Friday 3 February 2012

On Improvisation and Getting Better

Let me begin by making a statement which I don't think will be easily contradicted: Improvisation of rules is perhaps the most important skill for a GM to develop.

Let me make a further statement: Improvisation of rules is not very thoroughly theorised, taught, or even talked about, whenever anybody grapples with the issue of "how to be a good GM" or, more fundamentally, "how to make games fun".

Let me make a concluding statement: There is a good reason for this.

In the intro to the 2nd edition AD&D Dungeon Master's Guide, we find Zeb Cook giving probably the best GMing advice that I can think of:

The Player's Handbook and the Dungeon Master Guide give you what you're expected to know, but that doesn't mean the game begins and ends there. Your game will go in directions not yet explored and your players will try things others think strange. Sometimes these strange things will work; sometimes they won't. Just accept this, be ready for it, and enjoy it. 
Take the time to have fun with the AD&D rules. Add, create, expand, and extrapolate. Don't just let the game sit there, and don't become a rules lawyer worrying about each piddly little detail. If you can't figure out the answer, MAKE IT UP! And whatever you do, don't fall into the trap of believing these rules are complete. They are not. You cannot sit back and let the rule book do everything for you. Take the time and effort to become not just a good DM, but a brilliant one.

To repeat: Take the time and effort to become not just a good DM, but a brilliant one.

Coming as it does at the end of these two paragraphs, the implication of this sentence is clear: Zeb believes that becoming "a brilliant DM" is directly connected to having fun with the rules; adding, creating, expanding, and extrapolating; and making up the answers to things. I agree with this - it does. But there is another implicit question which Zeb doesn't answer: How is possible to get better at all of this?

That's not a question which is easily answered, because I tend to think that GMing is basically a form of tacit knowledge, and being able to improvise rulings is perhaps the area where this is truest. Improvising rules in ways that are fair, interesting, and workable is not a skill that can really be learned from a book. It's something that has to come through experience. To borrow an analogy from Michael Oakeshott, being a good GM is rather like being a good chef. There is a set of rules, which are like a recipe. But creating a really tasty dish is much more about the feeling and intuition gained from years of cooking than it is about slavishly following the recipe. Strict adherence to the recipe will get you so far, but proper chefs have the experience and know-how to take the dish to a level far higher, and often if you ask them why they are doing what they are doing, and what rules they are following, they'll find it difficult to tell you.

GMing is similar. A good GM probably isn't particularly aware why they are doing the things they are doing (beyond the obvious awareness that they're doing it "because it works"), and if asked to write down a "good GMing guide" they would probably not be able to produce anything particularly coherent or enlightening. This is especially true when it comes to improvisation. I'm sure you all know GMs and players who are good at this - but try asking these people to explain how they got good at it, or what principles they apply. Probably, they won't be able to do that. Probably, they're just good because they have a lot of experience in GMing, giving them an intuition for what will work, and the tacit knowledge of how to pull it off.

In a sense, this reinforces Zeb Cook's advice, and perhaps shows it in an even better light: Take the time and effort to become not just a good DM, but a brilliant one; or, in other words, develop your tacit knowledge and intuition for improvisation through continual play.

(Edit, because I think it was unclear in the original post: Let me emphasise that I am talking about improvising rules for situations not covered in the game text, rather than improvising content - like NPC personalities or settings. I think the latter is more easily taught than the former. The paradigm example of what I am talking about would be what crops up in this post: There is an elf who might try to sneak up on the PCs' campsite during the night. One of the players decides his character will put coins on the tops of rocks, as a primitive alarm system which will make a noise if the elf knocks them over. Rules improvisation on the part of the DM here is crucial - he needs to decide how he will adjudicate the success of this. And to do so successfully he needs tacit knowledge.)


  1. Lies.

    I'm going to write something coherent and enlightening.

  2. I'm not certain I understood the thrust of your statement completely, but I have a fairly decent collection of works that deal with improvisation in general (e.g.
    Impro) and improvisation that is gaming specific (e.g.
    Play Unsafe).

    This is just two illustrative (an excellent) texts in a larger body of work.

    Now I feel there is a difference between improvising content / narrative and improvising rules. Taking the structures and components of the game and re-arranging them into interesting and aesthetically compelling ways requires either long familiarity with a game or wide experience with many games. And you're right, there is less written about that aspect of improvisation.

  3. -C: I'm interested to see what you come up with.

    Adam: I haven't read either, but "Impro" seems to me to be a collection of methods for practising improvisation, which is really just a way of developing tacit knowledge in the manner I'm discussing. "Play Unsafe" I'm not sure about - I'm unwilling to pay the £12.81 it costs to find out!

    For what it's worth yes, I did mean improvising rules. I should have made that much clearer. And in fact I might go back and make some edits.

  4. does this imply that some games make it harder than others to become a "brilliant gm", or even impossible?

  5. I think this tacit knowledge is identical to normal knowledge it's just that there are many, many variables in the world and an expert has learned the basics such that they've moved on to more subtle ones, maybe even working with those variables subconsciously.

    If we closely observe enough sessions of DMs having to improvise rulings there will most certainly be patterns. Those patterns can lead to guidelines or rules-of-thumb.

    I think there are already a few of these floating around: "Yes, but . . ." or the idea to allow yourself to revisit rulings in a later session (meaning that we accept the difficulty of predicting all the implications of a ruling).

  6. Yeah it's a good question that, maybe it's best solved not by books but by conversations? By someone stating a situation, and another person saying how they'd solve it.

