Hey! Watch out!

[EggKookoo]

Imagine an encounter beginning with a NPC Foe hidden. PC Bob fails to notice Foe. PC Mary perceives Foe. Initiative order is Mary, Bob, Foe. Per the surprise rules, Bob can't take an action or move on his turn this first round.

What if Mary, on her turn, yells out to Bob about Foe? Per RAW, nothing happens, of course. But it feels like Mary's warning should serve some purpose. I'm musing about a homebrew mechanism here that would feel balanced. If Mary can just yell out and undo the surprise condition on Bob, that trivializes surprise. At the other end, Mary using her entire action to warn Bob probably costs more than it's worth to do. If I make it cost Mary her bonus action, well, that only matters if she had a bonus action to miss. Maybe that's okay, but it still feels like it might trivialize surprise too much.

Does anyone do anything like this?

[MaxWilson]

Imagine an encounter beginning with a NPC Foe hidden. PC Bob fails to notice Foe. PC Mary perceives Foe. Initiative order is Mary, Bob, Foe. Per the surprise rules, Bob can't take an action or move on his turn this first round.

What if Mary, on her turn, yells out to Bob about Foe? Per RAW, nothing happens, of course. But it feels like Mary's warning should serve some purpose. I'm musing about a homebrew mechanism here that would feel balanced. If Mary can just yell out and undo the surprise condition on Bob, that trivializes surprise. At the other end, Mary using her entire action to warn Bob probably costs more than it's worth to do. If I make it cost Mary her bonus action, well, that only matters if she had a bonus action to miss. Maybe that's okay, but it still feels like it might trivialize surprise too much.

Does anyone do anything like this?

Nope. If Bob is reading books in his library with Mary when they are suddenly attacked by werewolves, it doesn't matter whether Mary or Bob or the werewolves go first, Bob isn't mentally ready to be dealing with werewolves right now and he's surprised. There's nothing Mary can do that will snap Bob out of library mode faster than actually being attacked by werewolves has already done.

However, surprise rarely applies in my games either--about the only scenario where it would happen in a dungeon crawl is if you're attacked while resting. (If you're in danger-awareness mode, you're not relaxed enough to be resting, unless you're also Alert.) The way this scenario would play out during a dungeon crawl would just be that the hidden attacker automatically wins initiative.

[NaughtyTiger]

it means only 1 or 2 creatures will every be surprised in a surprise round... not very surprise-y
it also means higher initiative player gets punished cuz players slower than mary get to act normally.

wouldn't that work for monsters as well...
the first monster to perceive the party can warn the rest?

[EggKookoo]

I don't do surprise much either. In this case, I was trying to model an ambush, where the party walks out into the open (unless they think to be stealthy). I suppose it's simpler to give the ambusher a free attack unless the party has some way to know she's there, and then run initiative from there.

[Bobthewizard]

It might help you to picture it all happening at the same time. It's a sudden attack and then Mary barely gets her sword out in time to attack the Foe before it hits her, but Bob doesn't, no matter if Mary warns him or not. It's not like the characters are sitting around waiting for someone else to finish their turn like the players are. All the movements and attacks happen at about the same time, and initiative and surprise just tell us what order to determine the outcomes.

[Keravath]

It might help you to picture it all happening at the same time. It's a sudden attack and then Mary barely gets her sword out in time to attack the Foe before it hits her, but Bob doesn't, no matter if Mary warns him or not. It's not like the characters are sitting around waiting for someone else to finish their turn like the players are. All the movements and attacks happen at about the same time, and initiative and surprise just tell us what order to determine the outcomes.

I think this is a good way to think of it. Each round is 6 seconds and although we resolve everyone's actions sequentially, in some senses they are happening simultaneously. D&D is not a simulator, it is a role playing game so the details of sequential initiative in a parallel action world may not make a lot of logical sense but it keeps things simple, everyone gets a turn and they resolve their actions in order.

If you look at it like this, Mary, Bob and the Foe are acting at the same time. Mary can yell, make lots of noise, but Bob is still surprised since it is all taking place in this one 6 second window and until after his turn has passed in the initiative order he can't take reactions and can't take his turn until the next round. This works because the character actions aren't really sequential, they are all happening in the same 6 seconds and the game only resolves them sequentially because it is easy.

If you think about it ... in D&D ... if two characters start 20' apart, one can always run up and attack the other if they have at least 20' of movement. However, in a "simulator" movement could be actually simultaneous and before the one character covers the 20' the other one could be running away and the other would not catch them.

[Segev]

What really sealed this for me was the realization that, indeed, it is all happening more or less at once. Everyone isn't standing around while Mary takes her turn. The bad guy is attacking and Mary is reacting and Bob is being surprised all at once. Initiative just tells us whether unsurprised Mary moves faster than initiating bad guy; if she does, she might hit him before he's done aiming his thrown dagger at Bob.

The way I'd describe it is, "Roll initiative. Okay, Mary, you go first. You see a guy stepping out of the shadows, getting ready to attack Bob with a throwing dagger! What do you do?"

If Mary does something that makes the guy not throw the dagger, he still has his action on his turn to do with as he pleases, but it was his moving to start throwing a dagger that triggered the initiative roll in the first place, so knowing that was his intended first action is totally fair.

[EggKookoo]

Right, thinking of it simultaneously helps. Although it's fun to think of D&D combat as purely simultaneous. You can run up to me while I'm concentrating on a spell, hit me, make me lose concentration, which prompts me to cast a new spell that heals my ally who was incapacitated and at 0 HP on the ground. Since he hadn't gone yet, my ally can jump to his feet and attack you, scoring a good hit and driving you to 0 HP.

I mean, it all could be simultaneous, but it's funny to visualize.

