PDA

View Full Version : For Every Action A Reaction



Cluedrew
2021-12-15, 10:26 PM
Every time a character (well PC) does anything there should be a reaction from the world. Not equal and opposite, this isn't about physics but about the game progressing, but still something should happen and things should change. Sounds simple? It is, but it also seems to be an easy thing to miss.

Oh before I continue, I don't mean quite actually "does anything", but I couldn't find a neat way to fold it into the original statement so here it is as a tacked on qualifier. On one side you have flavour actions like: "After setting up camp I pull out a block of wood and do some carving." On the other side there are big divisive actions like: "I push the trigger and blow up the building." If it is something you might think of resolving with rules that is a good sign that it is significant, but that isn't exactly right either. I don't have a perfect way to define that difference, so let's say this is (as they saying goes) more of a guideline than an actual rule.

So there are two main reasons there should always be a reaction. The first is simple so I'm going to knock it out real quick. The alternative is boring. The alternative by the way is "Nothing happens." Useful maybe as a beat here and there, but best used in moderation.

The other takes a bit more explaining because it has to do with the core loop of games. It applies to a degree to almost any core loop but I'm going to focus on one of the classic role-playing game loops: GM and (non-GM) player.

Action with Reaction:
GM: "This is the current situation, what do you do?"
Player: "The character does this action."
GM: "This is the new situation, what do you do?"

Action without Reaction:
GM: "This is the current situation, what do you do?"
Player: "The character does this action."
GM: "This is the same situation, what do you do?"

The difference is the reaction, the situation's action in response to the player's action. This may seem unnecessary at first glance; after all don't the player's actions force a change/reaction of some kind? There are two cases that can slip though, when someone tries to do something and fails, or when someone tries to stop something and succeeds.

The first can be described with a super-basic version of any binary pass/fail roll. On a pass whatever you were trying to do just happens, on a fail it doesn't. In other words, nothing happens. Most people will layer on some extra consequence to change the situation a bit, take some damage from a fall, waste time fiddling with the lock, that sort of thing. This leads to the "fail-forward" model where every result of the die has some notable impact on the situation.

Oh, and don't wait for "the second", I tried to think of something unique and interesting so say about stopping something and succeeding, but its just the same but reversed. But there is a second major way to solve the problem.

The other solution is to put everything in a frame-work so that the attempt itself changes the situation. The simplest example of this is a turn system. Succeed or fail your turn is now over, and by the time it comes around the situation is going to be different (or if it is not, that is a different problem). This has a very different structure in terms of the rules, but the effect is the same, it keeps things changing.

So that's it, my big statement, why I think that and how many systems handle it. Actually, the fact that many systems handle it means you might not have to do much about this in a game. But it is something to keep in mind for adjudication and any homebrew you might do. Plus it is of some theoretical interest and I have been surprised many times what topics have sparked large debates on the forum.

Fun Fact: I remember enough about writing this post to be confident I did it. But forget when or why I did it. I found it mostly written in a text file, sitting there for who knows how long.

Duff
2021-12-16, 12:04 AM
I feel like you've stated the obvious. Which means I might well have missed your point.

Can you give an example where you feel this does not (or didn't) happen?

Razade
2021-12-16, 12:13 AM
Every time a character (well PC) does anything there should be a reaction from the world. Not equal and opposite, this isn't about physics but about the game progressing, but still something should happen and things should change. Sounds simple? It is, but it also seems to be an easy thing to miss.


You should look into Powered by the Apocalypse games, this is one of (if not the primary) philosophy behind them.

Sneak Dog
2021-12-16, 05:03 AM
Player: 'I'd like to listen at the door real quick.'
GM: 'You don't hear anything from the other side.'

If one listens at the door for a quick 10 seconds and doesn't hear anything, that's fine. Even if it's boring, it's part of the player exploring the scene and gathering tidbits of information. An informed player will then make a significant choice, and then the scene will change.

The player opens the door slowly and quietly. There's an empty room behind, with two open doors, one left, one right. From the left a soft snoring can be heard.

