Mark CMG said:
So, two out of three falls recommends less prep work?
I recommend not doing every last bit of prep work and only do the prep work that you are likely to need. Look, If your fights last 6 rounds, if you prep 12-rounds worth of material you've left yourself plenty of options.
If I'm prepping a caster I'll have 12 combat-use spells (including escape and detection), any persistent daily buffs, and the buffs the NPC will have ready and are likely to be able to use. Why do I need to spec out the other 2-dozen spells? After bluff, intimidate, sense motive, spot, listen, search and concentration, do I really need to spend the remaining skill points?
Most melees are pretty simple, excepting Bo9S characters. Work top down, choosing the 12 most powerful maneuvers they qualify for and just assume they have the pre-reqs. It can pretty much always be true.
At this level you know the party and their 3 most likely approaches. Figure out the NPCs likely reactions in a general sense (Caster erects forcewall, archers spiderclimb & shoot, swordsage shadowblinks & attacks mage, etc) so you aren't blindsided. yes, they will occassionally catch you completely offguard. Big deal, they might catch the BBEG offguard too since the BBEG won't have seen them in action as much as you (the DM) have. Start with the plan that works best and roll from there.
Simply reacting in a decisive and coordinated, if suboptimal, fashion is often enough to provide the apperance of competence to players. Heck, the players won't know it's suboptimal since they can't be entirely sure what the NPCs are capable of. The PCs will never know what spells the NPC had already cast or prepped in memory or what plans they might have had.
An old military saw is "Perfection is the enemy of the good." If you've got an imperfect, but workable, plan go with it rather than dithering around trying to make it perfect. It gives the other side time to prepare or, in this case, get bored.