Please see this IMPORTANT MESSAGE from The Angry DM!

In Defense of Opportunity Attacks

May 17, 2012

Or At Least, In Defense of Something Like Them

I want to defend attacks of opportunity for a bit. But, at the same time, I don’t. The problem is that I want to defend attacks of opportunity (AoOs) for the job they do, but I’m not foolish enough to believe the answer HAS TO BE AoOs. There are other ways to do things. I’m not picky about the mechanics. But the jobs that AoOs do need to be handled by something.

Additionally, the moment you bring up AoOs and you call them AoOs, everyone immediately thinks of the AoO rules from certain previous editions. Those are specific implementations and some feel they have had problems, but let us not forget that BECMI had AoOs too. If you moved away from an adjacent opponent, the opponent got to take a whack at you. So, while I want to defend AoOs, I don’t want to defend the specific forms of AoOs. I just want to discuss what they do for the game, why they are needed, and how we might be able to get rid of them if we must.


First of all, if the game is going to have any mechanics that give a tactical advantage for holding a specific position on the battlefield, AoOs create a risk or cost to obtain that advantage. Flanking bonus, higher ground, cover, concealment, terrain bonuses: all of these things that have existed in past editions give creatures an advantage for occupying a specific position. Some of them are fixed, like a piece of terrain that provides a certain bonus. Others are relative, like flanking bonuses. But they all benefit someone who moves to the right place at the right time.

AoOs provide a potential cost or risk to obtain those bonuses. They ensure that creatures – heroes and monsters – sometimes have to weigh the risk of obtaining the bonus against the need for the bonus. Yes, a flanking bonus is nice, but if the combattant has to risk an AoO to get there, they have to make a strategic decision. Without the risk/cost, there is no decision. Without a limiting factor like the risk of an AoO, every time the opportunity presents itself for that bonus, anyone who doesn’t take it is stupid. Simple as that. And for something like flanking, this can result in a bonus that is “always on.” When the monsters outnumber the heroes, the monsters will almost always have a +2 attack bonus against someone. When the heroes outnumber the monsters, the heroes always have that bonus.

What’s wrong with that? Well, technically nothing as long as the bonus is fairly trivial. If the bonus starts to involve things like lots of extra damage, it starts to become something that needs to be assumed in combat balance. But more than that, it raises the question of why you bother having a rule that specifies how to flank. If it is trivial for two heroes to flank a monster, why bother with the requirement of being on opposite sides. Its just wasted rules. Just call it ganging up: if multiple creatures are adjacent to a single enemy, they get a +2 to attacks. Or, hell, call it “outnumbered.” If either side in a combat encounter has a numerical advantage, they can specify one enemy against which they have a +2 to attack.

The same can be said for cover. If there is cover available on the battlefield, any ranged combattant should have cover against ranged attacks from their enemies. If you can freely move around the battlefield, there is no reason for ranged combattants to EVER not be in cover. So, again, you can simplify the cover rules immensely by just saying “if there is cover available on the battlefield, every ranged combatant gets a +2 to AC against nonadjacent opponents.” Again, it becomes something that really should be “always on” because only a moron would choose not to have cover.

The dirty little secret about AoOs is that they are actually what make all those positional advantages meaningful. Without something that limits and controls movement on the battlefield, the bonuses and penalties for position become far more complicated than they need to be. You don’t need complex flanking and cover rules, for example, if flanking and cover are so easy to acquire that you can assume that people are going to have them almost all of the time. Just add them in as “always there” and call it close enough.

AoOs also give you a way to emphasize certain characters’ mobility. The rogue, the opportunist who relies on flanking, has mobility powers or skills (whether it is a combination of the Tumble skill and the right feats or movement based Utility Powers) that allow him to move around the battlefield with impunity. It becomes a defining feature of the class, something special, something that is possible because its something special only they can do.