Khedrac
2021-12-16, 06:06 AM
I feel like you've stated the obvious. Which means I might well have missed your point.

Can you give an example where you feel this does not (or didn't) happen?

I'm with Duff here.

Also, I think that module writers have to limit the number of options they give to the party that actually make a difference. If they don't, then every module becomes infinite in size and cannot be published.
However, that said, it is then up to the GM to ensure that everything the players try gets some form of response even if there is no practical effect.

E.g.:
GM: You see a man standing on some sort of platform haranguing the crowd, he seems to be trying to work them up into doing something that sounds violent.
Player: I try calling out to interrupt him and turn it into a debate.
GM: The man ignores you, in fact you quickly realise he is deliberately ignoring you - he has no intention of engaging in honest discourse.
Player 2: OK then, I cast a magic that will silence everything without 10' of the platform the man is standing on.
GM: Just before you complete your spell the man seems to reach some sort of climax, the sudden absence of noise from the man definitely startles the closest people, but most of the crowd are beginning to stream away shouting and waving whatever they have in hand. Make an observation skill roll.
Players roll.
GM: you notice the instigator slipping away from the riot he has set in motion - it looks like your spell helped him even while it surprised him.

In the above example a not-very-well written module calls for the players to witness someone start a riot and then sneak away without being able to interrupt (better writing would have them only witness the very end of the speech without time to interrupt, and be able to get the story from other onlookers). The GM should still be able to have the desired effect happen without allowing the players to materially affect it, but with the players able to adjust the details - akin to the DM colouring in a black and white sketch and the players get to dictate some of the colour choices.

Is it "railroading"? Yes, but that is not always a bad thing, especially if the "meta" choice is "play what the DM has prepared or come back next week when the DM has prepared something else".

icefractal
2021-12-16, 05:13 PM
IDK, that example would definitely bug me if it happened in a game.
The initial attempt failing, I feel like there should have been a check. It won't be a debate, but a good enough orator on the PC side should be able to undercut him regardless of cooperation. Might be a rather difficult check, but it shouldn't be flat-out impossible.
But I think the "oh he just finished right before you did that" is worse, it's making it obvious that this is really a cutscene which you were deceiving the players into thinking was interactive.
And "your spell actually helped him, lol"? That's adding insult to injury. :smallyuk:

I don't like cutscenes much in any case. Yes, they're used in other media, but different things work in different formats, and I don't think cutscenes work well in a TTRPG. But if it's one where the PCs obviously have no way to intervene (they're too far away, they're seeing it happen by TV/scrying, it's on a scale they can't operate at, etc), then I'm ok with it.

Video-game-style "you're standing right there but you can't do anything because plot" cutscenes? Those suck even in video games, and more so in a TTRPG.

KorvinStarmast
2021-12-16, 07:10 PM
IDK, that example would definitely bug me if it happened in a game.
The initial attempt failing, I feel like there should have been a check. It won't be a debate, but a good enough orator on the PC side should be able to undercut him regardless of cooperation.
I don't get this sense of entitlement. Particularly if you understand crowd behavior.

Might be a rather difficult check, but it shouldn't be flat-out impossible.
OK, with that qualification, a chance for a success but at long odds sets up a risk/reward situation.

But I think the "oh he just finished right before you did that" is worse, it's making it obvious that this is really a cutscene which you were deceiving the players into thinking was interactive.
How is that deception, to set up the main tension during the scene/event? That's the attitude that I don't get.


And "your spell actually helped him, lol"? That's adding insult to injury. :smallyuk: No, it is case of "not everything the players do is gold plated." Your implied demand that it be so is what I don't get. A wide variety of actions that players do make inadvertently help the opposition. I see it nearly every week in play.

I performed a classic gaffe like that a few months ago when I was flying, and cast a web at the enemy second rank, thinking I was doing something brilliant but of course when I cast that concentration spell, I stopped flying, and when I fell the concentration I need to keep the web up failed due to the falling damage, so in one brilliant move I took myself out of the fight and failed to do any harm at all to the enemy at all.