[Segev]

Right, thinking of it simultaneously helps. Although it's fun to think of D&D combat as purely simultaneous. You can run up to me while I'm concentrating on a spell, hit me, make me lose concentration, which prompts me to cast a new spell that heals my ally who was incapacitated and at 0 HP on the ground. Since he hadn't gone yet, my ally can jump to his feet and attack you, scoring a good hit and driving you to 0 HP.

I mean, it all could be simultaneous, but it's funny to visualize.

I haven't seen this done, but it's conceivably possible to do away with initiative and just have everyone act simultaneously. Everyone writes down what they're doing and handles any rolls on their end, and then the DM goes around the table asking people what they're doing. They say it, any rolls targets need to make (e.g. saves) are made, movements are done, and then the results are applied at the end of the round all at once. If you want to move into melee with somebody, as long as you can get from where you started the round to where they are at the start of the round, you can do so.

This doesn't work in 5e or 3.5 as well as you might like, though, because opportunity attacks and AOOs and reactions and immediate actions and the like NEED off-turn to function, and this system eliminates "off-turn" as a thing.

[EggKookoo]

I haven't seen this done, but it's conceivably possible to do away with initiative and just have everyone act simultaneously. Everyone writes down what they're doing and handles any rolls on their end, and then the DM goes around the table asking people what they're doing. They say it, any rolls targets need to make (e.g. saves) are made, movements are done, and then the results are applied at the end of the round all at once. If you want to move into melee with somebody, as long as you can get from where you started the round to where they are at the start of the round, you can do so.

This doesn't work in 5e or 3.5 as well as you might like, though, because opportunity attacks and AOOs and reactions and immediate actions and the like NEED off-turn to function, and this system eliminates "off-turn" as a thing.

Yeah, whenever I'm tempted to do this I remember reactions are a thing.

[Segev]

Yeah, whenever I'm tempted to do this I remember reactions are a thing.

It does require either a super-simplistic existing system or making your own, yeah.

[Samayu]

I haven't seen this done, but it's conceivably possible to do away with initiative and just have everyone act simultaneously. Everyone writes down what they're doing and handles any rolls on their end, and then the DM goes around the table asking people what they're doing. They say it, any rolls targets need to make (e.g. saves) are made, movements are done, and then the results are applied at the end of the round all at once. If you want to move into melee with somebody, as long as you can get from where you started the round to where they are at the start of the round, you can do so.

I've tried this when running Paranoia. Yes, it's a super simplistic system. And this is still a pain, partly because I play it with hardcore gamers who don't want to be told their action had no effect because something changed in the meantime.

[Segev]

I've tried this when running Paranoia. Yes, it's a super simplistic system. And this is still a pain, partly because I play it with hardcore gamers who don't want to be told their action had no effect because something changed in the meantime.

Why would their action have no effect? The whole point of the way it's laid out is that your actions all have effect at the same time. If you could take it at the beginning of the round, how things were wet up then, it goes off appropriately.

[Shadhael]

Why would their action have no effect? The whole point of the way it's laid out is that your actions all have effect at the same time. If you could take it at the beginning of the round, how things were wet up then, it goes off appropriately.
A possible example from how I understand it:

An enemy archer is standing 25 feet North of melee Fighter PC. Enemy melee fighter is engaged with Fighter PC's Monk ally 25 feet East of Fighter PC.

Fighter PC wants to engage the enemy archer. Decides to move north and take the Attack action. Enemy archer wants to maintain distance, and decides to retreat north 20 feet and launch a ranged attack against the monk.

Resolved simultaneously, the Fighter PC can't close the distance to the archer as they both moved north. Fighter now has no enemy within range to attack, and feels like their turn is wasted.

Resolved on initiative order: 1) the archer is either too far away if they moved first and the Fighter can target the enemy melee combatant engaged with their monk ally. Or 2) the Fighter closes the distance first and make their attack against the archer.

So if resolved simultaneously, the conditions that lead to one choice may change to make that choice very poor and players feeling like they wasted their turn. Of course some tables may want that added layer of realism (the guy you're chasing is also fleeing from you at the same time as opposed to having the distance between the two of you yo-yo) and that's fine too.

[Segev]

A possible example from how I understand it:

An enemy archer is standing 25 feet North of melee Fighter PC. Enemy melee fighter is engaged with Fighter PC's Monk ally 25 feet East of Fighter PC.

Fighter PC wants to engage the enemy archer. Decides to move north and take the Attack action. Enemy archer wants to maintain distance, and decides to retreat north 20 feet and launch a ranged attack against the monk.

Resolved simultaneously, the Fighter PC can't close the distance to the archer as they both moved north. Fighter now has no enemy within range to attack, and feels like their turn is wasted.

Resolved on initiative order: 1) the archer is either too far away if they moved first and the Fighter can target the enemy melee combatant engaged with their monk ally. Or 2) the Fighter closes the distance first and make their attack against the archer.

So if resolved simultaneously, the conditions that lead to one choice may change to make that choice very poor and players feeling like they wasted their turn. Of course some tables may want that added layer of realism (the guy you're chasing is also fleeing from you at the same time as opposed to having the distance between the two of you yo-yo) and that's fine too.

Right. The key way I would design such a system is that EVERYONE's actions are valid based on the state of things at the START of the round. Because we're not re-describing the scene mid-round.

So the archer can move away and shoot, but the fighter can move up and hit him. The fighter got into melee and hit him and the archer moved away and shot, so at the end of the round, the archer is damaged but has moved successfully away from the fighter, who will have to move after him again if he wants to keep engaging him in melee.

This is actually one of the points where this doesn't work for D&D, because you have to resolve whetherthe fighter gets an AOO for the archer moving away from him, and that's not a state that is measured by this system. The archer wasn't next to the fighter at the start of the round, so his decision to move further away doesn't get unforeseen information added to complicate it in the form of suddenly being next to the fighter for part of the round.

