You are viewing davidberg

delve, glyph

Support for GMs looking to Plan Right?

Vincent Baker recently shared an annecdote about how designers from different corners of RPG-land publicly advocated against GMs planning outcomes in their games.

Commenter CC (I don't know his name) replied with an excellent example of a planned outcome that paid off, with the PCs deceived a second time by their hated arch-nemesis.

Here's my take on the issue:

  • Some GMs want to plan and guarantee outcomes because they're control freaks or idiots or dicks who don't care what their players want. They pursue such guarantees whenever they feel like it.



  • Other GMs want to plan and guarantee outcomes because they know their groups will love 'em. They pursue such guarantees only when an inspiration meets that criteria.



  • Still other (most?) GMs are in-between. A little team spirit, a little selfishness.


GMs from all three camps have historically wound up using costly approaches to guarantee their outcomes. Disempowering players is the most common cost, but there are others, like rendering the fiction implausible.

  • Some GMs who've seen, heard of, or experienced these costs have sworn off guarantees altogether, preferring to play games where outcomes aren't planned.



  • Other GMs have kept the guarantees but worked on minimizing the costs. It hasn't been easy. I'm not aware of any published designs that achieve this.


So, when accomplished designers tell an audience of 250 GMs, "Don't plan," they have a point. It's an admonition to the control freaks and dicks and idiots, an alert to the in-betweeners, and a reality check for anyone who thinks minimizing the costs of planning will be quick and easy.

But it's not the end of the conversation. The right outcomes, carefully conceived and brought to fruition, can add great things to play. The costs can in fact be minimized to some degree. How much? Enough? Are there solutions other than years of GM training? System-based solutions, perhaps?

I'm glad the "don't plan" call and efforts are out there. Now let's have some equal passion and effort for "plan right".

Update, Dec. 5 -- I've begun working on ways to support this in the Forge forums. One thread about the general topic, another about resolution, more to come.

Comments

Not sure those ideas are so smoothly opposed

To explain why I'll probably have to splurge a lot of stuff I've been thinking about to do with game goals:

Ok, I can agree first off that some people try to go for specific outcomes, because like I was saying before about the blindfold example, there is a value in trying to get a certain outcome; a certain arc, a certain end point or moment, even above the specific thing you want to show people.

Now this happens in loads of areas, people do it larps, in jeep, in storygames, wherever. You all aim at a target with varying levels of system assistance and constraint. Having achieved it plausibly is a victory for everyone, you can judge your success by how plausibly you got there, and have narrative catharsis that everything ended up "right".

I think this value of wanting to "do the plan", to complete the arc or depict what you wanted to depict is the kind of satisfaction an artist gets at a finished picture. You've known for ages what you were going to produce, but seeing that thing there made by your hand is gratifying. (There's also the thing about how reproducing something helps you see it in a new light, the way that artists or writers or programmers have new eyes from then on to look at everything)

The strange and unusual thing for me is when players then start to want to have certain outcomes without aiming for them with the control they have; heroic adventures where they play reluctant heroes, etc. This seems like on the one hand a perfectionist handicap on the mission of doing it right; a requirement for total authenticity and motivational logic, combined with a difficult starting point. On the other it's the superposition of another type of goal; "being there".

If someone really wants to get into the shoes of a particular character he has considered, and at the same time wants that character to achieve specific things before being killed by his brother, they are not only leaving out the legwork of justification of the latter events, they are introducing new elements to be fitted to it that might clash, rendering it almost impossible to achieve their goal (or just very very hard for the person they've asked to do it).

So it seems like there's this long term goal you want to be heading towards, and there is this mode of being, and both of those are reasons for play.

"Being there" can have it's own planning requirements, eg. the "tourist route" approach, where the GM pulls players around the setting or various other situations so that they can explore the situations they want to see. (Honestly I think this would often be best served by scene cuts or other relaxed continuity methods, but that's another story.)

Then there is the other side of things, another longer term goal that also requires constraint; getting to explore something interesting and unknown. For example, answering questions. Of course, no game can actually answer anything about the real world, but players can pose the question in the game world and then consider how their answer relates to the real world. The setup then becomes about embodying the fictional world with enough purchase in the real world, enough human life, that the parallels become interesting and thought provoking. (This description seems to fit pretty closely to "story now" concepts)

Or maybe they're not worried about the real world, and rather than the asking questions, they just want to explore new stuff. In that situation the long term goal is not that something will be reproduced, but that a new area will be poked at and you'll come up with "something" to fill it.