With that in mind, can you give me an example of what you would do with the previous scene (the speaker setting off a riot and sneaking off) that is a fail forward map?

Cluedrew
2021-12-16, 07:42 PM
To Duff: I thought is a fairly obvious point, at least it is in retrospect, but I have been surprised by "obvious" facts that have turned out to be very controversial in the past. But more to the point I think there is more stuff behind that obvious point worth talking about. Such as the two common strategies to avoid it.

Speaking of which, to my knowledge D&D is spotty (particularly in terms of what is in the DMing advice), but I believe that D&D has a solution for this in combat - the turn system - but I am not sure if it has a general solution for out of combat. Required the GM to adjudicate one. Is that correct?

To Razade: I've played several Powered by the Apocalypse systems, none of the big ones though. They definitely have the fail-forward system going on.

To Khedrac: I have been assuming at least a good faith effort on behalf of the GM. Mostly discussing the difference between the game moving forward and it standing still. Not so much the game going forward with and without player input. I suppose you could try to address it a bit by having some definite results in some of the player actions, but a belligerent GM can always make things difficult.

icefractal
2021-12-16, 09:20 PM
How is that deception, to set up the main tension during the scene/event? That's the attitude that I don't get.
By presenting a situation where the PCs were there able to interact, and asking the players what they do, you create the expectation that the action they're choosing could matter. If you've already pre-determined that the speaker will succeed at starting the riot then what the PCs do matters not and you shouldn't pretend it does.

So the exact moment that the PCs try to silence him happens to be right when he's finished speaking? And if they'd done that first, before the attempt at debate, would it have mattered? The way it's framed above looks like it wouldn't, the outcome is already pre-chosen and the "reactions" are simply different segues to the same place.


No, it is case of "not everything the players do is gold plated." Your implied demand that it be so is what I don't get. A wide variety of actions that players do make inadvertently help the opposition. I see it nearly every week in play.Yes, it works in the NPC's favor - if you as the GM decide the timing is perfect for that.

When you railroad, you assume responsibility for what happens - it's not the player making bad decisions if they don't have agency over those decisions, it's you as a GM choosing to have them fail (and in this case also look dumb).



With that in mind, can you give me an example of what you would do with the previous scene (the speaker setting off a riot and sneaking off) that is a fail forward map?
Fail forward? The scenario you present is fail only. There's no other path.

Simply having that same scene but the PCs can potentially change the outcome would already be fail forward, as you have the fail path right there. What it would need to have is a success-forward route, or probably a few of them - sabotaged the speaker, diminished but didn't stop the speaker, didn't stop the speaker at all but did capture or follow him afterwards.

Or, if you just want the riot to guaranteed happen - don't have the PCs get there in time to stop it. They get there once people are already rioting. This could be deceptive (if "getting there in time" is presented as important), but if it's just "As you walk through the ___ district, you hear the sounds of shouting and smashing wood ..." then no issue.

Khedrac
2021-12-17, 03:38 AM
IDK, that example would definitely bug me if it happened in a game.
The initial attempt failing, I feel like there should have been a check. It won't be a debate, but a good enough orator on the PC side should be able to undercut him regardless of cooperation. Might be a rather difficult check, but it shouldn't be flat-out impossible.
But I think the "oh he just finished right before you did that" is worse, it's making it obvious that this is really a cutscene which you were deceiving the players into thinking was interactive.

I agree it's not good DMing, and if the DM has time to write their own material or expand on pre-written they absolutely should allow the party to have a chance of affecting things - a good example of this happened to me in a Call of Cthulhu game where the party was supposed to turn up to stop a ritual in progress at Stonehenge (cue fight where the leaders of thecult carry one summoning and the party have to get through their minions to stop them) in practice the cultitsts turned up to find the party between them and stonehenge with better weapons and prepositioned in cover - only the wo leaders tried to get to the stones and they didn't get very far.
But that's a good DM who had prepared the material and knew us and our penchant for not following the trail of breadcrumbs. I was talking about a pre-written module where the author doesn't expect the party to even witness the mob formation, but includes it to give the DM something to work with if needed. Writers simply cannot cover every possibility, and the DM's job becomes enabling the players to seem to have an effect while carrying on with the module.
This gets even worse in Living Cmapaigns where a series of modules can be based on outcomes that the party did not have happen in the earlier modules (I think they usually go with feedback to find the most common outcome).

