Actions and simultaneous activity

16 posts / 0 new
Last post
One of the big paradoxes of 3E is that, while actions are resolved sequentially, a round's activity is simultaneous. The ambiguity is strong enough that a thread escalated to over 350 posts in the DM Rules Counseling forum, so I'm taking these issues here, in the hopes of clarifying everything in time for 4E.

Now, there is but one way to consolidate a sequential order of resolution and sequential activity: asynchronization and liberal overlap. In other words, while actions don't necessarily start at the same time, your actions will usually start while another's are taking place.

Simple enough? It gets more complicated.

We found that in 3E, on his turn, a combatant performs his entire round's worth of actions. Assuming that these actions take place over an entire round, distributed more or less evenly, this can mean one of three things:

1) The round in question refers to the upcoming round. In other words, if your turn comes up on initiative 15, the actions you're performing take place between that count of 15, and the same count in the next round.

2) The round in question refers to an ongoing round. In other words, if your turn comes up on initiative 15, this initiative could represent a mid-point of your actions.

3) The round in question refers to the last round. In other words, if your turn comes up on initiative 15, the actions took place in the round that ends now.

Now, many people might intuitively think that the first option seems right. I think that this perspective is actually the wrong one. If your action starts on your turn and ends one round later, some inconsistencies can occur. For example, let's say that I start attacking a target on my turn. This attack will take place over the coming round. Immediately after my turn, my opponent decides to disarm me. Since both of our actions are overlapping, it stands to reason that his attempts to disarm me would render my attack more difficult. In other words, his action, despite being resolved after my own, has the potential to affect the outcome of mine.

For a similar example, consider someone casting a spell as his turn begins could be interrupted by someone else with an attack of opportunity. However, if this happens at the beginning of the spellcasting, nothing's preventing the caster from using the remainder of his round's time to cast a spell. After all, if I'm interrupted only 1 second after I started acting, I still have 5 to do whatever I want, and this could include casting another spell, from a realistic point of view.

This is where the first option becomes counter-intuitive. The way the game is played, my attack will be affected in no way by my opponent's attempt to disarm me. If I hit, I hit. If I miss, I miss. Once I'm done, my opponent gets to see if he disarms me. Similarly, it seems inconsistent that my actions could be interrupted early on to the point of preventing me from acting the rest of the round.

The second option offers a more balanced point of view in that we can try to separate each of your actions when it comes to resolution. For example, you could say that your move action takes place before your turn, and your standard action takes place after your turn, within a time window of one round. However, it still presents the likelihood that your latter action be affected by the former action of someone acting after you.

The third option, however, seems most appropriate. If your actions represent the end of your activity, once your actions are complete, so too is your activity. As it becomes an element of the past, nothing in the future can change the outcome of whatever you have completed.

I believe this illustrates better the idea of acting at the same time as your opponents without being synchronized with their own actions. If your turn takes place after your opponent's, your reaction to his actions will come later. In my example above, I start attacking my opponent before he realizes he's better off disarming me. While he proceeds to disarm me, I'm already finishing my attack, and his late reaction will cost him whatever my attack will entail.

In order to prevent later actions from modifying the outcome of earlier actions, the earlier actions must be over and done before the later actions are resolved. If we take to the premise that your actions take place over the course of a round, the only way to make sure this happens is by having your turn represent the end of your actions.

In a lot of ways, this makes a lot of sense. After all, when you take your turn, you decide on your character's actions based on what's been going on. You know exactly what's been happening in the last round and the actual outcome of all actions that took place. However, you haven't been waiting idly by while things were going on. You were gradually reacting to the actions that ended before yours in order to make your action choices. The actions you're reacting to need to have ended. After all, how can you know that the ogre's attack will bring your ally to the dying condition, if this event happens after you start acting? In layman's terms, the ogre brings your ally to -1 hp on initiative 15 and your turn takes place on initiative 13. If the ogre's actions weren't complete by initiative 15, how could you possibly know, by initiative 13, the full extent of the ogre's attack?

