Rabalias grew up wanting to be a pirate. But a band of evil bureaucrats kidnapped him and forced him to work for The Man. Even so, Rabalias was patient and cunning. He escaped by gnawing his way through the walls of his prison and concealing the hole behind a picture of cthulhu. He fled to the coast, and stowed away on the Black Armada, where he worked his way up to the rank of Admiral.
Over at Department V, Smiorgan writes about Everway’s three methods to decide a conflict: Karma, drama and llama.
Smiorgan discusses the issue of who decides what the plot should be (in order to rule in accordance with drama), and how randomness (fortune) can introduce something new and unpredictable.
I mostly want to talk about the latter here. It’s a ubiquitous way to keep the game unpredictable: Pick up some dice and let fate decide what happens. An observation I make is that this is very often restricted to determining “can I do X”, which is in itself only one of the interesting things one needs to decide during a roleplaying game, but that’s a topic for another day. What I want to talk about here is an alternative approach to introducing unpredictability – one which I am increasingly favouring in my game design.
Here I am going to refer to the method as asking. More broadly, it is about giving away decision-making power to someone else. You see, your decisions as GM (or a player, for that matter) may be based on drama or karma or something else entirely, but to you they can seem predictable. You have perhaps already thought about what the needs of the story are, or what the demands of the fictional situation are, so making that decision can seem predictable to you.
So an obvious way to get the sense of unpredictability for as many people as possible is to spread those decisions around. I’m not talking about discussion and consensus; in many ways that feels like the most predictable method of all for resolving things. I’m talking about varying who makes the decisions.
In a traditional GM-and-players game, you get this a bit. The GM takes decisions about the NPCs, the world, and often some conflict resolution. so they provide a sense of unpredictability to the other players. And it’s often remarked by GMs how the players’ surprising actions make the game exciting and unpredictable. But it’s clear that the GM has much broader scope for making decisions, and it is they who provide the chief source of unpredictability outside of the dice: the question is not “what will happen”, but “what will the GM decide”.
What I’m increasingly finding is that having all the players involved in those GM decisions, by making individual calls, creates a fantastic sense of unpredictability for everyone. No one person has their hand on the tiller, so the boat goes where it will.
I’ve called it asking, because a very straightforward way to make it happen is by asking questions to another player. Instead of it being either the dice or the GM who decides what happens next, it’s another person whose mind you can’t read. And even the person you ask, moments ago, didn’t know what the question would be or that they would be answering it.
Similarly though, rotating roles (as seen in Microscope and Lovecraftesque, for example) ensures that the story isn’t moving in a straight line. Each person guides it a bit, and no one person could have forseen where it would go. In effect, here, it’s the system doing the asking, but instead of always asking the same person, it’s a different person every time.
It’s important to emphasise this is about one person deciding. If you turn to group discussion for this, you quickly find that you’re relying on negotiation, social dynamics and (often) a rather turgid laying out of the reasons for and against each course of action. This is far from unpredictable.
This is also the method that lies behind improv-based approaches to GMless roleplaying. Each person leaps forward and inserts their ideas into the story higgledy piggledy, like having a jam session. But what improv approaches tend to leave space for a small number of people (maybe just one) to dominate the game, subtly or not-so-subtly steering things so that they are not so much a product of the group as the product of an organising committee. This is why games like Lovecraftesque and Microscope impose a no-discussion rule, forcing every player to contribute to the flow of the game.
So there you go – karma, drama, llama and banana, I guess.
With apologies to regular readers, this is an advert for my local gaming club.
Refugees from Reality
We are a Roleplaying and Board Gaming club in Chesterfield, Derbyshire. We enjoy a wide variety of games, from indie to the big names. We welcome new players and are always open to trying something different. Find out more on our website or drop us a line if you’re interested in coming along. www.refugeesfromreality.co.uk
To celebrate the fact that Lovecraftesque is now a physical product you can hold in your hands, and since Halloween is coming up, we’re running a competition to design a scenario for the game. You don’t need to have read the game, you just need some creepy ideas and a little time.
Lovecraftesque scenarios are short – like, a couple of pages – and their role is to provide inspirational material linked by a common location and theme. They provide ingredients someone could use to make an eldritch horror tale, but the players choose which ingredients to use, they add ideas of their own, and build a story for themselves. You can find a bunch of examples from the kickstarter here.