Heavily related to Job 1, AoOs also give various creatures a way to deny opponent’s positional advantage or to maintain positional advantage against a foe once they’ve got it. Once you are flanked, it can be hard to move out of flanking position. You might want that cover, but if an opponent is trying to block you, you need to deal with that opponent or find another way to get it. In truth, the ability to control the battlefield in this way is also an important part of making positional advantage meaningful by giving it a cost or a risk. In fact, it creates another form of positional advantage to consider: position control.

Some positions on the battlefield become valuable only because they allow creatures to control access to other valuable positions. The path around the difficult terrain is valuable because my AoOs allow me to stand in the middle of it and charge a toll for anyone trying not to be slowed by the difficult terrain. This goes for both heroes and monsters.

Moreover, AoOs are an important part of protecting allies and working as a team. Whatever edition of D&D you play, you know the fighter can survive more attacks than the wizard and thus, the fighter has some responsibility to keep the heat off the wizard. But if opponents can freely move wherever they wish, the fighter really can’t do that. Of course, you can make up for this by giving the fighter some ability to establish a zone of control (a mark, an aura, whatever), but that means only the classes you specify (or the people who took the feat or whatever) can protect allies. In an emergency, the rogue can’t step in and help protect an injured ally. Or an NPC the party is escorting. Or anything else. The only way to protect someone on the battlefield, in that case, is by slaughtering everything that isn’t you. And its kind of nice to have other options, isn’t it?


Imagine a world with unrestricted ranged attacks. Imagine a world where a ranged combattant with a bow can freely attack anyone in range and with line of sight. If a bow and a longsword both do 1d8 damage, what idiot is ever going into battle with a sword. Seriously. If they both do the same damage, but one can attack from anywhere while the other can only attack by putting itself in harm’s way, how stupid do you have to be to rely on melee combat?

Ranged attacks (including most magical spells) need a limiting factor to make up for the fact that you can attack almost anyone with them without endangering yourself. The ability of a melee combatant to shut down an archer or spellslinger just by standing next to him is vital to ensuring there is any melee combat at all in your game. Otherwise, from a tactical standpoint, the game becomes a game of longbowmen having shootouts across the battlefield and melee-based monsters end up mostly extinct.

And then, with free movement, it will quickly devolve into a cover-based shootout. Both sides will hunker down into cover and fire away at each other until someone is dead. At least, that’s where the smart money is. Until a wizard flushes them out with a grena… err… a fireball. If this sounds familiar, that is because it is the form combat takes in non-fantasy settings in which firearms are the primary means of making your arguments known. And I am not saying its not a fun style of play. But it also isn’t swords and sorcery because swords are notoriously difficult to make ranged attacks with from behind a chest-high wall.

Now, different games handle this differently. In some games, the rule is that you just can’t attack an adjacent opponent with a ranged attack. Or that you can’t make a ranged attack at all if a melee opponent is standing next to you. That’s fine and dandy. But if it is also trivial to move away from the opponent, you get the “backpeddle and shoot” sort of kiting combat that, amusingly enough, effectively turns ranged combat into a weird form of moving melee.

But those rules leave me personally unsatisfied. Look, I get that you can’t raise a longbow and shoot a guy standing in front of you, dancing around, and waving a sword at you… but I’d kind of like to have the choice to take a shot at someone else, even if I am going to get stabbed to do it. I’d like to be able to trust my defensive armor, spells, and temporary hit points to get me through the warrior’s attack so I can shoot the evil wizard and disrupt his spell before he blasts my buddies.

And in the end, that’s what its all about: making choices. If something is free to acquire (like a flanking bonus in an AoO free world), acquiring isn’t a choice. It’s just a correct answer. If something is forbidden (like shooting one opponent while another one is gnawing on my face), not doing it isn’t a choice. It’s just what the rules say. But when something is available but risky or costly, it is always a choice. Do I take it? Do I risk it? Do I live without it? And choice is what puts the RP in RPG.