Thus, I propose that, for 4E, we clarify the way actions take place in combat:

On your turn, you get to perform an entire round's worth of actions. These actions represent your activity during the last 6 seconds of combat.
You do realize that this distinction is irrelevant.

Your point is that, if my attack happens in the 6 seconds after I declare it, and the enemy's disarm does too, I should have to take the disarm into account.

BUT, if my attack happens in the 6 seconds prior to my declaring it, the enemy's disarm, which also happens prior to his turn, hence relatively at the same interval to my attack as in the first option, doesn't affect me.

Sounds flawed.
Yeah, there's no distinction there. The exact same actions occur over the exact same period of time. The only difference is where you place the marker (the initiative count).

If it helps you get your head around the sequence of events to consider the initiative number the end of the period, go for it. However, it doesn't make any more or less sense to do so.
The distinction exists in that you're aware of the outcome of what's going to happen prior to it actually happening. If you know that you're going to take damage from an opponent's attack while it's taking place, what's preventing you from simply acting to avoid the attack?

Let me bring up a simple example. My turn comes right before an enemy orc's does. I decide to perform a melee full attack on the orc. It's pretty much assumed that my full attack will take place over the course of 6 seconds. If the orc decides to move away at some time before I'm done, how can I guarantee that my action gets completed successfully? After all, I doubt that the orc will require all of his round to move beyond reach, especially if he can move far beyond my melee reach.

This is an example of why the distinction is relevant. Turns measure reaction times. You can't react to something that's yet to come. You can only react according to what has transpired before. If I deal 13 points of damage to my enemy during my turn, this damage doesn't kick in one round later. It's in as of the end of my action. Once my turn is over, this damage is set in stone. In other words, my action is complete.

Here's another example. Again, my turn comes up right before an enemy orc's. I attack the orc, and deal enough damage to kill him. When does the orc actually fall? Now, or in one round? If my action starts now, then it's completed in one round, which means the orc is actually there for another round. The only way for the orc to be dead now is for me to have completed my action by the end of my turn.
It is a game.
It uses simplified mechanics so the game is not bogged down in trivial issues.
It is designed for players to have fun.
It is made so that each individual takes thier turn in a sequence so no one is skippped.
It is not meant to simulate an actual battle that has taken place.
It is a game.

And if you think the fight mechanics of this game are a bit off, go play FF 7.

Sorry if this sounded mean-spirited, but I really just do not see how this is an issue.
There really is no inconsistency. Your opponent's disarm attempt fails to disrupt your attack because you had the better Initiative, a concept which applies throughout combat. Your opponent's reactions are always one step behind your own. A round isn't so much a particle as it is a wave.

Also, your proposal may have an even stranger side effect than the alleged inconsistencies. Maybe it's just me, but narrating things in past tense and then rolling to determine success just don't seem to go together. Using the past tense implies the narrator knows what happened. Even if the narrator were describing a failed attempt, building up the suspense before telling me what happened, I'd expect him to know it well enough in his own mind not to have to pause the story to roll some dice.

I think the bigger concern (for what it's worth) is that 6 seconds is too big of a discrete chunk for a player turn. They could make 3-second rounds instead. Full actions could be split into two different actions, with the success of the first action affecting the success of the second, whether through bonuses or requirements. A round could then consist of two manageable passes: a swift action pass followed by a 3-second pass. This makes a simple pattern of preparation and action, or action and reaction. Swift, 3-second, swift, 3-second, swift, 3-second, ...
Or we could institute a second by second combat scheme. Each player rolls initiative. Whoever wins goes first, they say what they are going to do. They decide to do a standard attack (3 seconds). The fight advances one second and the person with the next initiative goes next. They want to cast a spell as a swift action (1 second). This spell is immeditaley cast and attack the first combatant. Since he took damage he gets penalties to his attack when he finally attacks (at the beginning of 4 seconds)....

Actually, I quite like the simple combat. There's absolutely no need for combat to be hyper-realistic in DnD. If you like that, fine, house-rule it in. Just leave the simpler rules for us simpletons.
Actually, I quite like the simple combat. There's absolutely no need for combat to be hyper-realistic in DnD. If you like that, fine, house-rule it in. Just leave the simpler rules for us simpletons.

