Faketeams makes the case that you should eschew this commonly practiced strategy.

I agree with them in the example they use - nominating Derek Jeter (assuming you're not prepared to pay the premium for his popularity and name value) because the shortstop position itself is so scarce. Their argument is that it'll be that much harder to get your real choice at shortstop once one of the few viable ones is off the board. In other words, if you're targeting Jhonny Peralta, for example, you might get him for cheaper if Jeter is still out there, as would-be bidders will have another fallback option.

But barring that somewhat uncommon scenario, I think tossing out expensive players you don't want is still a good strategy. You want those guys to go when there's still lots of money available to spend. Or, put differently, the last thing you want is for players you don't like to become bargains because then it's lose-lose for you. Either you have to price enforce and risk getting stuck with them, or watch them go to someone else for cheap.

This assumes that players go for more at the beginning of an auction, and while that's not always the case (I've seen ridiculous bidding wars for semi-scrubs by teams that waited too long to spend their money), in my experience it usually is. Still, the dilemma of having to price enforce with a player you don't want, or letting him go cheaply to someone else can come up at any time.