In that situation you cannot force the outcome ahead of time at all. There's no point, because you don't know what it is you're going for until you get there, by definition! But you can plan to exclude other possibilities and narrow the focus to see what you can come up with. You can also build up the background so that you have the fuel to make the new stuff complex and evocative, with alien strangeness or wittiness or absurdity or mathematical subtlety or whatever.

To me in both cases something like planning appears: In one case reproducing the structure itself is the key, in the other, the structure that is there serves as a stage to produce other stuff, new stuff.

Re: Not sure those ideas are so smoothly opposed

Man, almost maxed out the comment box with my intro, but here's mainly what I was getting to:

If that model of why people are playing is true, I feel like there are some people who really are not “playing to find out what happens”, not just by habit, but because building story arcs or special moments interests them.

They should probably be told (if anyone knows how to do it well enough to tell them) “here are ways to manipulate your player’s characters better” and “here are ways to insure you and them are on the same page about where the game should be going”.

Or “why not get the other players involved in planning too?”.

This feels like at least two kinds of "plan right". Do you abandon immersion and go for combined story telling, or do you make it so that the players who work without immersion cover for those who do?

For other players, who are playing for the unknown, they should be told ways to focus the scope on what they want to explore, prepare stuff to work from without stomping over their own interesting conundrums, and given various tools to do that nicely. For them explicit arc planning/choice funnelling probably is a stupid idea, and they'll want to find alternatives.

Loads of indie games take that "scope narrowing" out of the hands of the GM and put it in the rules, trying to build a starting situation that is self contained, a setting that flags up new and interesting perspectives and precludes cliched responces, or building the basics of an arc into the natural behaviour of play. Currency loops can be a form of automated planning, as can certain "reward systems".

In these cases the broader arc is not the main thing, it's cool and all, but better to just automate it and focus creative energies on the new stuff.

Re: Not sure those ideas are so smoothly opposed

Hi Josh,

Sure, there can definitely be different types of planning (good, bad, none) by different players in one game. I didn't even intend to address players' plans for their characters here -- that's a whole new interesting topic!

I think you make some good points about the various reasons someone (a GM, for this example) might want to plan. Authorship, emulation, immersion -- for pretty much any goal someone might bring to an RPG table, there's some appeal in having control over what you get along that vector.

I suspect it's accurate to say that, though most players of games want some types of outcomes left to chance (e.g., whether that character pulls off that tricky badass feat), most players of RPGs also come in with other outcomes they don't want left to chance (e.g., I'm immersed, this game is recognizable as Star Wars, etc.).

That said, I think we do in fact have game designs that offer players and GMs guarantees of the types of outcome they'll get (immersive, Star Wars-y, etc.). We just don't have designs for elegantly guaranteeing specific outcomes (like "arch-nemesis stabs PCs in the back").

Re: Not sure those ideas are so smoothly opposed

Here's the trouble I see here, that we'll probably have to get round; most of the games we make hard-code those types of outcome, you know?

And I think that's ok going in because the creative activity isn't there. You know as GM going into a game that the game will gaurantee x, and you're off doing something different.

Like D&D 4e at it's most combatty handles monster creation and level appropriateness, and leaves spatial stuff and tactics to you.

And I wonder whether any system that encourages planning has to do the same kind of step back:

It would set up the potential to fix certain kinds of outcomes, but leave the last steps of wrestling to the GM or player, so that they actually have something to do.

Like if you look at universalis, that provides a system to mandate certain things will happen, and to set up the game world with clues to your idea, and then fight off people prematurely triggering your arc etc.

I reckon with the right mindset going in, and careful use of gimmics so that you don't mess up the structure, you can run highly plotted stories in universalis. And that has that kind of "wrestling with unexpected hitches to produce a story", but whoever wins, you still get one!

But if you make a system that always leads to specific outcomes, doesn't that mean that the game will always leads to backstabbings etc? That people will only choose to play that game if they want to get backstabbed halfway through the game?

There's a whole knot of things here, and I'm trying to avoid the problem of giving a cook an automatic food-producing machine rather than a better kitchen, so that those people who don't just want food, but want to make food, still get to do it.

But of course, microwave ready meals are the best thing ever when you're starving hungry, and if you can find a way to auto-produce stories that are really good then some people will be totally happy with that.

There probably is a case for giving people a foolproof plan to implement, and they just pick which plan they want for the situation, like a big box of plot twists and story-trope machines to plug into their game.