You have until 11:59pm (UTC-11) on Monday the 7th November to submit your scenario(s) by commenting here with a link or emailling us at lovecraftesque at vapourspace dot net. Each person can submit up to three entries for consideration. The winner, chosen from among the entries by us, will receive a copy of Lovecraftesque softcover and cards for themselves or a friend (which we will ship to you wherever you are), and the runners-up will receive the PDF of the game (again, you can give to a friend instead).
We recommend that you follow these guidelines; we’ll accept variations but the scenario must be an inspirational set of ingredients, not a pre-cooked story or situation.
You retain your copyright over your entry/entries but unless you tell us not to we’ll publish your work on the Black Armada website for free download. We won’t use it any other way without your permission. We reserve the right not to publish anything which we deem distasteful or in breach of someone else’s intellectual property.
Georges Cuvier, the legendary father of palaeontology, boasted that he could deduce the class and even genus of an animal based on a single bone, because the “correlation of parts” ensures that every component of the animal’s body are related. In other words, every animal is specialised for a particular way of functioning, and therefore every part of the animal reflects that way of functioning.
You can say the same of roleplaying games. There are plenty of games which boast of being completely generic, or of being able to handle just about any situation. But upon close inspection of even individual parts of a game, it is possible to discern a great deal about that game’s way of functioning. Just as an animal has a specialised “design”, games have a literal specialised design that can be observed in individual components of the game.
Before I go on, let’s pause to consider the difference between the design of a game – its procedures, guidance and fictional material – and the way it’s actually played. A game is, after all, not really an organism but a tool. We can learn a great deal from studying a tool, but its full functioning only emerges when we see how it is used. Even so, if a tool is well-designed, its designer will have envisaged a particular way of using it which the design will then promote and support. You can use a kitchen knife to cut paper, and you may even get quite good at using it that way, but it is designed to cut food and that is where it comes into its own.
Let’s take as a simple case study the so-called “traditional” game. This is a class of beasts rather than a single animal, but its members have components in common from which we can deduce a common function. I’m going to look at three such components: the GM; difficulty checks; and ratings for weapons and armour.
First and most fundamental, the GM. Games vary wildly in how they implement this component, but the “traditional” approach is for one person to exercise their judgement and creativity to plan a fictional setting and events, which form the context for a situation that same person creates, populated by people,
creatures and phenomena that person describes and controls, regulated by rules that person adjudicates up to and including ignoring the rules in favour of their own rules or ad-hoc decisions. Everyone else describes and controls one character within that setting and situation.
Given that description of a GM, a game design Cuvier would conclude that this was an incredibly important role. The backbone of the game in which it featured. It is possible for a GM to take a highly collaborative, discursive approach, but the role’s natural oeuvre is autocratic – there’s nothing in that description I just gave about collaboration except at the interface between the player characters’ actions and the rest of the world. And because they control so much, they dictate the terms by which the other players must engage with the game.
Returning briefly to what I said earlier about the way a game is actually played, it’s really important to acknowledge that a lot of groups, whether their particular game tells them to or not, do in fact adopt this more collaborative approach. GMs may look to their players for subtle cues to help them craft an experience that will be satisfying for their players, or they may be much more explicit, discussing what the players want from the game and giving them creative input on setting and even situation. But that is not what the game component known as the GM is designed for.
No, the GM as a design component is clearly aimed at producing a specific experience: highly guided play, where one person decides broadly what the game is going to be about and then prepares and moderates the game accordingly. As an experience it’s close to a choose-your-own-adventure story, but with the vastly expanded flexibility for action implied by decisions being run through a human brain instead of a branching flowchart. I’m not going to go into a discussion of other approaches but it should be obvious that there are many other possible experiences a game could promote.
Moving on: difficulty checks. These
are a ubiquitous mechanic across a range of games, and in most games that have difficulty check mechanic, that is the single core mechanic for handling conflict (outside of combat which sometimes has its own dedicated mechanic). Characters have some numerical stat (attribute, skill, talent, whatever) and via some kind of randomisation they either succeed at a task (if they beat a difficulty number) or fail (if they don’t). Maybe there will be degrees of failure, yes-and, no-but, etc, but fundamentally the mechanic is about: can this character overcome this obstacle, avoid this risk, complete this task.