The tactics of this system change a little, too: the archer has to consider whether he wants to risk the fighter remaining in move-distance of him. Maybe he dashes (or double moves, or whatever) and foregoes his attack. Now, the fighter still moves and attacks him, and the archer doesn't get a shot off, but the archer ends the round a double-move away from the fighter, meaning that the fighter now would have to spend all of next round moving to even keep up, and not have a chance to attack.

But the point is: assess the validity of all actions based on how the round STARTED. Anything that could be done then, can be successfully attempted (because, you know, dice might say it fails anyway). The end of the round state is based on all actions that were thus resolved. This becomes the starting state for the next round, which determines what actions are valid to attempt for said round.

This also means in-combat healing directly negates in-combat damage during the same round, because you're not going to know whether the target hit 0 hp or not as the healing and damage go off "at the same time," effectively.

[Grey Watcher]

While I'm of the overall opinion that it's fine as-is (what with in-universe time being so much less than out), I don't know how much it actually screws up balance to add a Warn action. I mean, same result: somebody doesn't get their turn. (OK, movement, but maybe Warn only grants the target an action, not a complete negation of surprise.) Mary and Bob still only get one action between them, it just changes who gets to act. Of course, if Bob has something more immediately useful, it can change a lot, but no more than if Bob had successfully perceived Foe and Mary had not.

[Segev]

While I'm of the overall opinion that it's fine as-is (what with in-universe time being so much less than out), I don't know how much it actually screws up balance to add a Warn action. I mean, same result: somebody doesn't get their turn. (OK, movement, but maybe Warn only grants the target an action, not a complete negation of surprise.) Mary and Bob still only get one action between them, it just changes who gets to act. Of course, if Bob has something more immediately useful, it can change a lot, but no more than if Bob had successfully perceived Foe and Mary had not.

It introduces some weirdness if the DM calls for initiative because something hostile but undetectable is happening. The "transferring hex" debate in another thread is a good example. But that is a particularly corner case.

[EggKookoo]

While I'm of the overall opinion that it's fine as-is (what with in-universe time being so much less than out), I don't know how much it actually screws up balance to add a Warn action. I mean, same result: somebody doesn't get their turn. (OK, movement, but maybe Warn only grants the target an action, not a complete negation of surprise.) Mary and Bob still only get one action between them, it just changes who gets to act. Of course, if Bob has something more immediately useful, it can change a lot, but no more than if Bob had successfully perceived Foe and Mary had not.

I considered that Mary sacrifices her action to restore Bob's to him, but that felt too expensive. My reason behind asking the question was to help work up a way to explain it to the players. I'm fine with the existing rule (as the DM I'm often dealing with stuff that's more gamey than simulationist for balance reasons), but I also accept that it's reasonable to for Mary to think "Well, I wasn't surprised, why can't I yell out a 'heads up' to Bob?"

I like the idea that maybe yelling to Bob costs Mary her reaction (taken at the start of Foe's turn), and restores Bob's reaction to him, and that's it. It's a one-for-one. It's not a huge cost or a huge benefit, but it might save Bob in the right circumstances. And it feels right -- reactions are more about those reflex-based quick things.

It'll be a few sessions before I get the opportunity to try it out in the actual game...

[MaxWilson]

While I'm of the overall opinion that it's fine as-is (what with in-universe time being so much less than out), I don't know how much it actually screws up balance to add a Warn action. I mean, same result: somebody doesn't get their turn. (OK, movement, but maybe Warn only grants the target an action, not a complete negation of surprise.) Mary and Bob still only get one action between them, it just changes who gets to act. Of course, if Bob has something more immediately useful, it can change a lot, but no more than if Bob had successfully perceived Foe and Mary had not.

It's not so much a problem with game balance as with the game logic / fiction. Let's say the initiative order is:

Attacker
Mary: not surprised
Bob: surprised

Mary's perception is high enough that she is not surprised. What exactly is she doing to Warn Bob more effectively than the actual attacker stabbing Bob? "Watch out Bob, you've just been stabbed!"?

[Demonslayer666]

I don't see any problem with giving them a bonus on their initiative if another player shouts a warning. It may give them the opportunity to act in the second round before the ambushers, or give them their reaction before the ambushers attack in the first round, but I would not make it a thing unless there was an ambush and the person shouting the warning acted in the first round.

[EggKookoo]

It's not so much a problem with game balance as with the game logic / fiction. Let's say the initiative order is:

Attacker
Mary: not surprised
Bob: surprised

Mary's perception is high enough that she is not surprised. What exactly is she doing to Warn Bob more effectively than the actual attacker stabbing Bob? "Watch out Bob, you've just been stabbed!"?

My original order was Mary, Bob, Foe. Mary sees Foe, Bob does not, therefore Bob cannot move or act or get his reaction until Foe's turn is completed (if I'm getting the rule right). So the benefit of Mary's warning is that Bob might be able to do something before Foe's attack. At the very least, maybe Bob gets his reaction back, allowing him to cast shield or some equivalent.

[MaxWilson]

My original order was Mary, Bob, Foe. Mary sees Foe, Bob does not, therefore Bob cannot move or act or get his reaction until Foe's turn is completed (if I'm getting the rule right). So the benefit of Mary's warning is that Bob might be able to do something before Foe's attack. At the very least, maybe Bob gets his reaction back, allowing him to cast shield or some equivalent.

Yes, I realize that your original proposal was oriented around a specific initiative order, but the Warn proposal (as Grey Watcher named it) has logic issues when the initiative order is different.

I.e. I realize that it works as intended when the order is Mary, Bob, Attacker, but I'm a programmer by trade. I don't just think in terms of "it works in this test case, let's ship it!"

