I've DM/GM'd for several gaming groups, young and old, and have found that one of the things that separates my younger players from older players is their eagerness to elaborate/collaborate in the storytelling by throwing in added unmentioned details they can call upon to aid them in battle or exploration.

For example, when entering a wizard's study, and told that the room is dusty, and cluttered with magical curios and artifiacts of all sizes and descriptions, my adult players will ask for more specifics, like "Do I see any magical rings that fit the description of the item I'm looking for", whereas my younger players will say "I grab a magical metal detector lying in the corner of the room and start sweeping the area for the ring".

Obviously, you need to be willing to relinquish some control in the second instance, (and perhaps the idea is anathema to serious gamers and DMs) but I've found I enjoy this form of collaboration. My question is, how do I encourage this sort of improvisation among adults without breaking my game (aside from just providing this example)? How do I keep my rulings "fair" when setting limits to how far this improvisation can go?

  • \$\begingroup\$ I don't have anything to put in an answer that hasn't already been said, but here's my experience with giving the players control: gm.sagotsky.com/?p=265 \$\endgroup\$
    – valadil
    Dec 31, 2012 at 18:03

5 Answers 5


How To Introduce Limited Player Narration

I was introduced to this kind of play with Robin Laws' excellent Feng Shui RPG. It explicitly incorporates this kind of "light" player narration (as opposed to full co-authorship as is common in indie RPGs today). I still use it as training wheels to help people get into the mode of improvising. In Feng Shui, not only is use of tactically-correct battlemats contraindicated, but you're encouraged that if you are, for example, in a fight in a pizza parlor, to just say "I snatch up a pizza cutter from the table and cut the fool attacking me with it" without having to ask the GM or roll for the pizza cutter. If it makes sense and is genre appropriate it's fine.

Many people need that help to get started with the idea because most old school trad games definitely come from the paradigm that the player isn't allowed to contribute to the narrative at all - if the GM does not explicitly instantiate something, it doesn't exist. Which has some good attributes - too much player narration actually leaves me cold as a player since I prefer immersive character play - but some light player narration does make the game run more smoothly and be more exciting.

Just Say No!

In fact, in general I tell people they need to man up and improvise those kinds of things themselves because if they ask the GM I'll say no! "If you ask me how high the helicopter skid is, I'll say 'too high.' If you say 'I jump and grab the helicopter's skid,' I'll say 'roll it!' If you ask me if there's a pizza cutter within reach I'll say no..." That breaks them of the mother-may-I syndrome, which slows down the game. Having to say no once in a while is less time spent and more empowering.

Run a One-Shot To Introduce the Concept

With my current play group, I ran a one-shot of Feng Shui before launching into a longer campaign to train them on the mode of play. This gave them an opportunity to experiment with limited narration in a safe short mode environment. Then when we started a longer game in a traditional game system, the techniques port right over. You can use another game, just make sure it's yielding the type of narration you want - if you go too far into indie land then you'll get your normal D&D players saying "I declare the lich king is my long lost aunt and he's really after the strumpets on the moon!" (Fun fact: on Golarion the moon is full of strumpets, for real.) Which is fine if that's the kind of game you want, but that's not this kind of limited-narration environment which remains sim/immersion friendly while still getting a lot of the benefits of a narrative game.

Controlling the Narration

As for controlling the narration, we have very seldom had problems with that. We had the understanding that narration had to be very limited and appropriate - "belaying pin on the deck of a ship" is a valid pickup, "magical metal detector" is not - and the GM can always say no. Generally this is self policing, as "I come up from behind the pizza oven with a rocket launcher!" triggers everyone's cheese alarms and they just frown at the overexcited guy till he comes up with something less stupid (unless you're playing Toon or something very postmodern). All the "yes, but" stuff is fine for certain kinds of games, but it is unnecessary and disruptive for a game trying to be sim with limited narration only. A couple "no"s to set limits is much less hassle for everyone than trying to put conditions on inappropriate things. With adults it really won't come up much at all after some initial boundary exploration.