The fact that so many games make this sort of mechanic the core of their ruleset tells you one simple thing: they want you to tell stories that are about trying to do stuff and succeeding or failing. In other words, what Robin Laws would call “procedural beats”. There are lots of great stories you can tell that revolve around success and failure: action movies and police procedurals, for instance[*]. But, once again, there are many other possible experiences a game could provide.
Let’s also pause to note the link between these first two components: the GM typically decides when a difficulty check is needed, what the difficulty level should be and what the consequences of success of failure are. So in games with a GM, difficulty checks are a key tool that can and often are used to help the GM control pacing and dictate the terms of the story; they help to make the game even more of a guided experience. Once again, we observe Cuvier’s principle of the correlation of parts.
Finally, ratings for weapons and armour. A lot of games, and virtually all games that have both a GM and difficulty checks, include rules for inflicting damage on characters. This very often means a set of numbers telling you how much damage weapons do, and how good armour is at blocking that damage. (Sometimes the “weapon” is a spell, or the “armour” might be a mutation or something, but it’s the same principle.) In most such games, no other sphere of activity is delved into in that level of detail: we don’t (usually) have ratings for various investigative tools, or for how different types of terrain interact with stealth, or anything like that.
What this tells us should be obvious: the game is about fighting. We need all that detail about weapons and armour because we’re going to be doing a lot of fighting and we care about giving it a level of granularity (perhaps “realism”) that we don’t need for other areas.
Now this last one is so ubiquitous that many games that really are not supposed to be about fighting nevertheless include it. Call of Cthulhu is about investigation and being traumatised by gribblies. Fighting is mostly futile, and not a central part of the Lovecraftian genre at all. Yet it includes special skills for different kinds of weapon, and ratings for their damage levels (pre-7th edition, at least – I haven’t seen it). This is arguably bad design: it doesn’t seem necessary or particularly useful as a component of that type of game. Such games typically fall back on guidance to let you know what the game is “really” about, or a well-understood culture amongst the fan-base.
The point is, these are not “generic” rules or vanilla design choices, they support a specific type of play. Each of these rules reinforces that type of play, promoting a very specific experience. A single person plans and guides the experience, which is mostly about struggling with obstacles and fighting threats. Individual groups may graft on other aspects of the experience or house-rule or ignore the rules to get the experience they want: but a study of the anatomy of those games provides a clear view of what they’re designed to do.
[*] Aside: I’m not actually sure there are that many great stories about success and failure. When I think about stories I’ve enjoyed even within the broad category of “mostly procedural”, the heroes largely succeed unless they’re overmatched. When they’re overmatched it usually marks out the key obstacle of the story, which the heroes must then struggle to somehow overcome through cleverness, a macguffin, some kind of montage, whatever. It’s pretty rare for random success or failure to generate interest. But whatever, it’s a popular model.
I find it confusing the way that the MC Moves in Apocalypse World and its children focus on the outputs of the fiction instead of the entities in the fiction.
What do I mean by that? Let’s take an example: in Apocalypse World, you “announce future badness”. You have to decide for yourself what kind of badness we might be talking about and, notwithstanding the greater structure provided by AW’s clocks and fronts, it could essentially be anything from “a dangerous person with a stick of dynamite has arrived” to “the dangerous person with their dynamite is about to blow up the holding”. There’s nothing in the system of moves to distinguish between the first arrival of a threat, which is in itself a sort of announcing future badness, and the actualisation of the threat, the imminent inflicting of harm or other unpleasantness. It’s just “future badness”. Similarly, “inflict harm as established” is just harm, the move doesn’t tell you anything about where the harm comes from.
I find this very counter-intuitive. When I’m MCing, I am thinking about threats, characters, events – stuff in the fiction which might generate “future badness” or “harm as established” or similar. I don’t think “how could I inflict harm on someone right now”, I think, “What might Mr John Q Dangerous do next? Oh! He may inflict some harm with his stick of dynamite!” Every time I read through the list of AW moves I find myself having to translate it into a language I can understand.
So here’s how I would recast MC Moves. A list of Moves, from the softest of the soft to the hardest of the hard.
