Ted Lange alluded to Adam Harstad eventually chiming in based upon an article he wrote last year, but here is the gist:(
The object of each FF game of yours is NOT to score the most possible points, it is strictly to score more points than your opponent that week.
The example I believe he used was this (and remember, this was based on where they were playing and performances last year): Going into Monday night you and your opponent are tied and he only has Brees left. You have a flex spot left and can choose between Khiry Ribinson and Jimmy Graham. If your object is to score the most points, you go with Graham, since he is projected higher than Robinson. But since the goal is to win the game, not score the most points, the correct move is to start Robinson since unlike Graham, not all of Robinson potential points are directly linked to your opponent's player.
As for your specific situation, here is the general rule: if the players are close in projects points, pick the player linked to your opponent's player if you are the favorite, because this reduces variance. The opposite is true if you are the underdog - pick players not linked to your opponent as to increase variance, and thus, your chances of pulling off the upset.
Edit: Walking Boot beat me to a lot of it while I was typing.
Sorry, I missed the thread this weekend. Anyway, past seasons' articles are always available for free for non-subscribers. Which is particularly handy for me, since I'm mostly writing strategy pieces and very little of it is time-sensitive. If anyone wants to read the article in question,
it's right here.
I do agree with the general consensus. Prior to the week kicking off, unless you are a *severe* favorite or underdog, it's probably pointless to worry about "neutralizing" anyone. There's far too much uncertainty for it to be worth doing anything other than just trying to maximize your points.
After the Sunday afternoon slate has been played, there's usually a good time for reflection on the scoreboard. At that point, with 80-90% of each team's weekly score decided and only one or two players left to go, it's a lot easier to identify situations where reducing variance will make a difference.
In that case, your goal should be simplifying the story of your comeback, (or making the story of the other owner's comeback more complicated). If you have a big lead going into Monday night, and the opposing player is starting Cam Newton, and you have a choice between starting Indy's defense or Carolina's defense, and your defensive scoring is a heavy performance-based scoring where defenses could even possibly go negative... you should probably be starting Carolina's defense.
If you start Indy, the path to a comeback becomes much simpler. The other owner needs Cam Newton to have a great game and Indy's defense to have a terrible game. But a great game by Newton and a terrible game by Indy's defense are the same thing.
If you start Carolina, the other owner needs a great game by Cam Newton and a bad game by Carolina's defense, which is two different things. It's like forcing him to hit on a parlay, and there's a reason why parlays pay out so well; it's much harder to hit them, even when the odds of each individual component are pretty favorable.