Mechanics and Limited Narration

Mechanics to dictate this are not needed and in fact I find them to be intrusive and harmful. The goal of limited narration is to be able to improv; if you have rules it's not really improv-ing. Whenever I have to shepherd currency (like FATE points) to do something cool, or make a roll separate from the roll I was going to make anyway to do a stunt (like the rolls for improvised aspects mentioned in the comments), that's lame because it 1) brings in more metagaming, taking away from character immersion and 2) takes away from the velocity of the game, which limited narration is intended to increase.

The one mechanical note I'll make is that people will try innovation to the extent that it's not punished - if you put big penalties on improvising it will tend to work against it. Feng Shui, as this is what it's about, went to the extent that you actually got a bonus to your rolls when doing a cool stunt. In d20 it's easy to rack up "-8 to do that because you don't have the 'slash a fool with a pizza cutter' feat" - make sure and encourage actions you want to encourage with the numbers. Removing a penalty or adding a bonus based on how cool something was is completely within your purview as a GM, despite the bizarre newthink from D&D 3e/4e that says it's not.

In D&D 2e, I had a halfway mechanic that worked - a Luck stat one would roll to see "if there's a pizza cutter". This worked in that campaign only because it was a very, very gritty low power sim game. I wouldn't use it in a normal power game and would just go to the "yep, if it's not dumb you've got it" model.

Though be careful of trying to over-mechanize it like Exalted does as described in @SimonGills' answer - I remember playing Roanoke, where you get a +1 per descriptive word you use on every action, up to +5. It led to some pretty sad results,as people just strung together tortured 5-adjective strings to get the bonus since the rules rewarded "5 words" instead of the desired result of "a cooler game." I prefer just "that's cool you get the +2 bonus" or not, when you start trying to over-define it it turns into a rule to lawyer instead of an improvisation to add to the story. Also, our stunt level is usually at their +4/+6 level - "I scream and hack at the orc" is expected in decent RP, not bonus worthy.


One option is to borrow Stunting from Exalted.

The basic rules are these. When a player describes their action, check it against the following tests:

  • If the player only used mechanical descriptions, he gets nothing.
  • If the player added a new detail about his action, his emotions or his appearance, he gets a +2 bonus on the roll and heals 2hp if he succeeds
  • If the player added a new detail about the environment or used an existing detail in a new way, he gets a +4 bonus on the roll and heals 4hp if he succeeds.
  • If the entire table goes wow when the action is described, he gets a +6 on the roll and heals 6hp if he succeeds.

Some examples

  • +0 Stunts
    • I attack the Orc.
    • I climb the wall.
    • I hide behind the bush.
  • +2 Stunts
    • I scream and hack at the Orc.
    • I pull some chalk out of a pouch, dust my hands and clamber up the wall.
    • I hide behind the bush, holding my holy symbol and whispering a short prayer.
  • +4 Stunts
    • I dodge around the pillar, kick a small barrel out behind me to distract the Orc then jink the other way and stab him in his neck.
    • I pull a short ladder out from behind the tool-shed and use it to get to the handholds halfway up the wall.
    • I hide behind the bush, pulling some of the dead leaves on the floor over my body as camouflage.

+6 Stunts are tough. They will depend massively on your group and what they enjoy. This is an example that works from a post about how stunts can work and how they can go wrong.

The Solar characters were being attacked by a T-Rex made of blood-ice attached to its skeleton, charging toward our characters as we battle saber-tooth lions also made of ice and snow. Matt’s character attempted to kill the Rex by flying into its mouth, in the process casting the spell Obsidian Butterflies, which would throw a ton of razor-sharp obsidian shards down through the back of its neck, then using the spell’s momentum to blast himself back out of its snapping jaws.


Shared storytelling (as all RPGs are at heart) requires collaboration and a shared sense of limitations. Most games are explicit that players only (though exclusively) control the actions and equipment of their character. That makes it hard to get to what you want without some work. Younger people push at limitations naturally and that's why you see the kind of thing you mention with a younger crowd.

