There are several different things to consider.
1. nonlinearity in value: in most cases the guys who finish 1/2 in most categories are going to be significantly above the trailers. So, when you calculate something like standings gain points in roto, you have to recognize that it because nonlinear at the top and bottom of the scale.
2. Dynamic value: as you and your competitors make picks, the value of each player to each of you changes, depending on how the cat totals change.
So, you may want to adjust your valuations to account for that. For example, you could top code or deflate extreme values to make sure you do not over-value the nonlinearity. Or, you may want to adjust your values as the draft progresses (you can also do this in a simple fashion, by focusing on the player, within a value tier, who addresses your weakest cats best.
But, you do want to be careful. Even if Ellsbury has little value to you, at mid season, he may be very valuable to someone else. Don't ignore that each player has market value, even if their value to you is less.
"I don't want to play golf. When I hit a ball, I want someone else to chase it."