Yeah, but there's nothing simple about combat currently. It's because each round takes 6 seconds that there are move actions, standard actions, free actions, swift actions, full actions, immediate actions,... It would be much simpler just to have actions, or actions and reactions/preparations.

On the other hand, if you had something as fiddly as 1-second actions, I think after so many rounds of "I'm already doing this 3-second action", eventually the DM naturally change to a swift/3-second pattern anyway.
Marcus, your examples in the OP seem to me to fall into the trap of imposing a time delay on the protagonist's action, but not the antagonist's. If the fighter attacks on count 15 and his opponent attempts to disarm on count 14, the attack resolves itself before the disarm attempt.

I can't think of an action that imposes penalties on other creatures during the action itself. D&D is a binary system in that regard (note that I'm specifically referring to the actions themselves, rather than any condition (such as grappled) that the action may impose on the target).

Ultimately, I don't think the D&D 3.X abstract combat system can be aligned with real world timelines. Problems exist at either end of combats.

If actions happen "after" the initiative count (your option #1), you have a problem at the end of combats. Creatures wouldn't die until after other characters had had a chance to waste attacks on them. Ie. Ranger kills an orc on initiative count 15. Fighter attacks orc on initiative count 14. Just before 15 comes around again, the ranger completes his kill and the orc falls over.

If actions happen "before" the initiative count (your option #3), you have a problem at the beginning of combat. Surprise rounds don't make sense. If initiative is rolled as soon as the combatants encounter each other, how can they have already taken actions?

If actions happen both "before" and "after" the initiative count (your option #2), you encounter both problems.

The only real solution seems to be to avoid dealing with this issue altogether. The actions take place in the split second surrounding your initiative count unless the rules specifically state otherwise. You're doing stuff at other times during the round, but it is only at that specific point in the initiative count that you actually interract in a meaningful way with your surroundings.

I note this issue arose out of a discussion about when AoOs refresh. Utimately, I think it's easier to address that question than it is to address yours. It's as simple as adding the words

"A creature can make one attack of opportunity between the start of its turn and the start of its next turn. Some abilities or effects may increase the number of attacks of opportunity a creature can take."

to the rules for AoOs.
The distinction exists in that you're aware of the outcome of what's going to happen prior to it actually happening. If you know that you're going to take damage from an opponent's attack while it's taking place, what's preventing you from simply acting to avoid the attack?

The simple fact that you can't react fast enough. That's what the initiative mechanic refers to.

Let me bring up a simple example. My turn comes right before an enemy orc's does. I decide to perform a melee full attack on the orc. It's pretty much assumed that my full attack will take place over the course of 6 seconds. If the orc decides to move away at some time before I'm done, how can I guarantee that my action gets completed successfully? After all, I doubt that the orc will require all of his round to move beyond reach, especially if he can move far beyond my melee reach.

On the other hand, if your turn covered the last six seconds, so did the orc's - simultaneity, remember? That means that when you decide that your last six seconds were spent attacking, the orc can decide that his last six seconds were spent moving away from you. He doesn't move after your turn, since the turns are simultaneous. Instead, under either system, there was a six-second period where you got the drop on the orc and beat the hell out of him before he could move away. I just want to reinforce that: This holds equally true for both mechanics, the only difference is when the actions are announced.

This is an example of why the distinction is relevant. Turns measure reaction times. You can't react to something that's yet to come. You can only react according to what has transpired before. If I deal 13 points of damage to my enemy during my turn, this damage doesn't kick in one round later. It's in as of the end of my action. Once my turn is over, this damage is set in stone. In other words, my action is complete.

Unfortunately, what you react to in your turn hasn't "transpired before", because all actions are simultaneous. Since everybody gets to determine what happened in the last six seconds, the problem is exactly the same. The only difference is that you're announcing the actions as if they were in the past rather than as if they were in the future.

