Most RPGs use turn-based combat because it can provide a simple method for clearly resolving the chaotic realities of the battlefield. (Simultaneous action resolution, for example, can work really well with very small numbers of combatants, but then breaks down rapidly as the number of combatants increases.)
Turn-based combat, however, creates mechanical oddities: If you were in a swordfight with someone and they were like, “Hang on a sec. I’m just going to grab a pint and have a quick drink,” you’d just stab their dumb ass. But because we’re using an abstract mechanical structure in which everyone resolves their actions one at a time — even though, in reality, everything is happening simultaneously — suddenly you’re just supposed to stand there watching me take my drink because it’s not your turn.
To deal with this, we add off-turn mechanics that allow characters to react to things that they should, logically, be able to react to, even if it isn’t their turn. In D&D 5th Edition, these mechanics include the Ready action, reactions, and opportunity attacks.
We can add one more level to this by adding mechanics that allow you to, for example, avoid opportunity attacks. We might want to do this because a character is super-skilled at drinking in the middle of combat, or maybe just because it’s a bad-ass moment. (The Disengage action in D&D 5th Edition is technically one example of this. In D&D 3rd Edition you could make Tumbling checks to avoid attacks of opportunity from movement and Concentration checks to avoid provoking while casting a spell.)
D&D 3rd Edition sought to implement a lot of mechanics from previous editions of the game in ways that were both more consistent and comprehensive. This included inventing the term “attacks of opportunity” and classifying which actions would “provoke an attack of opportunity.” This made sense, but in practice it had a major drawback: It created a huge list that filled nearly two full pages of actions detailing which actions did (and did not) trigger attacks of opportunity, and you either had to reference that list constantly during play or you had to memorize it. Even if most of this list boiled down to common sense, the result was still Byzantine and arcane.
As a result, after 3rd Edition, there was a practical impulse to avoid this “grand and unwieldy list” of actions. In D&D 5th Edition, this simplification has been taken to an extreme: An opportunity attack is triggered only when a combatant you can see moves out of your reach, unless they take the Disengage action.
This eliminates the complexity of the list by boiling the mechanic down to a single trigger, but it allows a lot of immersion-breaking shenanigans on the battlefield. And even the implementation of the movement-based trigger is kinda wonky: You can literally run circles around an opponent while firing arrows at someone on the opposite side of the battlefield, but you can’t walk past them while swinging your sword at them. And, bizarrely, this means that creatures with longer reach are actually less effective at attacking people around them.
In my opinion, if you wanted to simplify opportunity attacks, it would be preferable to either (a) eliminate the mechanic entirely (can’t get simpler that that!); or (b) go the other direction.
HARDCORE OPPORTUNITY ATTACKS
By default, any action you take provokes an opportunity attack from any combatant who can reach you.
There are three exceptions:
- Any attack action
- Dodge
- Disengage
In addition, Ready doesn’t provoke, but the action you’re readying may provoke when you take it.
Bonus actions, reactions (other than readied actions), and your free object interaction never provoke opportunity attacks.
Movement provokes opportunity attacks normally.
OPTION: BETTER MOVEMENT OAs
When using this option, movement provokes an opportunity attack whenever a character moves more than 5 ft. within your reach on their turn or moves out of your reach.
Note: Disengage still cancels all movement-based opportunity attacks during your turn. You also don’t provoke an opportunity attack when you teleport or when someone or something moves you without using your movement, action, or reaction.
OPTION: DAMAGE SPELLS
Spellcasters can use their bonus action to avoid the opportunity attack triggered by casting a spell if the spell deals damage.
Note: This rule may be useful to make some 5th Edition spells work properly. It avoids needing to make all of those spells special case exceptions. Our goal remains One Rule to Rule Them All.
OPTION: HARDCORE RANGED ATTACKS
Using this option, only melee attack actions avoid opportunity attacks. Ranged attacks still provoke.
Note: As with a spellcaster, you might allow a ranged attacker to use their bonus action to negate the opportunity attack.
DESIGN NOTES
Positioning in combat should matter. You want your rogue to pick the lock on the door so you can all escape? Cover their back! You want your spellcaster to rain hellfire down on your foes? Form a defensive line and give them the space they need to do it.
By allowing characters to just do whatever wherever, the 5th Edition opportunity attack rules cheapen positioning.
Beyond making the battlefield a more interesting space for tactical challenges, I also want things to make sense: If you’re going to do something that isn’t directly focused on fighting, I want you to think to yourself, “Should I really be doing this where the guy with the pointy metal can stab me?”
At the same time, we want to avoid a bunch of Byzantine complexity. We don’t want a big list of what does and does not trigger an opportunity attack: What we want is a simple, straightforward rule that we can easily memorize and apply. We want One Rule to Rule Them All. By flipping things around and listing the very small list of things that DON’T provoke, we achieve that goal.