[EggKookoo]

Yes, I realize that your original proposal was oriented around a specific initiative order, but the Warn proposal (as Grey Watcher named it) has logic issues when the initiative order is different.

I.e. I realize that it works as intended when the order is Mary, Bob, Attacker, but I'm a programmer by trade. I don't just think in terms of "it works in this test case, let's ship it!"

But it only really needs to work when in that (or similar) order, and fails gracefully when the order is different. Ship that sucker!

What I mean is, if it's Foe -> Mary -> Bob, and we apply the same Warn mechanism, nothing breaks. Mary just wasted her action. No different than any other wasted action (casting cure wounds on someone at full HP, for example). It's self-solving since she likely wouldn't bother with that particular action.

[Segev]

But it only really needs to work when in that (or similar) order, and fails gracefully when the order is different. Ship that sucker!

What I mean is, if it's Foe -> Mary -> Bob, and we apply the same Warn mechanism, nothing breaks. Mary just wasted her action. No different than any other wasted action (casting cure wounds on someone at full HP, for example). It's self-solving since she likely wouldn't bother with that particular action.

If the person with the most useful action is Bob, it's still an improvement, and allowing them to "unsurprise" their buddy more than the action their buddy just saw happen does is...anti-"the fantasy."

[EggKookoo]

If the person with the most useful action is Bob, it's still an improvement, and allowing them to "unsurprise" their buddy more than the action their buddy just saw happen does is...anti-"the fantasy."

The whole surprise thing is weird, the more I think about it.

Foe hides.

Mary and Bob come along and make perception checks (passive or otherwise, doesn't matter). Then they roll initiative, and Foe ends up moving last.

Mary sees Foe and attacks him. At this point, why is Bob still surprised? Mary's not hiding from him, and he just saw her attack Foe. Why does he not know Foe is there? Why wouldn't he suddenly be on "combat alert" the way Mary was? Her action should pull Bob out of surprise. Granted, he may not be able to see Foe, but at most that should just give him disadvantage on his attack.

Maybe I need to make my own homebrew surprise mechanic.

[Segev]

The whole surprise thing is weird, the more I think about it.

Foe hides.

Mary and Bob come along and make perception checks (passive or otherwise, doesn't matter). Then they roll initiative, and Foe ends up moving last.

Mary sees Foe and attacks him. At this point, why is Bob still surprised? Mary's not hiding from him, and he just saw her attack Foe. Why does he not know Foe is there? Why wouldn't he suddenly be on "combat alert" the way Mary was? Her action should pull Bob out of surprise. Granted, he may not be able to see Foe, but at most that should just give him disadvantage on his attack.

Maybe I need to make my own homebrew surprise mechanic.

I'm going to restate the scenario with a few more specifics. Feel free to tell me if I have gotten anything blatantly wrong from how you were picturing the scenario, but please be specific if you do. Some differences will be just me trying to fill in details, others will be due to trying to make clear what's going on.

Mechanically:
Foe is hiding with a check that set the DC to spot him at 15.
Mary rolls (or has a passive perception of) 16; Bob rolls (or has a passive perception of) 12.
Mary is not surprised.
Bob is surprised.
Foe declares to the DM he plans to attack! (Or, more likely, DM knows Foe is about to attack.)
Initiative is rolled.
Mary rolls a 20.
Bob rolls a 10.
Foe rolls a 5.
Mary attacks Foe!
Bob ceases to be Surprised.
Foe attacks Bob!

Narratively:
The wiley bugbear is lurking in the bushes up ahead. Mary notices there's something there just as it shifts its stance, indicating it's ready to attack! Bob is startled as Mary unexpectedly lets out a battle-cry and charges the bugbear, who Bob is further stunned to realize is a mere 10 feet away from him! Mary's reaction is so fast that she lands her hit before the Bugbear managed to bring his glaive around to strike at Bob. Bob finally shakes himself out of it and readies for combat, himself! (Next round starts)

Alternatives might be that Mary dive-tackles the bugbear, or shoves Bob out of the way of an attack, or something. The important thing is that, because she wasn't surprised, Mary notices the Bugbear getting ready to attack and can, because her initiative was higher, react to him faster than he can actualize it.

Incidentally, this is why I look askance at those who want to roll initiative when there's literally nothing to react to, even when the initiative-triggering action is completed. Because what's Mary doing, acting on psychic intuition that somebody did something and thus flipping out and attacking somebody at random?

[Lunali]

A possible example from how I understand it:

An enemy archer is standing 25 feet North of melee Fighter PC. Enemy melee fighter is engaged with Fighter PC's Monk ally 25 feet East of Fighter PC.

Fighter PC wants to engage the enemy archer. Decides to move north and take the Attack action. Enemy archer wants to maintain distance, and decides to retreat north 20 feet and launch a ranged attack against the monk.

Resolved simultaneously, the Fighter PC can't close the distance to the archer as they both moved north. Fighter now has no enemy within range to attack, and feels like their turn is wasted.

Resolved on initiative order: 1) the archer is either too far away if they moved first and the Fighter can target the enemy melee combatant engaged with their monk ally. Or 2) the Fighter closes the distance first and make their attack against the archer.

So if resolved simultaneously, the conditions that lead to one choice may change to make that choice very poor and players feeling like they wasted their turn. Of course some tables may want that added layer of realism (the guy you're chasing is also fleeing from you at the same time as opposed to having the distance between the two of you yo-yo) and that's fine too.

Alternate resolution, the archer and fighter both move to a point 45ft north of where the fighter was, the archer gets off an attack on the monk before the fighter is able to close to melee range. The fighter doesn't get to make their attack because the turn ended, but is now in melee range of the archer.

[EggKookoo]

I'm going to restate the scenario with a few more specifics. Feel free to tell me if I have gotten anything blatantly wrong from how you were picturing the scenario, but please be specific if you do. Some differences will be just me trying to fill in details, others will be due to trying to make clear what's going on.