    The weird thing about that is that it's almost roleplaying within roleplaying, with the GM who wants to learn setting up a situation for other GMs to resolve!

    The less recursive version of that is people playing in other GM's games, doing stuff based on stuff their players did, and seeing how the GM responds.

  7. Interestingly, it's very similar to Gygax's introduction to the Type One DMG, which I was reading last night -- and it ties into Zak's latest post in that the negotiation of unknown rule-space is where a lot of people have problems (especially with OD&D) and the ability to handle it is one metric for some sort of DM skill.

    I thought that the soccer example Zak used was illuminating because he (possibly unintentionally) described creating a pick-up game with 11 people -- having been in that situation, having somebody who can quickly and equitably decide teams and rules for the game (we use this as goals, no shots from beyond this distance, no offside, etc.) is very valuable in terms of getting on with playing.

  8. Shlominus: Yes - I think the fewer rules there are, the quicker you will learn how to improvise rulings. So maybe the best GM tools are actually just sitting down and playing Risus a lot, or something?

    Telecanter and Josh W: Observing other GMs is something we almost never do, isn't it? I mean, even within one gaming group it's rare to have more than one or two other GMs. I guess conventions are a way of doing this, but I never go to conventions.

    Thomas: Yeah, I think it's not just one metric for DM skill, but just about the most important one. The only other important DM skills I can really think of are:

    a) Creativity in setting/NPC/monster/"mission" design.
    b) Being a good people person (listening properly to what people say, facilitating discussion, etc.).

  9. Regarding your comment about observing other GMs, it should be possible to record some of these awesome recent spate of G+ sessions via video-capture devices. If somebody were to capture and archive even a handful of sessions, that would be really interesting.

    And I think you may have just described the Three Pillars of DMing, in the sense that you really have to have all three of them to get the full use of the others (a creative DM who can fly by the seat of his pants won't run a good game if he's a jerk, a great guy who adjudicates quickly and fairly won't run a good game if it's just orcs in rooms, etc.).

  10. Thomas: Don't you think that all of those things revolve around tacit knowledge? It seems to me they do. I wasn't thinking about that as I wrote the comment, but I am now.

  11. I think that's a fair conclusion, yeah. I think that's why Gygax refers to DMing as "an art" in the Type One DMG.

  12. Hm, I think it's more about mindset than experience/training. I GM'd perfectly fine at age 11/12, running stuff like the Fighting Fantasy RPG. I DM'd poorly in my '20s, a bit better in the 3e era, and have been getting a lot better again I think in the last year or so. Part of that was playing more, as well as DMing. But the main thing is a mind set: "I am the judge, I will adjudicate, or roll a die. I will look in a rulebook if I deem it necessary, but I am in charge, not the book".

    That attitude is linked to the content of the books - minimal rules encourage this attitude - and to the GMing advice and other keys in the books, eg the incomprehensibility & inconsistency of 1e encourages adjudication, as does Zeb Cook's 2e DMing advice, whereas 3e's presentation encourages look-in-the-book, slave-to-the-rules play, and 4e is somewhere in-between.

  13. I think there is one piece of advice that apply here.

    Just like a potential author should read a lot (especially a genre author who *must* read outside the genre!) to see how it is done, I think the brilliant GM should read a lot.

    If you want to be able to invent good rules and rulings, nothing beats having a great reservoir of examples to pull from.

    In many parts of the rpg hobby, people start with D&D and stay with D&D until they "die". Even with quirky resources like the 1st ed. AD&D DMG, this is limiting.

    I'd suggest to everyone who want to be a brilliant GM to read all the rules they can find. Preferably both some hoary stuff from the 70-ies and hip and arty Forge games from the early 2000-ies. Read other games than they style you like.

    That will teach you a lot!

  14. S'mon: Isn't mindset part of tacit knowledge? You seem to be saying you've got better with experience... ;)

    Andreas Davour: Yes, that's excellent advice. I suppose "GM more in different systems" is a way of developing the kind of expertise to make really good improvised rulings.

  15. It's true we don't often get to watch people DM. We're pretty much limited to self-reflection and, now, blogs/forums.

    And wouldn't that be odd for a chef, to try to learn that field by trial and error. Hell, they have schools and television shows.

  16. noisms:
    "S'mon: Isn't mindset part of tacit knowledge? You seem to be saying you've got better with experience... ;)"

    Yes, I think I have got better with experience over the past 10 yearsm and especially over the past year.

    But: I was much worse at age 22, 1995, than I was at age 12, 1985, despite lots of DMing experience in the meantime. Part of that is because after I left school age 18 I only DM'd irregularly. A lot more of it though is because it was the '90s, and the games sucked, the published adventures sucked, and the GMing advice sucked.

  17. I think you make a very good point here.

    I think there's also some interesting discussion to be had in how rule systems can be designed to support GMs in this.

    To take three examples using your "coins on a rock" situation:

    (1) The system includes no rules for adjudicating surprise or detection. The GM has to provide an entire mechanic.

    (2) The system includes a basic 1 in 6 chance of surprise. The GM needs to figure out how to modify this system to account for both (a) a character sleeping and (b) the primitive alarm system.

    (3) The system provides a Perception skill and a Stealth skill and a unified structure of modifiers being applied to those roles. The GM's task is now simplified to simply determining the right modifiers.

    It's why I've said in the past that a properly structured rule system facilitates rulings.

    OTOH, of course, improperly designed rules can also impede and hinder rulings.