Here's another example. Again, my turn comes up right before an enemy orc's. I attack the orc, and deal enough damage to kill him. When does the orc actually fall? Now, or in one round? If my action starts now, then it's completed in one round, which means the orc is actually there for another round. The only way for the orc to be dead now is for me to have completed my action by the end of my turn.

Why couldn't the orc have done something in the last six seconds that made up both your turn and his? If you killed him at the end of your six seconds, that was also the end of his six seconds... why didn't he do anything in that time?

Instead, think of it this way: The orc dies "before the start of his turn", in other words, before he was able to react in a meaningful way. The important thing is that he wasn't able to make a meaningful contribution to the action for that round; any further precision about when his death occurred is inherently meaningless, both in your system and in the rules as written.
Strange, I've always considered the round to be six seconds in length, yes, but that the actions took place between one person's initiative score and the same point in the next round.

To me, it meant that whatever inconsistency would take place during that time was simply a delay on the other characters' actions imposed by the actions of the first character.

Sure, it means rounds actually have a relative duration rather than a fixed one, but it also eliminates most inconsistencies, I find... At least enough to suspend disbelief.

Not sure I see any point in disturbing that system by adding more complications.
There really is no inconsistency. Your opponent's disarm attempt fails to disrupt your attack because you had the better Initiative, a concept which applies throughout combat. Your opponent's reactions are always one step behind your own. A round isn't so much a particle as it is a wave.

That's just it. It's a wave, and its full effects can only be felt once it has completely passed. When the outcome of your action is known, it becomes a finite reality. Whatever happens afterwards depends on this reality. It becomes inconsistent to react to something that has yet to transpire.

Marcus, your examples in the OP seem to me to fall into the trap of imposing a time delay on the protagonist's action, but not the antagonist's. If the fighter attacks on count 15 and his opponent attempts to disarm on count 14, the attack resolves itself before the disarm attempt.

You're right. Maybe my examples were not appropriate. Here's one that should make more sense. On initiative 15, my opponent attacks me and deals 20 points of damage. If his action started on his turn, the damage will have completely sunk in around the next time initiative 15 is reached. My cleric friend, who acts on initiative 14, sees little to no damage on my person, at this point, so there's little reason for him to approach and heal my wounds. Then again, as players we know that this damage is going to kick in. It now becomes inconsistent to allow for the cleric's player to decide his character is going to spend his next round healing me, since the character should not be aware of the extent of my opponent's attack on me.

This, of course, doesn't reflect the way the game is played. At the end of my opponent's turn, I took some damage. This damage doesn't start to exist in one round; it comes into being immediately and all subsequent actions in the encounter will take this damage as existing.

The only consistent way for the cleric's player to decide his character heals mine is for him to be aware of the extent of my opponent's attack. The only way this extent can be known reasonably is if it already exists.

If actions happen "before" the initiative count (your option #3), you have a problem at the beginning of combat. Surprise rounds don't make sense. If initiative is rolled as soon as the combatants encounter each other, how can they have already taken actions?

That's just it. They haven't already taken actions. They're starting to act from the moment combat starts, and their actions are completed upon reaching their initiative, making their effects final and immutable. #3 works under the premise that you're acting right up until your turn ends, at which point your action is complete. After the end of your turn, you're already starting your next round's activity, which is completely resolved by the end of your next turn.

This holds equally true for both mechanics, the only difference is when the actions are announced.

There is also the difference of pre-existing knowledge of the results of past actions. If actions represent the end of your round's activity, at least you remain coherent in considering them a finite reality. The difference is as such:

Model #1: You start to act knowing how other actions ended. This breaks simultaneity.

Model #3: You do not finish acting in time to prevent other actions to end as they did.

Model #3 is more favorable here in that it makes an abstraction of when actions started, and concerns itself with inflicting the full effects of actions when they should: after they're completed.

Unfortunately, what you react to in your turn hasn't "transpired before", because all actions are simultaneous. Since everybody gets to determine what happened in the last six seconds, the problem is exactly the same. The only difference is that you're announcing the actions as if they were in the past rather than as if they were in the future.