Mechanically:
Foe is hiding with a check that set the DC to spot him at 15.
Mary rolls (or has a passive perception of) 16; Bob rolls (or has a passive perception of) 12.
Mary is not surprised.
Bob is surprised.
Foe declares to the DM he plans to attack! (Or, more likely, DM knows Foe is about to attack.)
Initiative is rolled.
Mary rolls a 20.
Bob rolls a 10.
Foe rolls a 5.
Mary attacks Foe!
Bob ceases to be Surprised.
Foe attacks Bob!


Ok, so maybe I'm misunderstanding the RAW, but I thought Bob continued to be surprised until step 7 ("Foe attacks Bob!"). Are you saying it works as you outlined? That would make much more sense to me.

If it works like this, then Bob should still get his turn, since it was Mary's action on init count 20 that popped him out of his surprised state, and he still had a turn at init count 15.



Narratively:
The wiley bugbear is lurking in the bushes up ahead. Mary notices there's something there just as it shifts its stance, indicating it's ready to attack! Bob is startled as Mary unexpectedly lets out a battle-cry and charges the bugbear, who Bob is further stunned to realize is a mere 10 feet away from him! Mary's reaction is so fast that she lands her hit before the Bugbear managed to bring his glaive around to strike at Bob. Bob finally shakes himself out of it and readies for combat, himself! (Next round starts)

Wait, why does the next round start? Foe hasn't gone yet. Or has he? Is that what you mean by "managed to bring his glaive around"?

But if Foe hasn't gone yet, and Mary's actions cause Bob to cease to be surprised, I still don't get why Bob can't act.

I know, I know. I'm looking at this in a linear-init-count way and we all more or less agree that combat is this weird mushy semi-simultaneous thing. But then the surprise rule has this sequential thing where once Foe acts, Bob gets his reaction back (and would get his action and movement if his init was lower than Foe's).

Sheesh, maybe I should start another thread about illusions and faking depth so we can discuss something that makes sense. :smallsmile:


Incidentally, this is why I look askance at those who want to roll initiative when there's literally nothing to react to, even when the initiative-triggering action is completed. Because what's Mary doing, acting on psychic intuition that somebody did something and thus flipping out and attacking somebody at random?

Yeah, the start of fights is often weird. Like, how does initiative model a Mexican standoff? I guess everyone's holding their actions? In practice, I have no problem with PCs and monsters having a bit of a game-of-chicken standoff before someone just up and attacks. I resolve that one attack, then everyone rolls init and we move on from there.

Edit: Ok, I went and re-educated myself. Bob isn't surprised until Foe acts, he's surprised until the end of his turn. Which is why it happens at step #7.

[Segev]

Ok, so maybe I'm misunderstanding the RAW, but I thought Bob continued to be surprised until step 7 ("Foe attacks Bob!"). Are you saying it works as you outlined? That would make much more sense to me.

If it works like this, then Bob should still get his turn, since it was Mary's action on init count 20 that popped him out of his surprised state, and he still had a turn at init count 15.First off, no, I'm going by the existing rules, not with this "warn" action thing. Mary did not "pop [Bob] out of his surprised state."

Mary wasn't surprised, and took her turn on init count 20. Bob was surprised, and ceased to be surprised on his turn on init count 10; ceasing to be surprised took all of his turn. The bugbear (the foe) attacked on his turn on init count 5.


Wait, why does the next round start? Foe hasn't gone yet. Or has he? Is that what you mean by "managed to bring his glaive around"?Yes. "The bugbear managed to bring his glaive around" was me trying to say, narratively, that he brought it around to hit Bob. Sorry that I described that poorly. That is the Bugbear taking his action.


But if Foe hasn't gone yet, and Mary's actions cause Bob to cease to be surprised, I still don't get why Bob can't act.Bob doesn't stop being surprised until his action comes up. Mary's action is what I narratively said he noticed, but mechanically, he stops being surprised as his only action on his turn. It could be anything that Bob "notices." Let's switch it up a bit:

Bugbear declares intent to attack, triggering initiative!
Mary rolls 20
Bugbear rolls 10
Bob rolls 5
Mary, not surprised, attacks the Bugbear on initiative count 20.
Bugbear decides Mary is more dangerous and attacks her on initiative count 10.
Bob stops being Surprised on initiative count 5.
Next round starts.

Narratively, this would still have Bob surprised even as the Bugbear and Mary go at it. It takes him a few moments (roughly 6 seconds' worth) to get his head in the game, whereupon he can take Reactions and will get his turn in the next round.

What initiative rolls higher than the triggering person's initiative represent are people being able to react faster than they can complete their desired actions. If those people are surprised, that prevents them from acting pre-emptively, but it does mean they can take reactions in response to the triggering action (or anybody else doing anything before it, but after their turn). Such people didn't see it coming in time to preemtively act, but see it as it's happening and can REACT if they have a relevant reaction to perform.

People who are not surprised, but still roll lower on initiative than the triggering action, can take reactions against the triggering action (and anything else, really), but still don't get their act together fast enough to take preemptive action before the triggering action is resolved.

People who are surprised AND slower on initiative than the triggering action probably don't even realize anything is happening until after the triggering action is resolved. They "miss" the first bit of combat, suddenly realizing things have taken a turn when their turn comes up.

Remember, this is all less than 6 seconds. Chaos broke out and the people who are low-init and surprised just took that long to register something is "up."


I know, I know. I'm looking at this in a linear-init-count way and we all more or less agree that combat is this weird mushy semi-simultaneous thing. But then the surprise rule has this sequential thing where once Foe acts, Bob gets his reaction back (and would get his action and movement if his init was lower than Foe's).