Now, in the past, I have heard people speak out vehemently against AoOs because they slow down the game, they break the flow, they lead to overly complicated rules, they encourage players to micromanage their movement or become paralyzed with indecision. Now, the overly complicated rules bit is one we can just ignore. That’s an implementation problem. The same can be said for breaking the flow. There is no reason that AoOs have to break the flow of the game. I don’t honestly see the flow problem in this:

  • DM: “The goblin starts to move away… giving you an opportunity to strike. Make an opportunity attack.”
  • PC: “I swing my longsword. 16 AC for 9 damage.”
  • DM: “You slash the goblin as he backpeddles. Blood sprays from the wound, but he recovers and retreats behind the rocks. He nocks an arrow and fires. 13 AC?”
  • PC: “Miss.”
  • DM: “But the arrow goes wide of its mark, the goblin distracted by his bleeding wound.”

But let’s talk about the slowdown and analysis paralysis problems. The problem here is not that you have some system of AoOs in place. The problem is that we have a game in which free choice is a major component and tactical positioning is a source of major advantages. People have to choose between low-risk low-reward and high-risk high-reward and people are generally risk averse. People have to count squares and pick positions. Those two things combine to really terrify some people who are afraid of getting it wrong. Now, you can remove the choice aspect. As I noted above, a freely available bonus is a non-choice situation. A forbidden action is also a non-choice situation. But I would argue that that is also stripping away some of what makes the game a  role-playing game. So the alternative is to make position less of an issue to begin with.

And honestly, that’s the other half of my argument. Everything I’ve said above immediately falls apart if you decide to remove the tactical movement/tactical position element from the game. Make it not matter. Dump flanking. Lose terrain bonuses. Simplify the hell out of cover. It works. And it can be a lot of fun. The AGE system (Dragon Age) really isn’t designed to give much of a crap about your position. Warhammer Fantasy RPG (3rd Edition) has a really neat, simple way to handle positions and ranges that downplay the tactical positioning element. Any game designed primarily for gridless, theater of the mind style combat works just fine without position being a major factor.

And, to be absolutely clear, I have NOTHING against that style of play. Personally, I love it. I love it just as much as I love gridded minutae combat. If you want to take out tactical positioning and a precise grid, I’ll still happily play the game. I honestly have no preference on that front.

But if you want to allow position to have a big role in the game – if you want terrain to be meaningful and you want to create classes and abilities focussed around positioning and mobility – you are making a trade off. You’re adding the “count the squares and be careful” component right up front. And in that case, as I’ve said above, you need a system in place that does the jobs that AoOs do. You’ve already chosen to slow down the game and risk some tough decisions. The best you can do is make sure your AoO system is nice and simple.


The key is to make AoOs simple to remember and easy resolve. Don’t make them complex. Don’t define a thousand actions and specify which four hundred of them provoke AoOs (like 3.5 did). Don’t turn AoOs into an opportunity action and allow a whole bunch of stuff to replace your AoO (like 4E did). Don’t limit the number, but lift the limit for certain feats (like 3.5 did) so you have to keep track of who made an AoO and who didn’t from round to round. Don’t make AoOs compete with a dozen other forms of “outside of the turn order actions” like interrupts and reactions (like 4E did). Make it f$&%ing simple!

This is a good one: if you are ever adjacent to one or more opponents who can make melee attacks and you do anything other than make a melee attack against an adjacent opponent, all of those opponents get to make their melee attack against you for free. In short: if you end up next to me and my buddies, you’d better hit one of us or we’re going to hit you. The moment you make that melee attack, you’re in “defending myself in melee mode” and anything else you do that round is safe. But if you don’t make that melee attack, you’re open to attack.

Its easy to keep in your head (though I admit my wording might be clumsy), it has no weird exceptions, and it doesn’t care what else you do. If you are engaged with adjacent foes, make a melee attack to lead your action or get hurt. That’s the basic rule everyone has to know. Then, you can worry about putting exceptions in feats, skills, or whatever like Tumble or Combat Casting or Positioning Trick. And keep those exceptions down to a manageable number too. Melee should be a quaqmire that is hard to get out of.