Quertus
2021-12-17, 10:14 AM
I believe that D&D has a solution for this in combat - the turn system - but I am not sure if it has a general solution for out of combat. Required the GM to adjudicate one. Is that correct

Afaik, yeah. And that's related to one little thing that bugged me about the OP. "I X." "OK, now here is the state of the world" is bad.

What?!

OK, no, actually, it's a great gameplay loop. But it's incomplete.

Another completely valid possibility is

"I X"

"OK, while you are doing X, you notice Y."

With an implicit expectation of, "I continue X", "I continue X taking Y into consideration", "I continue X and add Z", or "I cease X, and do Z".

I guess… Hmmm… I view the game as a timeline of intended actions. These intentions can change as the actor perceives the world around them. It's how I run my NPC plots.

Imagine, for instance, a tick-based combat game. A really detailed one. Where you know that, to cast a Fireball,

Retrieving sulfur will cost you one tick. [Visual 2, Olfactory 2, component pouch, bindable (arms) 1]

Retrieving bat guano will cost you one tick. [Visual 2, Olfactory 2, component pouch, bindable (arms) 1]

Rolling them into a ball will cost you one tick. [Visual 1, Olfactory 3]

Casting the spell will cost you 3 ticks. [Visual 5, Auditory 7, recovery 7, bindable (arms) 5, concentration, Spell (I'm having "Spell" do a lot of work here, that should probably be broken out into separate tags in a real system)]

Let's say your character intends to do those 4 steps to cast Fireball.

Let's further say that the system gives default actions, like "dodge" or "dive for cover!", that one can abort to.

On tick 1, while you're retrieving sulfur, someone tries to stab you. They've got a mundane dagger, you trust your Stoneskin to hold them off, you don't take any defensive actions.

On tick 2, however, a volley of arrows is released by formerly hidden archers. You don't have time to get Protection from Missiles up before the volley gets here in tick 3, and don't feel like getting turned into a pincushion, so you abort to Dive for Cover.

Because you aborted during a [component pouch] action, you have to roll on the "component mishap" table. Curse your luck, there was a mishap: you lost 6 uses of bat poop - mostly on yourself. This is a [visual 5, olfactory 7] mishap. But you did successfully obtain your component.

Dive for Cover has a recovery of 2. So, on tick 5, with an arrow sticking out of your shoulder, you're rolling that ball.

On tick 6, you consider standing up, but decide you'll just roll out of cover at the end of casting Fireball, and chuck the spell at what you can see. So you begin casting.

Then, on tick 8, you see the grenade roll behind your cover.

Normally, someone would choose one of several options here.

Yell "grenade"? You can't, not without breaking your spell.

"dive for cover"? you can't, because you're prone. And that would kill your concentration on Fireball, making you lose the spell and eat the recovery for nothing.

"Dive on the grenade"? Lose the spell, and your life, and, since you're starting prone, probably fail to save anyone else? Nah, bad plan.

However, in this system, a grenade deals 2 types of damage: physical shrapnel and AoE explosive. The physical damage can be soaked by your Stoneskin, and you have Evasion for the AoE damage. Normally, Evasion (or "rolling reflex vs AoE") would *also* break concentration, but you took a special feat / class feature / whatever that says it doesn't. You cross your fingers and roll reflex as the GM rolls damage.

That's the kind of back-and-forth I'd like to see in combat, and try to create in the general gameplay loop.

EDIT: my point? That each time anyone's actions or effects become visible to another actor, they should have the opportunity to reevaluate their current course in light of that new information. PC or NPC.

Vahnavoi
2021-12-17, 02:35 PM
Keeping track of time usually covers the same rule space as keeping track of turns, so "nothing happens" has a resource cost to it. There's also situations where a null result like "nothing happens" provides useful information. Or is useful in other ways - internal motivation requires some idling to manifest, too much feedback from part of the game master or system screws with it in unexpected ways.