Sheesh, maybe I should start another thread about illusions and faking depth so we can discuss something that makes sense. :smallsmile:I hope my description just above this helps.

Surprise tells you when you notice "something is happening" in the mushy semi-simultaneous milleaux. If you're surprised, you don't notice until your turn. If you're not surprised, you notice right away.

Initiative tells you how quickly you can act to do something about the "something [that] is happening." If you're acting before the initiative of the one who decided to start "something," you either had already noticed them ahead of time and thus are able to act immediately (i.e. you were not surprised), or you hadn't seen them until just now (i.e. you were surprised, and lose your action to becoming unsurprised), and you're only now noticing as they become visible half-way to completing their act, and you don't have time to formulate a response of your own unless you have a Reaction you can take on their turn to whatever they're doing.

If you're acting only after the initiative of the one who decided to start "something," and you hadn't noticed them before they did start "something" (i.e. you are surprised), you actually only notice the "something" at all (or at least, pull yourself together enough to be able to do more than stare in shock at it) when your turn comes around; you can't even React when they do it because by the time you notice, it's already done. If you're acting that slowly but had noticed them (i.e. you weren't surprised), then you could do anything fast enough in response (i.e. take your Reaction), but nothing that is so complicated that you have to initiate action on your own (i.e. an action or bonus action or move) in time to interrupt what he's doing.


Yeah, the start of fights is often weird. Like, how does initiative model a Mexican standoff? I guess everyone's holding their actions? In practice, I have no problem with PCs and monsters having a bit of a game-of-chicken standoff before someone just up and attacks. I resolve that one attack, then everyone rolls init and we move on from there.Actually, in a Mexican Standoff, if nobody has started shooting yet, initiative hasn't yet been rolled.

In the classic version, nobody is hidden, and everybody knows everyone else is there and armed. Until somebody says, "I shoot Bob," (or whoever they say they're shooting), nobody rolls initiative.

When Jane says, "I shoot Bob," initiative is rolled. Yes, Jane twitched first. But anybody who rolls higher than her in initiative noticed her twitching and aiming for a kill-shot on Bob and manages to move faster than she does. EVERYBODY (because nobody is Surprised) can use any relevant reactions to Jane's shooting when Jane's turn comes up. They can also use Reactions to others' actions before Jane's turn.

Initiative models a Mexican Standoff by being the mechanic that determines who actually shoots first, even if it's clear that someone else "started it."



Another scenario to picture: Some college kids are at a frat party, and one jock is getting beligerant with a nerd who keeps poking his buttons. Finally, the jock snaps, and rears back to punch the nerd, only for the nerd's judoka girlfriend to spin-flip the jock to the ground.

Mechanically, the V. Human 2nd level Fighter with Prodigy for Athletics rolled higher initiative than the jock, and she grappled him and action surged to knock him prone before his turn came up. Her nerdy bard boyfriend didn't even have to use vicious mockery.

[EggKookoo]

Sorry, I got so caught up in trying to reinterpret things that I lost sight of the original rule. I didn't mean to make you re-explain yourself. I'll go sit in the corner for a bit.

I suppose surprise comes up infrequently enough that there's probably not much value in creating a "warn" mechanic. But if I did, like I said, doing so would at most return Bob's reaction to him in case he had some kind of defensive ability available. Seems like something that's not too powerful but potentially very useful.

[Segev]

Sorry, I got so caught up in trying to reinterpret things that I lost sight of the original rule. I didn't mean to make you re-explain yourself. I'll go sit in the corner for a bit.

I suppose surprise comes up infrequently enough that there's probably not much value in creating a "warn" mechanic. But if I did, like I said, doing so would at most return Bob's reaction to him in case he had some kind of defensive ability available. Seems like something that's not too powerful but potentially very useful.

From a narrative standpoint, the problem with a "Warn" action is that Mary calling out, "Bob! Bugbear!" is no more going to get him to react to it faster than is him being attacked by the Bugbear.

The actual point of weirdness is that, if Mary were on the Bugbear's side, it would mean Bob wasn't surprised because he is aware of Mary. Though in such a situation, it would likely be that the DM substitutes a Deception roll on Mary's part against Bob's Insight to see if he realizes she's leading him into an ambush rather than is the friend and ally he thought she was, and substitute that.

The Surprise rules do hinge on being able to identify who's on whose side. In that, if the only people you're aware of are those on your side, you're Surprised.

[EggKookoo]

From a narrative standpoint, the problem with a "Warn" action is that Mary calling out, "Bob! Bugbear!" is no more going to get him to react to it faster than is him being attacked by the Bugbear.

I think it's more that Bob only knows there's a threat once it's affected him. But Mary could alert him to the presence of the threat (assuming she goes before Foe). So when it comes to Foe's turn, Bob is now aware of him.

So, RAW:

Mary sees Foe and is not surprised. She attacks Foe.
Foe attacks Bob, who had no idea Foe was there.
To model Bob's confusion, he can't move or take an action this round, and only use his reaction after his turn ends.

But with "warn":

Mary sees Foe and is not surprised. She attacks Foe.
On Foe's turn, Mary uses her reaction to warn Bob.
Bob becomes aware of Foe's presence and regains his ability to use his reaction.
Foe (we're still on his turn) attacks Bob. Most of the time this is the same as RAW, but if Bob has something like shield, he gets to use it.
Then to keep the impact of surprise, Bob still can't take an action. Perhaps the warn mechanic allows him to move on his turn. I guess that would be subject to some testing.

But yeah, surprise is probably so rare that this isn't necessary. I'm just thinking about how to handle it narratively if Mary were to call out a warning to Bob. Per RAW, Bob gains no benefit from that. It just feels off.


The Surprise rules do hinge on being able to identify who's on whose side. In that, if the only people you're aware of are those on your side, you're Surprised.