1. Introduce someone or something that has the potential to be a threat or an opportunity. A guy with a gun. A person with power. A mighty storm. Right now, their mere existence is all you’re announcing. You’re saying “here is a threat that isn’t currently threatening you”.
2. Actualise the threat or opportunity by giving that someone or something a reason to become threatening, or if they already had a reason, by giving them new access to or awareness of the players and their concerns. So, the guy with a gun realises the players are working for his arch-enemy. The person with power, who is looking to recruit new followers, notices the players in her territory and she decides to go after them. The storm enters the valley the players are in.
3. Activate the threat or opportunity by having it make a concrete move to harm the players or their concerns. The guy with the gun points it at the players, he’s about to fire it. The person with power sends a squad of goons to round the players up. The storm begins to lash the players and lightning flashes all around – they need to seek shelter or they’ll take harm.
4. Enforce the consequences if the players fail to block or evade the activated threat, or if they take appropriate action to take advantage of the activated opportunity. The players take harm from the guy’s gun. Having treated with the powerful person, she gives the players new equipment and authority. The players are given the condition “freezing and exhausted” by the storm.
That fourth category contains many of the moves in AW: inflict harm, trade harm, capture someone, take away their stuff and so on. But many of the AW moves cover the whole gamut of the first three categories: you can announce future badness by introducing a threat, actualising it or activating it. Similar thinking goes for opportunities, off-screen badness, and so on.
By the way, I realise this isn’t entirely novel. The countdown clocks do something similarish. But hopefully this provides a helpful general set of steps for moving a threat from first appearing on the horizon to being all up in your face.
I’ve long had an interest in designing a Dream Askew hack, thanks to very positive experiences playing the game tempered by some issues that I wanted to address. I’ve been tinkering around with this concept for quite a while, and finally managed to get a working prototype to a playtest this weekend.
The game is currently a fairly thinly reskinned version of the original, since I want to concentrate on streamlining and reworking the design framework. But I didn’t want to just copy Avery’s game, so I’ve changed the setting. Space Askew is set in the belly of a space station, where outcasts and misfits live in the shadows below a more prosperous settlement.
I felt that DA would benefit from some more in the way of relationship-building in setup. I’ve added a set of Hx-style relationship seeds to choose from, and some Hillfolk-style unrequited desires. In both cases the process involves choosing something yourself, then asking another player a question, the answer to which provides a completed background element.
I wanted a clearer and more intuitive set of MCing guidelines for running the Situations. I’ve brought the MCing system a bit more back towards the way Apocalypse World works, supplementing each Situation’s Principles and Moves with a set of general Principles and Moves, and giving clear guidance for when an MC makes Moves. The Principles are a bit different from AW’s, focused more on small-scale interpersonal drama than constantly shifting external threats.
I’ve placed question-asking at the heart of the system. When you want to create a bit of content for the world (a character, a location, a piece of technology, a rumour…) you don’t create it yourself; you ask someone else about it. This is true in setup and during play.
I’ve created a more developed process for deciding what scenes should focus on, and for deciding who is MCing at any given time.
I added a very simple harm system. Whenever someone tries to inflict harm on another character, they say what they’re doing in the fiction, then ask someone else what the outcome is. I want harm to be kept simple and fiction-based, and I want the decision to inflict harm to recognise that, once the bullets start to fly, you can’t entirely control the resulting pain and injury.
I’ve switched the psychic maelstrom to a Battlestar Galactica-esque set of gods who you make sacrifices to and petition for aid.
I’ve added a new skin called The Foundling, who was once part of a networked hive mind connected to a parent AI and has somehow become separated from the parent. They’re a bit like the Hollow in Monsterhearts, lacking a clear identity and anxious to understand humanity.
How did the playtest go?
Character gen was fantastic. Beyond my wildest dreams, really. It took 45 minutes and yielded well-realised characters with charged relationships but plenty of undefined space to explore in play.
The play itself went well, but I think that was more a testament to the quality of my players than the system I wrote. I suspect people were relying on their own habits of running fairly systemless games rather than my rules. (The excellent setup will have helped, of course.)
This was partly the result of my failure to effectively teach the rules, and I’m clear that the game needs a Lovecraftesque-style teaching guide that guides you through the rules systematically.
