Your Distract maneuver forces them to make the AoO if you succeed, though. What about just letting them take one if they want it? Does that even require a maneuver? If not, why not? Shouldn't you just be able to just stand there?
Sure. That's equivalent to taking the Distract maneuver and then deliberately failing your deception check.
Can you do this to varying degrees? Helpless if you just stand there and take it (Ben Kenobi style), but also just enough to let them attack, but then defend yourself? If not, why not?
Sure, but I see no reason to complicate the rules by allowing more than a few degrees of freedom. Presumably, a person could not perfectly judge how much he was protecting himself anyway.
I think it would come up with my PCs. They'd finish off downed foes without worrying about wasting actions.
Downed foes most of the time are finished off after the battle ends anyway. If they are downed, most of the time they aren't getting up again. I think it would punish the PCs more than anything, as they are the ones more likely to recover from being downed and whose survival would be more threatened if the goblins or whatever could attack them while still otherwise engaged in combat.
This is interesting to me. In 3.X, I'm not sure you should really need to really spend a feat on this.
The feat allows you to make a single additional attack at your highest BAB during your turn, regardless of the action you are taking (Withdraw, Double Move, Full Defense, Charge, etc.) against any undefended object provided you have an unused opportunity. It's very much intended to provide for a character who wants to do swashbuckling sort of stunts. I'd hitherto never considered making it just something anyone could do, perhaps because it feels like something that would require a bit of skillfulness when I imagine it. However, I could just integrate it into the combat engine. If I did, I'd probably make the standard maneuver inflict an AC penalty and have the feat eliminate that penalty. However, that would probably make the feat too weak, so I'd need to then bundle it with some related ideas.
This is if you hold a particular spell, sure. But if I don't do anything on my turn, is there a reason I wouldn't be able to just kinda pick the spell I want when it comes up? I don't mean rules-wise (since I'm talking about revising those), I just mean from a "common sense" / game balance approach.
Despite the fact that 3e doesn't track a spells casting time, I've always assumed that it still takes 2-4 seconds to prepare and cast a spell and therefore the decision to cast the spell has to be made a 'considerable time' before it actually goes off. I don't see using a (non-quickened) spell as a reflex sort of thing. I might allow an AoO to be made with spells that can be cast as a free action. Again, this is in my opinion a common sense/game balance approach.
This is a pretty good point. Again, though, what if you have an archer who hasn't fired in the round?
Then he's holding or delaying his action. If he isn't then the concrete reality the abstract rules are trying to capture is that his last arrow was fired moments before the beginning of his (abstracted) turn (an event with no in game reality) and he's been spending the moment since then preparing the next arrow he's going to fire.
Or has, but has a high base attack and feats or class features that allow more attacks? Should you be able to do something then?
Sure, but this situation now represents a case of unusual and uncommon skill in preparing an arrow, selecting a targeting, aiming and firing. In my rules unusual and uncommon skills are represented as feats. So if I wanted to allow a ranged attack as an AoO, I might have a feat like 'Snap Shot', that a) gave you a 'virtual' threat zone and b) allowed you to fire an arrow as an AoO at some cost. Common sense to me suggests that cost is a deduction from the maximum number of attacks you may make during your next own action. I would never allow a ranged attacker to increase his maximum rate of fire by making AoOs. That doesn't make sense to me from a 'common sense' perspective and from a game balance perspective it would move the game in the direction of my experience with early 3.X, which was, that the best party strategy was for everyone to become a ranged attack specialist.
There are two issues going on here. First, it's obvious that a turn based system cannot accurately capture the reality of a real time combat. We do have a mechanic that we could use to much more accurately simulate that reality and that would be dividing the round into a large number of segments or impulses, each representing a fraction of a second. I could devise such rules pretty easily and they'd result in highly cinematic combats with very high degrees of verisimilitude, however, given the already high burdens of adjudicating combat in 3.X adding this additional complication to the combat in my judgment would make the game play to slow.
The other question I see being raised by my answers is 'what is a feat, and what is a maneuver'? In my mind, if it is something I can imagine a five your old doing, then it is a combat maneuver open to everyone (though with varying degrees of success depending on skill). If it is not something that I can imagine a five year old doing, it suggests to me that it is a combat feat - something that certainly unskilled people can't do and which not even every competent person could do. This is I admit not entirely satisfactory. For very high levels of skill you'd expect a person to at least be able to attempt most feats, even if they weren't quite as good at them as someone that had been especially training to do them. However, it seems to work well enough from both a common sense and game balance approach. What it does mean though is that things like 'Distract' or 'Throw' or 'Aggressive Stance' are maneuvers in my game (like Trip, Grapple, Disarm, etc.), not things you get as feats, because they are things I can imagine a 5 year old doing in a play ground fight versus his 5 year old nemesis. This requires me to not use a lot of published 3e feats because often they aren't written with this theory in mind.
One of the biggest problems I see with a lot of 3e published rules is they make the incorrect assumption that because a Feat is a short bit of text that it is easy to design a good one. In fact designing Feats is really really hard. Getting my feats right has been a huge challenge. Ten more years of play testing and I may be happy with them.