They hinge on identifying who is a threat, for sure. If you detect one threat, you're not surprised even of other threats are hidden. This is partly why it seems like it makes sense that Mary's warning would alert Bob to the presence of a threat, and therefore render him "not surprised" any more.

[Segev]

But yeah, surprise is probably so rare that this isn't necessary. I'm just thinking about how to handle it narratively if Mary were to call out a warning to Bob. Per RAW, Bob gains no benefit from that. It just feels off.



They hinge on identifying who is a threat, for sure. If you detect one threat, you're not surprised even of other threats are hidden. This is partly why it seems like it makes sense that Mary's warning would alert Bob to the presence of a threat, and therefore render him "not surprised" any more.

Surprise isn't that rare in my experience as a DM, actually. Happens every 2-3 sessions.

That said, I think where you're hung up is that, if Mary notices the threat before initiative is even considered tobe rolled, and tells Bob about it, then Bob won't be surprised. It's only when Mary notices the threat just as things are going down that her warning comes too late to penetrate Bob's inattention before things are already happening.

There's no difference between Mary shouting "DIE, BUGBEAR!" and "BOB, LOOK OUT!" as her action, let alone reaction. Either way, the fact that the Bugbear is attacking just as Mary is, just a little slower, will catch up to Bob at the same time Mary's sudden outburst does.

[EggKookoo]

That said, I think where you're hung up is that, if Mary notices the threat before initiative is even considered tobe rolled, and tells Bob about it, then Bob won't be surprised. It's only when Mary notices the threat just as things are going down that her warning comes too late to penetrate Bob's inattention before things are already happening.

I think what I'm zeroing in on is that I'm used to initiative being an active thing on the part of the PC (not the player, the PC). As soon as I ask Bob('s player) to roll initiative, I've adopted the mindset that Bob is an active participant in the fight. Except he's not. The fight has started around him but he's technically unaware of it. Sort of.

Ideally, a surprised creature would not roll initiative while everyone else does. Except that the rules then say it gets to use its reaction once its turn has passed, and we need init to know that. Maybe a middle ground approach is to use the surprised creature's initiative score to determine their turn position for that first round, then have the player roll starting with the second round, and go from there. It allows the RAW rule to work as intended but also implies the creature's head really isn't (yet) in the fight.

Most of what I'm trying to do here is figure out how to present this to the players in a way that feels right to them.

[Lunali]

I think what I'm zeroing in on is that I'm used to initiative being an active thing on the part of the PC (not the player, the PC). As soon as I ask Bob('s player) to roll initiative, I've adopted the mindset that Bob is an active participant in the fight. Except he's not. The fight has started around him but he's technically unaware of it. Sort of.

He is an active participant in the fight, he was caught with his guard down, but he is full aware of the fight. Once again, surprised should have been left being called flatfooted in this edition.

[MaxWilson]

From a narrative standpoint, the problem with a "Warn" action is that Mary calling out, "Bob! Bugbear!" is no more going to get him to react to it faster than is him being attacked by the Bugbear.

The actual point of weirdness is that, if Mary were on the Bugbear's side, it would mean Bob wasn't surprised because he is aware of Mary. Though in such a situation, it would likely be that the DM substitutes a Deception roll on Mary's part against Bob's Insight to see if he realizes she's leading him into an ambush rather than is the friend and ally he thought she was, and substitute that.

The Surprise rules do hinge on being able to identify who's on whose side. In that, if the only people you're aware of are those on your side, you're Surprised.

It's not actually weird to be wary of a potential threat. If Bob has known (or believed) all along that Mary was likely to knife him in the library, he's not going to be all wrapped up in reading his fantasy novel when Bugbear suddenly kicks down the door and starts swinging a halberd at his head. The threat is coming from a different direction, yes, but he's not going to be in Condition White, to use the lingo of self-defense.


In condition white, you are relaxed and unaware of what is going on around you. Ideally, a police officer is only in white when asleep, but realistically we often drop our guard when we are at home or in some other environment we assume to be safe, like the squad room.

Since even police stations have been attacked, it is better to be alert even when you are in your “lair.” As the Lakewood coffee shop ambush proves, you simply cannot be on white when you are in uniform, whether on duty or off.

If you are attacked in condition white, you may very well die – unless you are lucky. I prefer not to depend on luck.

In condition yellow, you remain relaxed, but are aware of who and what is around you. This means you are paying attention to the sights and sounds that surround you whether you are at home or moving in society.

Condition yellow DOES NOT equate with paranoia or any other irrational fear of persons or places. Instead, you simply have moved your alertness to a level of attention that will prevent you from being totally surprised by the actions of another person... You will also be running a cursory “what if” mental visualization of where a threat could appear and what your reaction(s) should be.

Source: https://www.policeone.com/police-trainers/articles/coopers-colors-a-simple-system-for-situational-awareness-Np1Ni2TbRj9EkGUN/

I would suggest that adventurers should usually be in condition yellow or higher for reasons similar to police officers, but partly for game balance reasons I would rule that being in conditions yellow or above is strenuous enough not to count as a short or long rest, unless you have the Alert feat or similar that makes constantly running "what if?" scenarios in the back of your mind second nature. Therefore, surprise is by definition possible whenever you are relaxed enough to be attempting a short or long rest.
Note though that there is still a huge difference between losing six seconds getting your threat response together (momentarily "surprised") and being caught unarmored, unarmed, in the open and on your own (while taking a bath or out shopping). Even by these slightly-unrealistic rules about resting, while resting you can stay fairly safe simply by e.g. making sure you rest in an area with good enough visibility to see threats when they're still at least six seconds away.


I think what I'm zeroing in on is that I'm used to initiative being an active thing on the part of the PC (not the player, the PC). As soon as I ask Bob('s player) to roll initiative, I've adopted the mindset that Bob is an active participant in the fight. Except he's not. The fight has started around him but he's technically unaware of it. Sort of.

