Maybe I'm treating the question as more specific than it needs to be, but in your example it appears to me as though player 1's agency is being denied. Twice she stated her action clearly, and yet somehow she failed to get the results of that action back from you.
You don't have to wait until all players are agreed before allowing a player to act. Now, OK, you don't want the party to split unless really necessary, but until they're actually out of earshot of each other, let each player get on with what they want to do. Running through your example hopefully will show the principles I mean:
GM: Okay, you're standing on the edge of the forest.
Player 1: Okay, I ready my bow and start forward.
GM: fine
Player 2: Hold on, I'm still talking to this other guy.
GM: Player 1, are you scouting ahead alone?
Player 1: I'll take a look, I won't go far until the others catch up.
Player 3: Did we stock up on bread? I need to put on my night vision
goggles.
GM: If the bread's not on your equipment list then you didn't stock
up [or if that's not your style of game, "yes, you stocked up on usual provisions at the supermarket yesterday"]. You now have your goggles on.
Player 4: [Crazy roleplaying thing!]
GM: [Crazy roleplaying response]
GM: Okay, are you guys going into the forest?
Player 1: I am, yeah.
GM: You start to walking into the trees...
Player 2: One sec, I need to cast detect evil.
GM: No problem, but Player 1 was already doing this while you finished with the other guy and that roleplaying thing happened.
... the undergrowth is mostly pretty thick, but you easily find what
might be an animal trail. You'll be able to travel much faster along
the trail if you want to. Player 2, if you've finished your conversation with the other guy, you cast Detect Evil. Nothing in
range. I assume you maintain it as you start into the forest?
Player 2: of course
GM: You can't see exactly where player 1 is, but she can't have got far. Okay, actions from everybody else?
Principles:
- Stay in the conversation
- Respond to actions and requests for information as quickly as possible. But in the case of Player 1 setting off ahead alone, I think it's also reasonable to give the others an opportunity to react and for Player 1 to confirm before doing something that takes a little while and might be foolish. You don't want a conversation that goes, "1: I ready my bow and start into the forest, "2: hang on a sec...", "GM: 1, you fall in a spiked pit and get ambushed by 40 goblins", "1: well, actually, I didn't mean I'd leave the party, I was about to say I'd have waited when 2 said to hang on but you interrupted", "GM: Ah. I totally didn't get that from what you said".
- If something can happen, narrate it happening.
- You don't need to design the adventure specially to avoid delays. Once the players get used to the style, they can keep the action going just by acting. Or they can stop when they really need to, they're in control.
- Don't ask the players what's happening. Ask the players what they're doing, tell them what's happening.
Now, there's a whole other scenario you have to be able to deal with:
GM: Okay, you're standing on the edge of the forest.
Player 1: Okay, I ready my bow and start forward.
All other players: WAIT!!! We need to make a plan.
Player 2: Also, I'm still talking to this other guy
GM: Player 1, are you scouting ahead alone?
Player 1: No, I'll stick with the others until we have a plan.
All players: refuse to move while they spend 30 minutes arguing the
best tactics for hiking.
In this case you have mass analysis-paralysis. There are a few ways to break it:
Specifically tell the players that their precise plan doesn't matter. As soon as any encounter begins you will prompt them for their formation, and you will allow for sensible precautions. This doesn't suit all playing styles, but it saves a lot of time making preparations for things that never happen. The game style issue you want to resolve is, will the players be punished for acting without thorough explicit preparation? If so, then that might be why Player 3 is worried about bread, and it's the nature of a "10 foot pole to check for traps" game.
An extreme version of this is to run the game such that whatever the players suggest is considered reasonable. The world conforms to their expectations rather than the other way around. Then they don't need to plan at all, and the way to reach group agreement is not to wait until everyone agrees, it's to take turns to speak, and agree with and build on whatever the previous person said. They can improvise, they can take turns to contribute, and no matter how stupid what they do is, you will respond "yes, and...". Not "no, because" or "oh, this other thing first needs to happen first", or "you do that but you die because you never said you'd put your armour on". Again, this won't suit all styles, it's not very simulation-y. But they'll stop doing pointless boring things, because any interesting thing they think of is worth saying.
Guide the players through making the plan. You don't want to give them too much OOC information, but usually the characters have expertise that the players don't, and you can bolster their confidence by confirming their guesses and supplying general information at the right times. Confident people make decisions quicker:
GM: Okay, you're standing on the edge of the forest.
