy football playoff says "We are going to select two weeks at random and let that determine the outcome of our league." If you have a great team but have a mediocre week, you lose. If you have a just-good-enough-to-make-the-playoffs team and your team has a great week, you win. But there's nothing special about the playoffs--you really are just selecting two weeks late in the year to determine your champion.
It's foolish and illogical. It's not a good way to determine a champion--it just rewards having a team that performs well in select relatively random (at least to the players on your team) weeks.
Going to the bench
Simply drafting the best team is not enough in fantasy football; you have to manage your bench.
It is very likely that for a few weeks, one of your top picks could be putting up mediocre points. During this same period of time, one of your later round picks, selected to be a backup, could be putting up great fantasy scores. You need to determine if or when to replace your highly touted early pick with a high performing later pick. You don't want to abandon a stud early: sometimes players struggle in their first few games and then go back to their high performing ways (Barry Sanders in 1997, Ladanian Tomlinson in 2003). But you don't want to leave a real stud on the bench for too long just because you drafted him later and he's not as well known as your starter: after all, this year's late round steals are next year's first round picks.
This is a bit of a gut decision, with a lot of luck. I, for example, have Roy Williams backing up Reggie Wayne and Steve Smith at WR: I might need to decide whether to replace either Wayne or Smith with Williams, or to sub Williams in for my second RB. I also have Vince Young backing up Donovan McNabb: while McNabb is a dominant fantasy performer, I'm still going to be leaving a lot of rushing yards and touchdowns on the bench.
I also think that you're in trouble if you switch your lineup a lot from week-to-week. If you stick with a primary starting lineup, you're going to get good weeks and bad weeks from your players. To switch up your lineups often relies a great deal on luck: you're gambling on picking the starters that will have a good single week, while you chould then just as easily miss out on good single weeks from the players you've just benched. At the end of the year, you might hit on more hot weeks than cold weeks, but you also might hit on more cold weeks than hot weeks. But if you draft decent starters, you're usually better off just riding the starting lineup, because you have to hope you drafted players good enough to have more hot weeks than cold ones. You just need to decide when one of those starters needs to be replaced with a different long-term starter.
Drafting Peyton Manning in a snake and auction
On Sunday I made the argument that in the first round, after what you feel the sure thing elite RBs are gone, you should select Peyton Manning rather than a RB you are less sure of. We'll see if this strategy works out.
This is my argument for a snake draft only. In a snake draft, you are stuck in a slot, you don't have much control over who will fall to you in that slot, and you want to make sure you get good production from the pick.
This does not mean I recommend using a huge chunk of money on Manning in an auction. In an auction, you have more control over which players you really want: rather than getting stuck with a slot pick, you can target the players you really want (so you can decide whether to pay up for the elite RBs or pick out the better budgeted RBs you think will perform well). Also in an auction, you might be able to get a QB that puts up 90% of Manning's numbers for 50% of the money. Spending a lot of money on Manning is a budgetary decision with a lot of ramifications. For example, if you draft Manning in a snake draft, you are choosing Manning over one other starter; in an auction draft, Manning might cost as much as two or more other starters (indeed, in the Hazelweird draft Manning went for 75, while I got Donovan McNabb for 35 and Reggie Wayne for 40). But there are other ramifications that depend largely on your draft.
So yes, I do think that if you don't get one of the top 3-6 RBs, you should draft Peyton Manning in a snake draft. And honestly, Manning is so good that targeting him in an auction could pay off, too: he's always very good, and in 2004 his 49 TD passes helped KS to a Hazelweird Title. But in an auction, I'm not as confident that spending a large part of your budget on Manning is the wise move.