From a practical point of view, if we end up in a situation where this One Rule to Rule Them All doesn’t make sense, it’s much easier for me as a DM for me to “break” the rules with a ruling that’s more permissive to the PCs than one that isn’t. In other words, saying, “Actually, it would make a lot of sense if this thing you’re doing that would normally provoke doesn’t provoke in this situation,” the players will be much happier accepting that than if I have to say, “Actually, this thing you thought wouldn’t be bad for you is actually going to be bad for you.” A restrictive framing, therefore, can paradoxically give us a greater liberty to make bespoke rulings when and if they’re needed.
ADVANCED DESIGN NOTES
Taking a step back from opportunity attacks, there are two different broad approaches to modeling the idea that you can’t just run willy-nilly around a battlefield or do a crocheting project in the middle of a melee without consequences.
First, you can try to mechanically enforce it: Thou shalt not.
Thou shalt not move past someone threatening you with a melee weapon. Thou shalt not drink a potion if someone has marked you as their target. Thou shalt not run through an area under the effects of suppressive fire.
(My house rules for combat in 1974 D&D, where the procedure effectively makes melee “sticky,” is another approach to this.)
You can also mitigate this approach: Thou shalt not X, unless Y.
For example: “You can move through a threatened space if you succeed on an Acrobatics check.” Or, “If you get hit with an opportunity attack, you have to stop moving.” (If you can avoid the attack, then you can ignore the Thou Shalt Not.)
The other approach is that you can impose a cost for doing it — e.g., “You can do X, but you’ll suffer a penalty.” Or risk getting hit by an extra attack. Or lose an action.
This approach gives more flexibility: If you really want or need to do something, you can still do it. You just have to pay the cost.
And, once again, you can add conditionals that allow characters to mitigate or entirely avoid these costs.
Of course, the highest the cost becomes, the greater your need or desire would need to be to endure it. Conversely, the more negligible the cost becomes, the less influence it will have over the characters’ decisions.
If you think about opportunity attacks within this design paradigm, 5th Edition’s opportunity attacks are clearly aiming for the second method, and my argument is that they have become so trivial that you would be better off either (a) eliminating them entirely in order to streamline combat and encourage even more movement on the battlefield or (b) using the hardcore opportunity attacks house rules (or something like them) to make them actually matter.
Alternatively, you could abandon that paradigm entirely and maybe try to implement something of the Thou Shalt Not variety.
I agree that opportunity attacks in 5e are on awkward footing. Cutting them out completely does give problems with 2-3 popular feats: Sentinel, Polearm master, and Warcaster (this one is debatable). You could scrap Sentinel, ignore the minute Warcaster nerf, and just play Polearm Master as the only thing giving you an OA. I’m just not sure that I love that (or that I didn’t forget something).
If you go the other route and give out more opportunity attacks, you will run out of reactions very quickly. This already happens with the above mentioned feats, and also at higher level with some class features and reaction spells. So either you add more reactions in (or only for OA?) or you will have the same dissonance due to the many things you can do in close combat, just a turn later in each round.
I’ve become engrossed with PF2e of late, and rather like its approach to positioning and attacks of opportunity. Lots of things trigger them (though thanks to action tags it’s now much easier to parse at a glance than in 3.x), but most of your party and your opponents won’t have them. Combat becomes less sticky and rewards mobility more, and being able to attack someone if they’re in your reach and not taking you seriously becomes a special feature of fighter-ish classes. Go ahead, drink that potion in front of me. See what happens.
However, this means that most characters aren’t able to punish opponents for taking those kinds of non-combat actions while under threat. But since things like drinking a potion is pretty action-intensive (one to take it out, another to drink, unless you are already holding it in a hand, limiting what you can do and just begging an enemy rogue to pinch it) I think the tradeoff works out.
I’ve been toying with this idea myself. I like the idea of melee engagement being the definition of control, not requiring special features or special attacks to do so, especially within a classic D&D/OSR context. To use an expression from the total war pc games, melee combat should be a tar pit. Fight, run, or die. This gives power back to fighters innately.
The gamist sense of implementing this where there limited legal moves you can take, in melee you must attack/defend, or withdraw/retreat, the end, should be acceptable, but as we know the simulationist ethos, “you can -try- to do anything,” is almost sacred. From 3e onward, it’s systematized such that we get opportunity attacks that are a slap on the wrist/easily avoided and maneuvers that are penalized so heavily that they’re a waste of time (unless you invest in the feats/combos).
I’m leaning more toward high lethality opportunity attacks or noncombat actions that can be interrupted in some way like spellcasting.
Our groups has been playing with “OPTION: BETTER MOVEMENT OAs” for years, adding to that the variant DMG Flanking rules. I attest to this being a very good addition that naturally enriched our games tactically for little complexity.
I’ve gone the opposite direction.
Opportunity attacks in D&D are problematic for me. One of my hobbies is fencing with a side interest in Hema. One thing that’s noticeable in 5e for me as opposed to a ‘real’ group fight is how static characters are once they enter melee. In the real fight people are continually shifting position, retreating and disengaging.