It was also partly the result of information overload. I should have realised this would be a problem, because I think DA was already pretty hard work and Space Askew added a bunch of extra stuff.
I hadn’t fully appreciated how Play to find out and Ask questions aren’t particularly obvious or intuitive ways to play. As the only person who had read the rules outside the playsheets, I needed to do more to explain this (and the teacing guide would need to include this).
So what am I doing next?
I’ve already begun work to streamline the amount of information in the game, reducing the number of Principles and Moves to a manageable level, and focusing on what is really core to the game.
I’m going to write a teaching guide which ensures certain rules that aren’t on the playsheets gets mentioned, that key principles are explained in more detail, and that character gen is more structured.
I’m going to turn the Situation sheets into something a bit more resembling a character sheet, complete with setup questions.
For those who haven’t seen it elsewhere – I have created a list of UK game designers. I have mostly done this because I intend to try and play and run more games by UK designers, and then… well, I got sucked into being completist.
As part of my endless quest to spend every waking moment of my life obsessing about game design, I have spent some time analysing the MC moves in Apocalypse World. And here’s what I learned: what at first looks like a long and fairly complex list of options could actually be boiled down to “say something, anything”.
Ok, it isn’t quite that simple, as I’ll explain in a moment. But let’s take a look at the basic moves (i.e. those which don’t come from a front).
Announce future badness, announce offscreen badness. This translates to “tell us something bad is happening, or about to happen”. Obviously, it also has to be something the players weren’t previously aware of, else it isn’t announcing, obvs. Minor subtlety: it has to be badness. Bad for whom? The game doesn’t specify, and at times it makes it clear that moves don’t have to be against the players, they can be against anyone. If that applies here too, then this is truly a flexible set of moves indeed – announce anything that could be bad for someone. In that case, I could just write “QED” here and be done. It probably doesn’t apply here, though.
Offer an opportunity (with or without a cost), tell the consequences and ask. This translates to “tell us something that might happen depending on what we do next”. Now, a particular corollary of this, combined with the “be honest” rule is that whatever that something is, it will definitely happen as described unless someone acts appropriately. So this move boils down to “say what is about to happen and then do it unless someone stops it”. Already we have the basis for essentially anything to happen.
Remember that all the previous four moves can be used with just about any time period attached. Future badness could mean “in the next five seconds” or “in a year’s time”.
Put them in a spot. This translates to “make something bad happen”, more or less. Ok, perhaps it’s a bit more specific – it implies they’re going to have difficult choices to make or challenges to overcome. But that pretty much boils down to “make something bad happen”.
Capture them, separate them, take away their stuff, inflict harm (as established), trade harm for harm (as established), turn their move back on them. Obviously, these are much more specific. The harm moves are the system’s means to link what the MC says to the harm system, and to prevent the MC from just killing a PC (the principles prevent you doing that anyway, but this bit of system reinforces it). The others are more-or-less just specific examples of someone being put in a spot i.e make something bad happen.
Make them buy. This is just a sub-type of offering an opportunity and/or telling the consequences. It amounts to a prompt to think about barter and other such trade/negotiation.
So essentially everything can be boiled down to “say what might be about to happen” or “make something bad happen”. This looks like it rules out nice things happening, but of course it doesn’t – if you offer an opportunity and someone takes it, something nice will happen. But clearly, for the most part, nice stuff is there to prompt the characters to action so they can get the nice stuff.
So this is maybe not quite just “say anything”, but it’s pretty close. It’s extremely close – maybe indistinguishable from – “say anything that the players won’t want to ignore”. (This seems jolly close to the Dogs in the Vineyard formula, from the same game designer of “do something they can’t ignore” when engaging in conflict.)
What’s my point? Well, mostly it’s just a bit of analysis I did, and I felt like writing it up. But it matters to me because, when I first started playing PbtA games, I remember staring at the moves list when it was my turn to act. Blinded by the sheer range of options. Paralysed, at times. But in practice, if I’d just fallen back on the principles and said something – anything – that the players would be expected to give a damn about, it would probably be fulfilling one of the moves.
I think the AW moves list is probably intended to function as a prompt, to help MCs mentally brainstorm their options in the few seconds before they open their mouth. Occasionally I think I’d find that useful. But I think for the most part I’ll just be saying the first thing that comes to mind, in future. I’m pretty sure I’ll end up sticking to the rules as I do so.