KorvinStarmast
2021-12-17, 05:37 PM
Fail forward? The scenario you present is fail only. There's no other path.
it wasn't my scenario, someone else presented it and you made a snide response to it.
Nothing you responded to me with adds anything new, nor covers any ground that isn't decades old, but thanks for the effort in any case. But I wonder if the scenario as presented was something like what Kedrac pointed out. Any DM can be faced with that.

I was talking about a pre-written module where the author doesn't expect the party to even witness the mob formation, but includes it to give the DM something to work with if needed. Writers simply cannot cover every possibility, and the DM's job becomes enabling the players to seem to have an effect while carrying on with the module. That's a different piece of the DM art, yes.

This gets even worse in Living Campaigns where a series of modules can be based on outcomes that the party did not have happen in the earlier modules (I think they usually go with feedback to find the most common outcome). A1 - A4 in AD&D 1e is a good example, but they were first used for con/tournament play.

Cluedrew
2021-12-17, 09:08 PM
Afaik, yeah. And that's related to one little thing that bugged me about the OP. "I X." "OK, now here is the state of the world" is bad. [...] it's a great gameplay loop. But it's incomplete.No, it is complete, it is just abstract/high-level because everything you mentioned can be folded into the new situation. Now if you think it is too abstract then that is different. But even at this level it shows the problem.

Also, my question is about what is presented in the rule-books of D&D. I know a lot can be done with that cycle but I was just wondering if "D&D by the book" calls much of that out.

Quertus
2021-12-18, 09:07 AM
No, it is complete, it is just abstract/high-level because everything you mentioned can be folded into the new situation. Now if you think it is too abstract then that is different. But even at this level it shows the problem.

Also, my question is about what is presented in the rule-books of D&D. I know a lot can be done with that cycle but I was just wondering if "D&D by the book" calls much of that out.

Now, granted, I’ve had a fever, so maybe I’m just being dumb, but… I don’t get it.

So, let me explain what I mean when I say that I don’t get it.

I’ve played with GMs who use that loop. It goes something like this:


I craft {item}

Ok, while you were crafting {item}, there was a zombie apocalypse that killed off the town, and then a fire, that killed the zombies and burnt down the town. You now have {item}, and your tower is all that remains of the town - possibly all that remains of humanity.

Or, the example I had wanted to use,


”The situation is X.”

“I take over the world.”

“…”

And… you aren’t interested in general discussion of the topic of responding to actions, only in what’s printed in D&D rules? :smallconfused:

KorvinStarmast
2021-12-18, 09:38 AM
Also, my question is about what is presented in the rule-books of D&D. I know a lot can be done with that cycle but I was just wondering if "D&D by the book" calls much of that out. Which edition?

Cluedrew
2021-12-18, 10:44 AM
To KorvinStarmast: Any edition is fine, just say which.

I am also happy to hear about other systems. I could talk about Powered by the Apocalypse does say how things go wrong, in broad strokes at least, to really reinforce the fail-forward system. Somewhat unique in that those are the only fail-forward systems I know that have individual rules for each "skill roll".

KorvinStarmast
2021-12-18, 03:00 PM
To KorvinStarmast: Any edition is fine, just say which.

I am also happy to hear about other systems. I could talk about Powered by the Apocalypse does say how things go wrong, in broad strokes at least, to really reinforce the fail-forward system. Somewhat unique in that those are the only fail-forward systems I know that have individual rules for each "skill roll".

I don't have a perfect way to define that difference, so let's say this is (as they saying goes) more of a guideline than an actual rule. As a general observation, every edition of D&D handles that, though I'll not speak for 3.x nor 4.x as my system mastery in both is low and nil respectively.
(I'll need to dig into Tunnels and Trolls for it's loop, as it was a little different than D&D but our campaign got curtailed before I got used to the rhythm).
The 5e loop is simple as laid out.