In my D&D (actually Dungeon Crawl Classics) games I’ve binned opportunity attacks completely. This frees up the battle field wonderfully and makes the fight much more dynamic and lifelike. I rule that you can only take your action after moving so you can’t attack and then run away which doesn’t really happen in a real fight.
Interesting point about characters not having a melee weapon being at a disadvantage. Hadn’t thought about that because they are sitting ducks. In real life even if you’ve got a weapon but it isn’t pointing directly at your opponent then you’re still a sitting duck (e.g. a 2 on one fight). Having played a bit of 2nd edition Runequest this models that much better as parrying without a weapon is a realistically bad idea and you only get a single parry per round in any case. There is no dodge in 2nd ed Runequest with only a small defence modifier for agile characters. I think that’s about right. Maybe give advantage in D&D if you want to model that?
In order in emphasise the primacy of melee weapons up close I rule that you can’t move and shoot or cast a spell. Well you can but you’re at disadvantage. (In DCC you have to roll to cast a spell)
Interresting… In 1EAD&D it seems like the disengage is action intensive and gives attacks of opportunity on a sort of both individual and group level in a weird way. Mages can cast a spell and move 3″ (as opposed to what seems to be 12″ of movement so long as a meleeist doesn’t remove themselves from the combat? I’ll need to review this)
House rules wise we had 12″ or Attack or split attacks/movements (so a 2/1 would be 1/1 with 6″ of movement)
but that could be very wrong.
Also Holmes which refers to 1EAD&D a LOT seems to have some additional rules like charging that aren’t seemingly in 1EAD&D that I recall but the DMG is pretty obtuse.
all of this is IIRC of course but how costly should it be to withdraw to drink anything is a good question, but I stand by my Paladins of Dionysus being able to swig beer all fight long…
In Rolemaster there are things called wrist vials that in game solve the problem of drinking a potion while in combat. Thusly if the rules get more restrictive about drinking potions in combat there are science-based workarounds anyone can use if they just combine a little thought and imagination.
D&D combat (especially in 5e) is so divorced from a simulation of ‘real’ combat that it might not worth be getting into the weeds about.
The idea of ‘hardcore opportunity attacks’, where you get attacked if you’re doing anything other than fighting or retreating, is a step in the right direction if you’re aiming for verisimilitude, but I don’t know what it would do to the overall game balance.
Realistically, if you’re in a fight and you spend six seconds doing anything other than fighting, whether that’s picking a lock or drinking a potion or casting a spell or whatever, you’re going to be hit – not attacked, hit – a bunch of extra times by every opponent. And a rule encoding that would destroy 5e’s gameplay loop.
I’m not sure how you’d implement this without breaking something else. I don’t want combat to be completely locked down – as Giles said, it’s desirable and logical for melee combatants to be able to move around. I feel like this punishes melee more than anything because your average caster has means of locking down melee monsters via summons or control spells, but your average melee PC has no real means of avoiding them. An option is to say “you must have a melee weapon to avoid OAs” but that is trivially easy to ignore with quarterstaves. You could say “you must have a MARTIAL melee weapon,” but this excludes spears and maces for no compelling reason while encouraging the infamous wizard with a fighter or cleric dip.
GURPS has those 1 second rounds where you can do almost nothing, but add some prep rounds and you can do anything. As a result, they need no interruption mechanics, as anything interruptible just takes more than 1 second, and your opponent has normal turns before you can finish, which, he might make a quick jab, which is probably defined in one of their endless option books.
You can do the same thing with 6-second rounds, by having spellcasting a thing where you start your spell at the end of the 1st round, and finish it at the start of the 2nd, and then start another at the end of that 2nd round as well. You still get 1 spell per round, but people can interrupt it, without any extra mechanics. Someone can see you starting to cast a spell, and shoot you on their turn. Could even have fancy saves or whatever to have the spell work anyway, if desired. Or spells like Magic Missile that require no prep and always hit so are great for disrupting other casters, unless they have Shield up.
And the same for potions and and scrolls and such, anything that’s interruptible, takes a move action to start, and then a standard action to complete on the following round.
Even for running from a fight. Round 1: move action, disengage, you step just out of reach of opponents, allowing you to move out of combat as a standard action next round. Round 2, standard action to run a double move away, or a single move around corners. And between, you can have -2 AC while disengaging or something like that, and problem solved. You are otherwise in melee until the other side is all disabled.
@tussock: I’m not quite seeing how the disengage part of that isn’t just “this is impossible unless they let you go,” but the spells one is really intriguing!
I’ve often thought I’d like D&D-esque spellcasting batter if casting a spell took two (maybe sometimes even more) rounds, but there are some obvious implications there in terms of both balance and bad feels. But, as you say, this would keep the action economy pretty close to the same, whilst not letting people just stop to chant and wave their hands around (with gestures so intricate that a heavy coat would interfere) in the middle of a sword fight. Very cool.