I can't go back to the fully fuzzy method as we don't have factors we throw into a formula any more. For instance, in the 2.0 engine, when checking for a pass success it would determine a base result like 60% and then adjust that for the team match up and then the player match up. So you end up with a 58% chance of completing the pass. The only thing the defensive play tendency affected was that it adjusted the team ratings going in to that calculation - combined with the defensive formation. So for a 3-4 Balanced defense, the team ratings would be adjusted by 1.125. For a 3-4 Heavy Pass defense, the team ratings would be adjusted by 1.25 and 3-4 Heavy Run would be 0.9. As you can imagine, this doesn't really give you much within a single play. Over 1000 plays, sure, you will see the effect, but within one play that 0.9 or 1.25 really doesn't amount to much. It basically might adjust the chance of complete +/- 2-3% or so.
With the 3.0 engine, I'm trying to give a more pronounced effect on each play. This means instead of adjusting things by 1.125 or 0.9, that I actually can mold the entire chance of different results on the call. For instance, when checking to see how well covered a player is, I can shift the advantages and therefore the range of results up or down based on Rush defense or Pass defense. Basically I can code things to know it is a pass defense or a rush defense and I don't have to rely on those adjustment factors.
I do understand the issue as you might want to have a more balanced approach to your defense, but I think game-wise the either/or approach makes more sense than a 50/50 meaning that you are essentially just reacting and not committing to either pass or rush defense. I think the defense should have to make a call one way or the other. I'm not even sure what it would mean to say you are 70% rush defense and 30% pass defense. Theoretically I understand it, but execution-wise I'm not sure how that would play out.
The real problem, I think, is really the way we call plays in the first place. In real life, a defense might have things that the offense are doing that they can key on to look one way or the other, but they would probably always have something they expect for each play. This certainly deserves more discussion, but I don't want to get back to the 70/30 type fuzzy effects unless I can come up with a good way to work it in like the current rush or pass setting fits in.