DM describes the environment
Player/players describe their actions/intentions/attempts
(Dice may or may not be rolled)
DM narrates the outcome and you have just established the new environment/situation, or it hasn't changed much and the players either make new decisions or try something else.

But, it gets complicated somewhat by the reaction (as a mechanic) which is a turn interrupt, and even moreso by a whole series of meta features/skills that disrupt the loop.

Examples include the shield spell and the bard's cutting words.

Each of them creates a retcon of a result that disrupts the flow from steps 1 through three. There are other features that do this, but these two are the ones I have seen the most.

The reaction itself is the game world, or the player, reacting to the action loop in situ, but we are at the micro level here since that is combat.

For the environment to react you need to describe what the environment is:
A bridge?
A traffic jam?
A storm?
A rabid dog?

What is it that's "not the player" that is reacting to the player? (Which makes your question very soft around the edges).

An appealing element of using guidelines as a preference to rules wherever one can is the Play and Find Out idea: the world reacting to unforeseen choices and decisions by the players needs some room to move and to flex. Constraining it to an if/then statement is what computer games are for.

As I browse through Worlds Without Number, I am finding some interesting world building schemes and faction play that I need to experience in a campaign before I try to figure out if it's along the lines of what you are looking for. There is some nice macro/meta play that moves the game world and not always in a direct response to player actions and decisions, but in a dialable way - some direct responses and some broader in scope.

By the way, IME "D&D by the book" is a great way to ruin D&D play. And the play's the thing.

Cluedrew
2021-12-20, 07:52 PM
Oh, I forgot to actually make a reply. Woops.


(Which makes your question very soft around the edges).Maybe, but its also a pretty narrow question, what do the rules say you should do on a failed skill roll? Without any special rules and out of combat. As far as I can remember its just "when you try to do something make a skill check. If you pass, it happens. If you fail it does not."


By the way, IME "D&D by the book" is a great way to ruin D&D play. And the play's the thing.True enough. Actually that is why this is still a question because I actually have no idea where D&D's rules on the matter end and advice passed around the community begins.

KorvinStarmast
2021-12-21, 12:14 PM
Oh, I forgot to actually make a reply. Woops.

Maybe, but its also a pretty narrow question, what do the rules say you should do on a failed skill roll?
Why do you need a rule for that?

Actually that is why this is still a question because I actually have no idea where D&D's rules on the matter end and advice passed around the community begins.
I don't see why you need a rule for this. Maybe that's our disconnect. The game has a referee for that reason, among others, since proto-D&D in the Twin Cities. The current edition uses the three step approach I mentioned up there, and step 3 is "the DM narrates the result" which works in play.

Cluedrew
2021-12-21, 06:39 PM
To KorvinStarmast: You don't need a rule for it, but some structure, or just advice in the rule-book, can help the referee make good calls. It is the difference between "support" and "allow" if that makes any sense.

KorvinStarmast
2021-12-22, 12:23 PM
To KorvinStarmast: You don't need a rule for it, but some structure, or just advice in the rule-book, can help the referee make good calls. It is the difference between "support" and "allow" if that makes any sense.
I am with you here. :smallsmile: (This is something that Worlds Without Number has that I like: plenty of DM/GM advice).

Tanarii
2021-12-22, 04:35 PM
Sometimes the situation remains the same, because the player attempts an action in which they cannot succeed, and they only get one shot due to a time limitation.

OTOH certainly your point still applies if they can keep taking the action until they succeed, with no state change or negative consequences except time, and time isn't a negative consequence in the situation. In which case, the DM should skip the straight to the state change. That's what D&D 3e's Take 20 and 5e's Automatic Success rule are all about.

Technically if they only get one shot due to a time limitation the situation doesn't remain the same, if you look at the larger picture including whatever's causing the time limitation.

Example:
GM: "This is the current situation, what do you do?"
Player: "The character does this action."
GM: "This is the same situation, but you're now N minutes closer to the next random encounter check. What do you do?"

In 5e, if the player says they will take ten times as long, then instead it would be:
GM: "This is the new situation, but you're now Nx10 minutes closer to the next random encounter check. What do you do?"