To encourage this in adult gamers, you'll need to establish some parameters so that their shared sense of limitations can expand in the direction you want to encourage.

To my mind, your biggest obstacle is going to be that adults can too readily see the dangers of this kind of collaboration. After all, if they can invent advantages or exert direct influence on the story, then they can take it over completely without even trying to do so. It's a small step from inventing a magic detector to inventing a +5 vorpal sword of life-bane just waiting for some hero to pick it up and rule the world.

Some suggestions for expanding shared limitations:

  • Give a solid-basis for judging player-invented story additions. This can be anything from how often players can invent things to the kinds of things they can invent. This doesn't have to be something objectively measurable as long as it can be something people agree on. For example: "It has to enhance the immediate scene in an entertaining or interesting way without creating long-term commitments for the GM or interfering with other players." This just needs to be something that can be used as a guideline so that players know where they stand.
  • Work to keep these (expanded) shared limitations simple, yet flexible. If you find people arguing about whether an invention is legit or not then you probably need to simplify or clarify.
  • Have a way to give feedback on player invention. Again, simplicity is a goal and can be anything from XP to character perks to "action points". Any boost that is permanent (like XP) you should consider awarding on a group basis to avoid competition getting out of hand or undermining your intent.
  • Consider having a way for players to give feedback on the invention(s) of other players. This is trickier, but if your player dynamics will support it, can be very helpful. Maybe at the end of each session, have the players vote for favorite player contribution (and maybe have this tie into your own feedback/reward dynamic).
  • If your players are hesitant, suggest a trial period with a deadline for re-evaluation. The trial period gives them an opportunity to try something new and the deadline is reassuring because it gives them a date to aim for in case they really hate it (or fear that they'll really hate it). This also gives your players time to build trust in you and each other that the story won't break as a result of this innovation.

And finally, good luck. I hope it works out for you. My best experience with this kind of thing was with the indie game InSpectres. I recommend obtaining a copy and seeing what they do with collaborative play. We had a great time in a short one-off and the player collaboration in it worked very well.


One specific technique is called Yes, but.... (A cousin of the "Yes, and" idea used in improv.) When a player suggests an idea, you try to accept it -- but add an interesting complication.

It lets players have some agency in the world, without undermining the idea of a challenge.

So if the player says they grab the metal detector, you say yes, but...

  • ...the ring isn't actually in the study -- they find something else hidden that'll play into the story you had in mind.

  • ...the metal detector is missing a particular crystal, and they need to hunt around the room for it. (You've just replaced hunting for the ring with hunting for the crystal, so the actual challenge is essentially unchanged -- that might be the easy way out, but works if you're stuck, or if you wanted to allow a particular character's high search skill to shine.)

  • ...the device turns out to be imp powered or otherwise sentient, and tries to run off with the ring for itself.

Whatever works best for the story!

The idea is to play off the input the players give you, using your role as the one who gets the final say to keep things interesting, but not to outright say no.

  • \$\begingroup\$ This is a great answer to the second question I asked, but I'm still searching for an answer to the first implied part (perhaps I need to highlight it in my question better) - how do I encourage players to make suggestions in the first place? :) Perhaps you could elaborate? \$\endgroup\$
    – Cat
    Dec 31, 2012 at 0:24

The first part is simple, the encouraging of player input, improvisation and collaboration amongst adult or older players I find is done easily by simply asking what they'd like to see for one, such as before starting a campaign or designing it asking things like "more magic items or less". But in game can be trickier and one way I found worked was to slowly include things at the start and point them out, like the example above about the metal detector, keep changing it from session to session and then just slowly or suddenly stop giving it to them as a spot check or search check and because they'll be use to it they'll ask for things, maybe vaguely but they'll think Bout what would be good at that time, like with the ring, they might ask after doing this, "is there anything in the room to help me find it?" and you can be inventive.


You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .