If you want to model it (and let's be honest, that's gonna be hard), the side on offense has to pre-plot their movement, and at each encounter they'd have to roll on a chart that basically says - 'engage, stooge around, or withdraw'. And the defenders would do the same, except their withdrawl would be to random locations, based on how the commander thinks it should happen.
Defenders would have already planned on where to go in case of a fighting withdrawl. While you CAN say that a defender would engage only if forces were roughly equal, you also have to allow for the human condition of a commander thinking he's seeing a window of opportunity or something along those lines. It's how command works in reality - the local commander will make decisions based on his standing orders as well as his analysis of the situation in front of him. Heck, some might go blindly charging into terrible odds because they have to die for the Imperium or something silly like that.
You'd also have to set up nodal command points and plot out, for each fleet, what they were to do for X turns before information could flow back to the command staff and they could make a decision and get the commands back to the fleets. Pincer movements taking advantage of new information can easily blow up in your face if the enemy retreats or reinforces a location and changes the expected odds, not to mention you could waste weeks of preparation by emerging in a system w/o any real enemy presence to crush.
Most people really don't get the communication lag concept today since we really don't have it. There is a Starfire novel called Insurrection where the information about an insurrection on the frontier first shows up back on Terra. As usual someone says "we'll crush them with XX", and then someone points out that the information is already 4 months old. So there's gonna be a lot of random rolling and commanders will say "let's reinforce XX fleet" and that part of the line of battle gets reinforcements which may, or may not, turn the tide of battle.
Your attacker is obviously going to have it easier in planning, as they will set out their goals in advance with plans for fleets to link up or reduce nodal fleets/planets. Defenders have to react w/o much planning and fall back along pre-determined routes - but as defenders they typically enjoy certain advantages as well. Like all battles, eventually the success of the attacker means their lines of communication and supply get extended and the defenders get shorter, so things can easily flip at certain battles.
I think because you have so much randomness tossed in a strategic war of this kind is not very fun. If you played some of the early computer games like SSI's Star Fleet game (i forget the name), but you plotted out 32 impulses of play, and you put in about when you expected to fire your weapons, which meant you were looking into the crystal ball thinking where the enemy is going to be in XX turns. At the end you got to see where your ship was in relation to the enemy and then you repeated it. It wasn't a very fun game, though it was a closer simulator of having to think ahead and make decisions like you are wanting to do for your FFW game.