I was listening to one of the Metatopia panelcasts from last year, and the panelists[*] mentioned that there are different types of feedback and wouldn’t it be nice to have a way to say what kind of feedback you wanted. Well, I agree, and it’s something I’ve been meaning to write about. So here goes.
Before I start, let me say that when I send my games out for feedback (playtesting, normally) I always provide a list of specific questions. This is partly to ensure that specific things I’m wondering about get covered; it’s partly to avoid feedback I’ll find unhelpful; and it’s partly to provide a structure to help people think about the play experience. But anyway. Let’s talk through different kinds of feedback.
Drafting feedback. This includes identifying spelling and grammar errors, as well as areas where language might not be as clear as it could be. You might want this when your game is in its final draft form. You probably won’t find it that useful before that point, because you’ll be redrafting anyway.
Comprehension feedback. This is a bit like drafting feedback, but a bit higher level. It’s asking whether there are aspects of the rules that are confusing. Can you understand the game? This might be particularly useful for an early draft read-through. I normally check on it with playtesting as well.
Experiential feedback. What did the game feel like to play? Was it humorous or scary? Was a particular mechanic hard work? Did you get emotionally invested in your character? This is generally a key component of playtesting for me. I want to create a game that feels a particular way, and so I need you to tell me what it felt like to play it. That’s much less useful if you’re just testing out a mechanic in isolation, though. You also might not need it so much if, say, you’ve already playtested the game quite a bit and you’re just testing a modification to the original design.
Mechanical feedback. What happened, mechanically? Did you seem to crit fail constantly? Was there an exploit where you could build up unlimited bennies? Did some mechanics just never get used? Did anything break down at the table? You’ll probably want this sort of feedback at some point in playtesting, unless your game is super freeform. Some people like to playtest mechanics individually, outside the context of a full session. It’s not something I do, but worth considering.
Design advice. It is often said that it is very annoying when people try to design your game for you through their feedback. And generally, I do agree with that. But, sometimes that may be exactly what you want: you know something isn’t working in your game, and you want suggestions on what to do about it.
So, when you’re asking for feedback on your game, be clear which kind(s) of feedback you’re looking for and, where appropriate, which kinds you aren’t looking for. I would add that you can, and probably should, say which specific bits of your game you are asking for feedback on. If there’s a particular mechanic or aspect of play you want to hear about, say so! Even if there isn’t one particular aspect, you might want to break your game down into specific areas you want covered.
Of course, it bears noting that you might not always realise that you need feedback on something. Maybe you think your mechanics are working perfectly and you don’t need feedback on them. If a playtest reveals they broke down completely, I’d hope my playtesters would tell me that, even if I was only asking for experiential feedback.
I hope that’s useful. I’ve probably missed something. Comments welcome!
[*] I don’t know exactly who said it. Panelists included Emily Care-Boss, Julia Ellingboe, Avonelle Wing, Shoshana Kessock and Amanda Valentine.
So, I was reading some stuff about the OSR, and came across the concept that spot checks and detect traps rolls aren’t used in the OSR: instead, you identifying potential danger zones and have your character check them, and the GM tells you what you find. This article is about the more general case of this dichotomy: when is it appropriate to allow a player to describe their way to success, when is it appropriate to reduce it to a roll, and – in the absence of a roll – when to punish a player for neglecting to describe some particular action in the fiction. But yeah, I’ll talk about traps a bit because it’s a convenient example.
Caveat: I’m talking about games where you have Player Characters trying to overcome obstacles through skill or luck, and where those obstacles exist in the GM’s head or in their prep i.e. not invented after a roll is made. I realise not all games are like this, but that’s the scope of this article.
Some dudes are walking down a corridor. There’s a pit trap ahead. Do you have them make a detect traps check (as you would in, say, 3e AD&D) or do you hit them with the trap unless they take an action which will allow them to detect it and disarm or avoid it? Supposedly the latter is the OSR approach. He describes how the dudes, having lost their ten foot pole, look for cracks in the ground and then detect the edge of the pit trap by pouring some water on the floor and looking where it collects. Then they just walk around the trap.