Player 1: Okay, I ready my bow and start forward.
Player 2: Hold on, I'm still talking to this other guy.
Player 3: Did we stock up on bread? I need to put on my night vision
goggles.
Player 4: [Crazy roleplaying thing!]
GM: Alright. Player 1, you're going to scout ahead, and you're ready
for trouble, that's sensible. Player 2, sorry, the other guy really
doesn't have anything else to say. Player 3, yes, you're fully
stocked, the night vision goggles will negate the darkness penalty
under the trees. Player 4, I like your style. Do you all fall into
formation behind player 1?
[In your transcript, at this point you said "are you guys going into the forest?". That is to say, you asked the group for a consensus before giving them any feedback on their individual issues. If you do that a lot, it's probably the main reason things get stalled.]
Player 2: One sec, I need to cast detect evil.
GM: I'm fine with that if Player 1 will wait?
Player 1: For one round? Sure.
Player 3: I'll stick close behind Player 1 since I have the best
sight.
GM: Good. 2, Results of Detect Evil are [whatever]. 4, once you're
done invoking the wrath of Gragnar on any fool who dares oppose you,
where are you in formation?
Player 4: Rear-guard, if everyone's happy with that. The Book of Gragnar commands us to, "Pity especially the fool who tries to sneak up you in a forest".
GM: Sounds good. Doing that thing you normally do when you're
rear-guard? That leaves player 2 in the middle. Onwards!
Design adventures so that the party doesn't have a lot of time to waste. You don't have to railroad, but make sure that there is always something happening to them. They can deal with it however they like, but they must act. After all, arguably if there's nothing happening to them and they're free to delay as much as they like then that's practically the definition of "downtime between sessions" ;-) So, players don't arrive at the edge of a forest at their own leisure, they arrive at the edge of a forest as a consequence of dealing with the previous problem:
GM: Okay, you're standing on the edge of the forest. You can hear those
enraged villagers with pitchforks approaching, but as you already know
they're very superstitious about the forest, and you suspect they
probably won't enter it this close to dark.
All players: start to plan
GM: [after a couple of minutes or so, representing the party's head start] The villagers have crested the hill behind you, and the front few break into a run. They're in missile range and will reach you in a minute or less, but then again you never thought much of their combat ability.
All players: No, we're not slaughtering the whole village! We get into the forest.
Finally, be aware that "keeping the game moving" doesn't need to mean physical activity or plot progression. If Player 4 goes off on a crazy roleplaying thing that the other players react to and enjoy, then it's irrelevant that it isn't part of your plan for the session. It's as much a part of the game as anything you invent. So encourage it to play out properly, and as long as it's not boring the forest can wait. Similarly, if the players just plain enjoy bickering in character over their plans, you can let that be part of what defines that particular campaign. Just build 20-30 minutes per significant group decision into your session plan. Less work for you!
To do this you need to get buy-in from the "more action-oriented players". If all they like is combat then that's pretty much a non-starter, you can't run a game this way for them. But otherwise you need to stop them tuning out by soliciting their responses, and making those responses matter in the conversation. A frustrated character who spends the whole argument saying, "we need to stop arguing about this and get into the forest" in 10 different ways is still an active player. One trick is to keep track of who is speaking, and if someone hasn't spoken for a while specifically ask them, "what do you think, what are you doing while this is going on?". That gives the player the freedom to take a turn in the conversation, or for that matter to wander off into trouble if they like. If you frequently find that you ask a player what they're doing and they say "nothing" and tune back out again, then you still have a problem and need to address personally with that player what they need from the game. Some players enjoy spectating for some of the time, and might look tuned out when they aren't, so you do need to ask.
What is Agency?
I personally define agency by three criteria:
- The player has control over their own character's decisions.
- Those decisions have consequences within the game world.
- The player has enough information to anticipate what those consequences might be before making them.
What does that mean?
To elaborate on those conditions, I'll give examples of ways that agency can be violated.
- A group of goblins surrenders to the PCs. Alice decides that her character, Johanna, would rather just kill all the goblins and says that she starts executing them. Devin (the DM) decides that Johanna wouldn't do that and forbids the action. In this scenario, agency condition 1 is violated because Alice is no longer in control of her characters actions.