Is that more what you were going for?

Re: Not sure those ideas are so smoothly opposed

I don't see much appeal to a game that guarantees you will get backstabbed.

I want a game that elegantly* guarantees whatever outcomes the GM wants to guarantee.

..within limits, of course. The GM should be helped or forced to choose wisely in this regard.

*"elegantly" is the hard part! we already have the rest

Re: Not sure those ideas are so smoothly opposed

I don't know, could be quite funny. Will it be him? Or him? Or her? Actually, it was the dog, he was mind controlled.

Now I've mentioned them I really like the idea of plan fragments; imagine that instead of having an "adventure path" or module that people got sold, they instead were given an interesting encounter style and all the scaffolding it needs so you can run the game towards it. In other words you'd have a checklist of things to leave outside of player power in each of the scenes leading up to the particular one, what sort of minimum state the players should be in by the point of the encounter and suggestions of how to fit that in to different scenes or deal with minor failures/deviations. They'd also have forshadowing examples, so you could add litle bits of the future event to the current situation.

Basically the plan fragments being small enough would allow you to build your own plan from them, and the more you used, the more tightly constrained previous scenes would be!

On the broader stuff, I'd like to see people analyse what the tradeoffs are in that kind of "planner-GM and blindfolded PCs" play; restricting player character power in these areas allows you to make these kinds of outcome assured, but means you cannot build challenges involving x, vs "players playing their characters along certain lines can lead to this, but also that". Etc.

There's some facinating room to explore, and see where a GM actually has "room to breathe" to react to players and what she can do with that, assuming only that prep that is shared is set in stone.

But I think trying to make systems for the GM to produce any arbitrary outcome is insanely ambitious. I think we're more likely to get progress either by slowly colonising the possibility space of robust plans, or by getting games like universalis and finding better stratergies for them. Or in other words, you get games that can do 3 things totally reliably, or games where you are slowly more reliable at doing anything!

Re: Not sure those ideas are so smoothly opposed

Insanely ambitious? Perhaps. But is it any more ambitious than the typical indie approach, crafting systems to always produce some outcome that is really quality in all the ways that count?

Also, by saying "any arbitrary outcome" you're making it even harder. Tackling a finite scope of outcomes is fine (see "within limits" above).

Separately, the goals you mention for the Plan Fragments sound great, but I have no idea how what you're describing would achieve them.

Re: Not sure those ideas are so smoothly opposed

I think it is more ambitious, vastly more ambitious:

Because with indie games there is scope for playtesting; you know off the bat that in every one of 60 games you'll want event x to occur, or you'll want players to be able to explore theme y, or you'll want this pattern of causality or that relationship between players.

That's only part of the game, but you are saying as a designer "I want the GM/players to be able to do x".

Whereas there are people all over the place who really do expect the GM to be given a set of random plot pieces and experience parameters and make them work, first time!

GMs will say things like "ask me what you want to do and I'll try to find a way to make it happen".

Helping GMs in that situation is a higher order problem; they are effectively having to design a game on the fly!

Giving them an arbitrary system to start them off may not help, if the system doesn't match the kinds of problems they are likely to be facing etc etc.

Loads of GMs of that style would be better off swapping systems throughout the game; sometimes you roll social damage, sometimes you roll saves for your relationships, sometimes you roll for whether you succeed at climbing a clocktower, sometimes you have to match cards to evade the people pursuing you.

When I start to contemplate GMing a game in that way the possibilities are mind blowing!

Of course GMs can always say no.
They can specify what they think they're going to be able to handle; limit the ways that players interact with the world, the powers of their characters, the resolution mechanics and game world verbs available, set out game world parameters etc.

They can say that certain character types will not fit, or that certain goals are out of type for this kind of game etc.

But that in itself is no negligible task, how do you distinguish in advance how player goals are going to interact? What you'll be able to handle?

Most people develop a feel for "I can see that working", or "I've no idea how that could work". Or you can see how someone else's game works and try it out.

My feeling is that we're

reasonably good at sharing resolution systems with each other,

ok at sharing ways to slowly shift gameplay recursively (reward cycles, bangs, fronts, propagating changes through relationships, all that jazz),

but not so good at sharing "I wanted them to get to the temple before nightfall so I could describe the light of the sunset on it's mirrored surfaces, so I realised I needed to reduce the difficulty of this travel challenge".