Some dudes enter a room. There’s a moose head with a concealed compartment behind it. Do you have them make a detect secret doors check (as per AD&D 3e) or let them find it if and only if they investigate the moose head (OSR). In the example the dudes fiddle with the moose head and discover it slides to one side.
So, looking at “detecting hidden stuff” as a category of action, we can see that you can just skip over the business of describing how you find it and make a roll (perhaps the GM describes how you succeed or fail after the roll), or you can have the player describe in some detail what they actually do and judge what the effects of those actions might be.
We can go further, though: some hidden stuff will jump out and mess with you if you don’t detect and deal with it, some hidden stuff is something nice you’ll only get if you detect it. I think this is an important distinction. If the world is full of stuff that will hurt me unless I take the correct action, then this raises some questions:
What warning, if any, must the GM give me before the bad stuff happens? Is “there’s a corridor” sufficient warning that there might be a pit trap?
How much detail do I need to go into in my description? In the example, simply touching the moose head seemed to be sufficient to get it to slide, but what if I can only open it if I tickle the moose under its right eyebrow? Do I need to describe all the weird combinations of action I might take to get to that?
Now, this brings us to GM philosophy. Think about Apocalypse World’s “be a fan of the player characters”. In a system where you’re relying on detailed action description rather than “just make a detect hidden stuff roll”, it would be a dick move to have a moose head that only opens if you tickle its right eyebrow. That’s just too obscure. On the other hand, at the margin, it’s a total guessing game whether the particular hiding method you’ve decided on is too obscure, too easy, or just that nice level of challenge.
I suspect the OSR answer to all this is “who cares”. You’re going to get hit by traps sometimes, and sometimes they will kill you, and sometimes it’s because the GM put something in place that turned out to be a bit too obscure for you to pick up on it. Sometimes it will be because you were slopped and forgot to investigate the obvious moose head; sometimes it will be because you investigated the obvious moose head and it turned out to be a trap. But as someone not particularly signed up to OSR philosophy, the idea that my character’s life or death hangs on the question of whether the GM’s idea of fair warning and mine align, or whether the GM’s warning of a reasonable level of description and mine align.
Let’s think wider than hidden stuff. You may wish to base a category of action resolution on the players’ detailed description if you want your game to be about describing that thing in detail. (Duh.) If you like the idea of describing turn after turn of agonisingly detailed trap searching, weighing up the risk of wandering monsters against the risk of arbitrary death at the hands of a concealed trap, then OSR D&D clones may be for you. Equally, if you want a game that is about complex political negotiations, you might not want to boil every interaction down to a roll – you want to ensure there’s enough fictional positioning required that it feels like you’re actually negotiating, not just rolling a bunch of dice. Conversely, DON’T do that if you want to avoid such detailed description. If you make the intricacies of character position a crucial factor in a fight, then every time there’s a fight you’ll get painstaking description of character position, obviously. If your game isn’t about fighting, you probably don’t want that.
If you are going to make a category of action resolution all about player description (with or without dice rolls) then you’re also going to have to think about how to get everyone on the same page about that. Establish what a reasonable level of description is. Establish what fair warning is. This goes wider than traps: does my political negotiation description need to give the gist of what I’m saying, or the detail, and if the latter, do I also need to roleplay my impassioned, emotional argument, or just describe what I’m saying? If we’re not on the same page about this, I’ll be pissed off when you have my argument fail (or saddle me with a fat negative modifier to my roll) because you felt I wasn’t impassioned enough. I’ll be annoyed that you thought describing the quirk of your NPC’s eyebrow is fair warning they’re about to stab me in the face.
So this brings me back to the OSR. I read in the primer that OSR is about rulings, not rules. Fair enough; but one thing rules do is get everyone on the same page. Quite literally. If we all read the rules, we can have common understanding of how a given situation might play out, and even if we don’t then at least we have a fair way to check the arbitrary power of the GM. If we don’t have that, then that sense of fairness depends on the players and the GM being on the same page, metaphorically.
Anyway, what this has got me thinking is, there’s a space for an OSR-style game that provides exceptionally clear explanation of the above factors: how much detail is it reasonable to expect, how much of a warning sign is it reasonable to expect. Providing some parameters to your rulings, without forcing you to conform to highly detailed rules. Maybe it already exists? Comments welcome.