- Devin has planned for the PCs to be ambushed by bandits on their way out of town - and he foreshadows the ambush by having the PCs overhear in the bar that a merchant got attacked by bandits on the road they're about to travel on. The PCs look at the map, and choose to take a longer route to avoid the bandits. Devin decides to spring the bandit ambush on them anyway - Devin moves the bandit lair on the campaign map so that they will still encounter it. Here, agency condition 2 is violated because the PCs decision to avoid the bandits was made meaningless.
- The party is making a plan to infiltrate a dungeon. Devin decides that the evil wizard who rules the dungeon is Scrying them and therefore knows their plan. Unless there is a reason which the players could have been aware of for why the Wizard might be scrying them at that exact moment, this is an agency condition 3 violation.
- The party encounters a troll, which keeps regenerating on them. Carl remembers that trolls are weak to fire or acid, and so he has his character, Percy, attempt to torch one of them after it goes down. Devin says that Percy wouldn't know trolls were weak to fire and so forbids the action. Agency condition 1 is violated.
- Same as above, but Carl convinces Devin that that's railroading. Devin still thinks it's unfair for a player to use that knowledge, so he changes the trolls into homebrew "trulls" which are like trolls, but their regeneration is countered by lightning instead of fire or acid. Agency condition 3 is violated, because the players have no reason to expect that lightning would behave any differently than other damage types.
And what is it good for?
Imagine if any one of those examples above led directly to a player death - or worse, a TPK. Any of these situations could be a group-killer:
- The GM forbids the player from killing the surrendered goblins - then has one of the surrendered goblins stab them in the night.
- The GM ignores the players' decision to avoid the bandits - then the bandits kill someone during the surprise round.
- The players go forward with their plan that assumes they will have the element of surprise. They get ambushed and die.
- The GM forbids the player from using fire to kill a troll, which allows the troll to mop up the party with impunity.
- The GM replaces the trolls with trulls that aren't weak to fire, and then the trulls make mush of the party before the players figure out the switcheroo.
A defeat in which the players had no agency will always feel arbitrary - the players will feel as if the GM cheated them. A defeat that follows from Agency is one that the players can feel responsible for - because they knew the risk (condition 3) and did it anyway (condition 2) by their own free will (condition 1).
Is more Agency always better?
Probably not, but it depends on the group. I maintain that an undesired outcome (especially a character death) will never feel satisfying unless the player had sufficient agency to prevent it. Outside of that bubble, however, there a numerous other good things that it might be worth it to give up some agency for.
As a comment KRyan pointed out, running a game with absolute 100% agency would mean that the GM is never allowed to ever surprise the players - and many players want to be surprised sometimes. Additionally, agency might be worth suspending to prevent disruptive or egregious metagaming, or to make an overly-gregarious player to share the spotlight.
Can I run a game without agency?
Not completely - that would be silly. If you wanted to completely squash agency, you'd have to dictate to players what they do on their turns for them in combat. However, there's an entire school of play (called Participationism) where the DM basically controls the party outside of combat and dungeon exploration.
If you go down the road of playing a low-agency game, you should first make sure that your players are on-board with it and won't be trying to make decisions for their characters outside of the rails. Second, you make sure that you clearly define where the border is between your dictatorship and their agency, and make sure that you respect the line. Lastly, make sure that either any defeat they incur is a result of their agency, or you have their absolute trust that if you lead them into a defeat/setback, they will believe you that the game will be better off for it.
Best Answer
Yes -- if you use that specific phrasing, "your character would not do that", you are denying their character's agency. The player is an authority over what their character wants to do; your authority is over what the character can do. Rather than tell the player that their character doesn't want to do something, instead express it as their character's inability to do something.
The above is still sort of inadequate, because a motivated player can find a way around it -- for example by asking the wizard to polymorph the druid's leather armor into plate mail, or by asking the fighter to knock the druid out and dress them in plate armor while they're unconscious.
The problem we're having here is arguably caused by a failing of the 5e Player's Handbook, which states that the druid will not wear metal armor, but doesn't otherwise describe the consequences. To fix the problem, you need to do some worldbuilding. What happens to a druid that wears metal armor, perhaps against their will? You need to fill those details in. Ultimately, you want to tell the druid something like: "well, technically you can choose to wear the metal armor, but it's a really bad idea because you'll face the following consequences..." Make up some consequences so horrible that no sane druid would ever wear the armor.
or:
or:
Now that you've phrased it like this, it's no longer denying the character's agency -- now you're technically offering them a choice, but a choice that is heavily weighted so that it's functionally identical to the Player's Handbook rule.