Ideally, a surprised creature would not roll initiative while everyone else does. Except that the rules then say it gets to use its reaction once its turn has passed, and we need init to know that. Maybe a middle ground approach is to use the surprised creature's initiative score to determine their turn position for that first round, then have the player roll starting with the second round, and go from there. It allows the RAW rule to work as intended but also implies the creature's head really isn't (yet) in the fight.

Most of what I'm trying to do here is figure out how to present this to the players in a way that feels right to them.

Give up on PHB initiative then, it will never feel right. Even when surprise isn't involved, PHB initiative is still oriented towards forcing players to do nothing for the majority of a round, not even talk to the DM, (because it's "not your turn" yet) and surprise just adds another full round of doing nothing on top of that.

Switch to a concurrent resolution system that makes it natural for the DM to declare, "Suddenly, out of nowhere, a halberd slams into your chest. Bob, you take 18 HP of damage and are too flaggerghasted to react, but Alice, you were suspicious of this library all along. The halberd is held by an evil-looking bugbear. He's getting ready to hit Bob again. What do you do?"

It's one thing to ask Bob's player to wait patiently for a moment while Alice resolves her action for round 1, and then have them all declare actions for round 2 and perhaps roll initiative. But vanilla PHB rules would make Bob wait three times as long (Alice, skip Bob, Bugbear, Alice, Bob) to even do anything, even though this is a very tiny combat with only three combatants. In a bigger combat it is even worse for Bob's player.

[Segev]

I don’t see how the proposed concurrent system would play out any differently. The DM still is to resolve his monsters’ actions. He still has to listen to Mary and then to Bob (or Bob and then Mary), and he still has to resolve all the rolls they make, record the numbers, etc.

Saying it’s “concurrent” doesn’t magically give the DM more ways to split his attention, nor do away with the dice rolling and number crunching and analysis paralysis.

[EggKookoo]

Give up on PHB initiative then, it will never feel right. Even when surprise isn't involved, PHB initiative is still oriented towards forcing players to do nothing for the majority of a round, not even talk to the DM, (because it's "not your turn" yet) and surprise just adds another full round of doing nothing on top of that.

It's not the sequential initiative that doesn't feel right. It's how surprise interacts with it.

[MaxWilson]

I don’t see how the proposed concurrent system would play out any differently. The DM still is to resolve his monsters’ actions. He still has to listen to Mary and then to Bob (or Bob and then Mary), and he still has to resolve all the rolls they make, record the numbers, etc.

Saying it’s “concurrent” doesn’t magically give the DM more ways to split his attention, nor do away with the dice rolling and number crunching and analysis paralysis.

Bob doesn't have to wait for the enemy on round 1 because the DM already rolled the dice and resolved the attack as part of announcing the hook, and Bob doesn't have to wait for Mary on round 2 because Bob and Mary both have immediate access to the DM to announce their plans. The only waiting he's doing is on turn 1 while Mary resolves her attack, and on turn 2 Bob only needs to wait to announce his die-rolls to the DM and to Mary. In practice it's very fast, and the players wind up interacting with each other as much as with the DM as e.g. Bob and Mary discuss how to handle this bugbear.

If you force analysis paralysis and die-rolling to be sequential for each player, they take even longer than if you do them separately.

In short: more people to interact with (including a party leader in large groups who can make sure everybody has a chance to declare their action, so the DM can focus on roleplaying) + not needing to wait for other players to roll dice before you can begin analysis = shorter cycles, less waiting.

I'm going to steal a diagram from F# instead of doing ASCII art.

Instead of this: https://www.compositional-it.com/wp-content/uploads/2020/01/Timeline-sequential.jpg

Instead of "worker" pretend it says "turn." Bob is #2, 5, and #8, and his activity during round 1 (turn #2) is very brief because he just acknowledges that he's surprised, maybe yells something, and then waits for Mary and the Bugbear to finish. Also, pretend Worker #0 is already done before the diagram starts.

you get something more like this https://www.compositional-it.com/wp-content/uploads/2020/01/Timeline-batch-custom.jpg

You can see that there's still some waiting around required because there ARE synchronization points required each round. It's not a full-on real-time combat system. But in practice it is faster and better at encouraging player-to-player interaction than sequential turns are, and Bob spends less time in an idle state playing with his phone.

[Segev]

Ah, so this is like my "write down what you do" thing. Because that's the only way I can think of to keep it from being something that forces itself back into sequence anyway.

Also, knowing the players I've gamed with in umpteen groups, there will be those who are quite offended that the bugbear got to go first without having to roll initiative.

[MaxWilson]

Ah, so this is like my "write down what you do" thing. Because that's the only way I can think of to keep it from being something that forces itself back into sequence anyway.

Also, knowing the players I've gamed with in umpteen groups, there will be those who are quite offended that the bugbear got to go first without having to roll initiative.

IME players don't mind if the bugbear goes first as long as it seems fair, especially if it doesn't guarantee the bugbear goes first on every round.

If someone objects you can point out that they were in the middle of a long already-declared action, like continuing to read the library book. Letting you *change* your action at the cost of an extra second or two maintains the narrative die without putting the bugbear at a permanent advantage for the whole combat, but is also more fair than making the player wait for a new round to declare a new action.

In my game I call waiting to see what happens before declaring your action for the round "Delaying", and the rule is that if you're not definitely doing something else and you're ready for possible threats, you're doing an implicit Delay: ready to act, but not on a hair trigger for a particular readied action. That bugbear that hit you? Now you can hit him back.

So surprise per se is rare and primarily occurs to PCs only when you're suddenly attacked while resting. Surprising monsters is more common, if the party is sufficiently stealthy and the monsters are complacent.