GM Constraints

ChainsI have been chewing a bit on the mechanization of GM restrictions. Often they take the form of things that the GM cannot do, but such restrictions are usually designed to curb abuses. While that’s admirable, it often has elements of fighting the last war, which feels wasteful.

But what if you begin from a position of high GM trust? It’s the position I like to take – I am happy to empower any GM who is good enough to know when not to use that power. For that GM, is there a way to impose limitations on their actions which produce fruitful results the same way that some game designs have gotten a lot of energy out of constraining player choice (constraints breeding creativity and all that).

It seems like rich territory, especially because things that might be a bit to “meta” for players might work very well for the GM.[1] If the GM has broad authority but must work towards a particular end (even if it’s as simple as a “this will go well/poorly”) directed by something other than her sensibilities, that could force play in unexpected directions.

The trick, of course, is to make the direction useful. If it’s merely random, then it’s likely to produce random results. The constraint needs to be something that moves play in rewarding directions. This is, on paper, what a GM is often trying to do when “railroading” players, but in that case it is based on the GM’s decision to trust her sensibilities over the organic direction of play.

And that’s the dark specter of this. If the GM is subject to constraints which spill over into play, is it ultimately a different flavor of railroading? I don’t have an easy answer to this, and I don’t know any real way to find out other than testing out ideas.

So, here’s one for the curious, using Fate/Fae. It has two modes: hard and soft, and some variants.

The basic mechanic is simple: Make a deck of your player’s aspects.[2] Put it in front of you and flip up the first one.

That is the only aspect you can compel. When you do so, discard it and draw the next card. Now that is the only aspect you can compel.

Now, there are two caveats to this:

First, this will work better if your table allows players to pay each other for compels, even retroactively. I like the bowl of chips in the middle of the table that players can just grab from when they self-compel (or recognize a tablemate doing the same). Yes, that demands high trust, but all this is predicated on an assumption of high trust – if the GM is the only person at the table who can be trusted with authority, your game has deeper problems than someone hoarding fate points.

Second, this does not demand that the GM push hard to compel that aspect, but I would certainly suggest acting as if it does. I think this gets most interesting if the GM treats this as the primary driver of play. Because it’s the character’s aspects, the game will always tack towards the players, but because the “chain of events” is unpredictable, the route may be totally crazy.

A few options and hacks:

  • The default treats cards as single use (until you go through the deck), but if you shuffle them back in, then you get the possibility of reincorporation, but lose the guaranteed distribution of hooks.
  • This model gets easier to implement if you also use anchors
  • This probably should be kept hidden from players, just so they can be surprised, but it may not be required.
  • If the GM really wants, she can introduce a number of cards into the deck equal to the number of aspects 1 player has. She cannot dictate when and if they come up.
  • Alternately, if the GM wants to use an abbreviated deck to drive toward s a theme, that might be cool.
  • If you need more flexibility, flip up the top 3 cards, and you can use any one of them. Replace the card used when that happens.
  • If you really want to go nuts, do 3–5 cards, but keep them face up where players can see them. Every time you compel one, replace it, and put one fate point on each remaining one. When the time comes for that one to compel, pay out the full amount on it, rather than the usual 1.


  1. GM Moves, from the *world games, are not really constraints, though they may appear to be at first glance. They are closer to best practices.  ↩
  2. You should do this anyway, since it’s super useful for a lot of things.  ↩

2 thoughts on “GM Constraints

  1. warlock69

    The GM’s two required responsibilities are being a referee and world builder. If you can’t trust your GM to not cheat or be a dick, no amount of rules are going to fix that. Rules aren’t going to fix a bad GM. You can only lay out a series of best practices for your game, and if the GM is a good sport, he will follow them. A bad sport will railroad players in any game, not matter what it says in the rulebook.

    I do like your suggestion of using constraints on the GM to encourage more interesting play. In most games, the GM is purely a referee and not much of a player. The players in their role of the PCs are challenged creatively to come up with solutions to problems, working within the limitations set by the fiction and the game mechanics. The GM in the role of the setting is just there to react to the PCs and present new events as the game’s timeline progresses. The GM is challenged creatively as a narrator to describe the setting and characters, but not as a game participant. The GM is merely elaborating on his game prep. Sure, there are major NPC actions, but they are not always the focus of game play. If the GM had to creatively solve problems throughout the game, then it would make the GM a more active participant but also take the game in unexpected (hopefully more interesting) directions.

    Your idea about using cards for compels in Fate gave me an idea for other games. I recently bought the GameMastery Plot Twist cards from Paizo on a whim. Each of the cards feature a plot twist header with a mechanical effect and four suggestions for narrative effects. The deck is an equal mix of good fortune and misfortune. The cards are intended to be tools for players, but I think GM’s could use them in a similar fashion to your Fate suggestion. The GM would draw one card at a time. When he sees an opportunity in the course of the game, he can throw that plot twist into the mix, and then draws another card. There may need to be some kind of pacing rule so that a plot twist can only be thrown every so many scenes. You would want to pace the plot twists to reflect the tone and genre of the game.

    If you don’t have Plot Twist cards, I suppose you could create something similar using a random table.

  2. Pingback: GM Constraints… Wait, WTF? | The Rhetorical Gamer

Leave a Reply

Your email address will not be published. Required fields are marked *