We're not so good at communicating or systemising how to keep track of dependencies through time, and setting up current events to support later ones. Or what consequences to be aware of when doing this.

This is of course because of the emphasis on different kinds of play, and on player equality, reducing hidden information and all that.

Plan fragments are basically that, what you need to keep track of today to make what you've planned for tomorrow run smoothly.

My theory is that although a lot of experienced plan-first GMs do this, they internalise the dependencies, so that they only appear when someone tries to kill an npc they are going to need later, or they think things are taking too long etc.

My idea is to help GMs to put it on the paper in front of them, so that they can clear their heads a bit more, and also share the information with people in other games.

Re: Not sure those ideas are so smoothly opposed

Couldn't get this comment to post for some reason, been trying on and off for months in case it was an intermittent bug. In the end realised I should have just got a livejournal account! Probably a little out of date now given the forge stuff, but I've waited long enough so I'll let it stand:

Apologies about my style in these comments, trying to come up with an alternative to categorical statements like "don't plan" got me into too much of a dogmatic/systemising mode. Sometimes rewarding, but a little too much hard work!

Coming back down to earth, creating plan fragments is my idea for how to sort out the disconnect between the in-play and pre-play parts of planning:

Players tend to recognise planned play when it interferes with their freedom, but that can be a red herring, as GMs can be obstructive for lots of reasons.

GM's often see it totally differently, with planned play being about considering the future of where you want things to head, and the various lovely vistas/complex encounters you have in your head. That can be a total red herring, because you need to actually find ways to get the players there.

In the middle you've got something a management person said:

Planning is about what you have to do now to make the future you want happen.

To actually work off this insight, you can just (for example) write down each of the scenes you want to happen on A4 sheets, landscape, with a margin a fifth of the way in, and list on the left hand side of the sheet the things that need to have been set up for the things on the right hand to happen properly. Then when you lay all of these sheets down, you can sort of fan them out like playing cards with the current sheet for this scene/encounter at the front on the right, and all the constraints of future scenes beneath it and on the left.

With some experience to categorise those constraints, and graphic design tricks to make it clear, then you could always have in front of you information about what you need to do to make this scene work in the context of the larger game.

You could know when the player choices in the right hand side are liable to hit into the constraints on the left, and act to avoid this, and you could also look for potential points of conflict between the constraints for different future scenes.

The next kind of thing to have would be a goto set of advice for dealing with negotiating conflicts between constraints, and dealing with diversions, and you could write on the back of those scenes cards what sort of alternative scenes could replace them if they become invalidated before you can do them.

I'm thinking that as you start working with these sheets, you'd start shifting from "this specific npc must live" to "there must be a character that has this relationship to the player characters" and other forms of generalisation, with the old npc becoming separated out into a different pile, of NPCs particularly suited for various roles. Slowly taking the plan part and making it more abstract, while honing in on the actual details you want.

For preparation methods that don't follow the same kind of sequence different stuff may be required, but that can still be approached in a similar way; what must be achieved in order to set up this future set of events? What events or story elements move smoothly with this? Or even tend to help set it up? Etc.

Once you can keep track of the dependencies, I suspect planning becomes a lot easier, and hopefully when you have this kind of format, more experienced people than me can turn it into something foolproof, taking plan fragments and building tricks around them to make them work, or adjusting them to improve them etc.

Re: Not sure those ideas are so smoothly opposed

Hey Josh,

Just wanted to let you know this comment finally got through! Cool stuff. I'm busy now, but will respond soon.

Re: Not sure those ideas are so smoothly opposed

So, my thought on your idea is that it absolutely would work if it could be implemented, but I'm not sure it could be implemented.

I've found it both difficult and boring to map out what needs to happen in order for the stuff I want to occur. I have friends who have taken this exact approach and most of the time it took them forever and didn't work. They avoided awkward railroading at the end... by doing awkward railroading at the beginning.

I think that, for the system you suggest to be fun and easy for GMs to use, the game itself would have to identify the dependencies. Like, I say, "dusk arrival" and then I consult a card or chart and quickly get some list of likely delays to avoid.

I do think that the planning a lot of GMs already do could be aided by a well-designed visual format. Honestly, I'd love to see something like that, and give it a spin. But ultimately I think it's the wrong direction. No reasonable amount of prep will ever obviate the need to occasionally assert "I need this to happen". And a game that needs to accommodate that anyway probably shouldn't focus on ways to avoid it.
I've followed up on this at the Forge. One thread about the general topic, another about resolution, and more to come.