Don't neglect the effects of past actions. If I took 16 points of damage right before the start of my turn and we follow model #1, then by the time I start acting, I already feel the full extent of the damage before my action even starts.

Why couldn't the orc have done something in the last six seconds that made up both your turn and his?

Because the last six seconds didn't make up both his turn and mine. No two actions take place at the exact same time. Initiative tie resolution rules and the transposition of rounds justify this. We know that actions are simultaneous in that they overlap at some points, but they're not synchronized. If the orc's turn comes after mine, then he started his round's actions less than one round ago if I kill him on my turn. Since he didn't have sufficient time to complete his action, it becomes reasonable to conclude that his actions had no effect.

The key concept to keep in mind here is that the full extent of an action should only be a factor once the action is completed.
the best mechanic for realism is the announce in reverse order and act in initiative order, or to let everyone announce what the want to do and then roll for initiative (this only works when players don´t fight players, because otherwise they won´t know who should announce first)
OK, this is an interesting discussion about what it actually means when actions take place during a course of a round. When I play D&D, I just look at it as the character with initiative 15 does all his actions during that initiative count. Then the character with initiative 14 does all of his actions during this initiative count. Continue until every character takes his turn and start the next round. I never concern myself with how the actions of the init 15 character overlaps with the actions of the init 14 character over the course of 6 seconds.

The current combat system is designed primarily to make combat playable around a table top with miniatures. By playable, combat must be simple enough for most people to understand and adequately simulate a situations where multiple characters are acting more or less simultaneously. The D&D combat system is turn based, so it makes it simple enough by completely handling one character's actions at a time before it handles another's. I'm sure there are many situations where this system makes things a little weird. I've seen many such examples. However, for the most part, it works and gets the job done.

We can go on and on about what actions in a D&D combat round means in real life. As has been mentioned, there have been other large threads that discuss this. I just don't see how more specific definitions regarding a round of actions is going to improve my experience more. I'm just happy with "character A does all his stuff" then "character B does all her stuff" and we move on from there.
<\ \>tuntman
I think you are creating a false dichotomy.

Orc attacks you on 15 on round one

Cleric cannont heal you on 14 on round one because you aren't hurt yet

Damage is dealt on 15 on round 2

Seems to be your proposition.

I see initiative as more dynamic ... on count 15 or round one orc slice launches a series of thrust at you with his spear, eventually he connects

on count 14 of round one, Brother Bob of the Holy Hand sees your in trouble and starts moving towards you and casts his healing magic.

Any system that is broken into a series of alternating steps is going to face these issues.

And I think Dnd 3rd/3.5 is immanently more playable that the other popularly given alternative of announce in reverse order of initiative and then act where early actions make later actions improbable inappropriate or even impossible.
The dichotomy exists in that a player decides what his character will do based on what's been going on. If the orc deals damage to you, but doesn't kill you, the party cleric will heal your damage. However, if the orc deals enough damage to kill you, the cleric might instead opt to revive you. Since the decision making is based on the situation in the game, actions have to be final at the moment they're resolved.

Look at it in another way. I take 5 points of damage when an orc attacks me on initiative 15. When do these points of damage kick in? Now, or in one round? If it's now, it stands to reason that the action is complete. Otherwise, we end up asking ourselves why the orc spent the rest of his round doing didly-squat, since his round activity supposedly ends right before his next turn. If the action is complete, and your actions represent a round's worth of activity, then it becomes obvious as to why your actions are actually ending a round's activity.

In a simple analogy, consider pay checks. You get paid every week (or every other week). Your pay represents the work you've been doing in the last salary period, not the work you're going to do in the next salary period. Similarly, you get actions every round. These actions represent the work you've been doing in the last round, not the work you're going to do in the next one.

And I think Dnd 3rd/3.5 is immanently more playable that the other popularly given alternative of announce in reverse order of initiative and then act where early actions make later actions improbable inappropriate or even impossible.

I wholeheartedly agree. However, what I'm proposing doesn't change the way the mechanics are run. It is simply a clarification as to what the actions represent, in order to better work with the oxymoronic aspect of a round's activity being both simultaneous and sequential.