Transcript of April 5, 2014 ifMUD Discussion on Time Simulation

As last time, I have modified this transcript from the original: I’ve done some formatting, removed some typos, and added explanatory links for many games and references to past events. In some cases I have also reordered lines of dialogue when there were two separate strands of conversation happening, in an attempt to make the conversations easier to follow.

Meeting startup

Emily says, “hi, all”
DavidW says, “hey”
Emily says, “looks like we have reached the appointed hour”
Adam has joined the channel.
Beluki says, “Hi.”
Emily says, “this meeting we’re talking about how games model time, especially at the claim that (a) parser games usually model it in small, consistent-sized chunks, the duration of an action like taking something; and (b) that that has a strong effect on the kinds of stories parser games can tell”
Emily says, “there are some links to do with this idea here”
Emily says, “and David Welbourn has put together some sample texts at http://www.plover.net/~davidw/topics/if-time.html”
Jacqueline has joined the channel.
Emily asks, “so, with that: do we think that the current time-slicing approach of parser games is an inevitable feature of the form?”
Adam says, “For values of ‘current’ = 2001 or so… NEIN”
baf says, “Not inevitable, but natural. It’s what occurs if you don’t take steps to do otherwise.”
maga says, “I think it clearly isn’t inevitable, but the current world-model and verb-set makes it, if nothing else, much much easier”
zarf says, “natural to the standard library we’re used to, rather thn the form”
zarf says, “dislodging that library is hard, because both the authors and players stick to it, because they know that players and authors stick to it”
zarf says, “what maga said”
maga says, “(I am aware that this is #theoryclub, but it’s usually less useful to talk about what a form *could* potentially be made to do, rather than what the tools make easy)”
maga says, “(unless the next step is about making some tools)”

Action-to-Action storytelling in IF

Adam says, “Who’s read Understanding Comics? I thought that the different kinds of panel-to-panel transitions McCloud lays out shed some interesting light on the turn-to-turn dynamics of IF”
maga says (to Adam), “yeah, that analogy’s come up a few times in granularity discussions”
Emily says, “I’ve read it, though it’s been some years now. but I think I know what you are referring to”
Adam says, “Aha”
maga says, “so in McCloud’s terms parser IF is by default action-to-action”

Turn count, time per action, and other ways of measuring time in IF

DavidW says, “oh, I just thought of an example game that I should’ve included but didn’t: Adventurer’s Consumer Guide.”
Busta says, “I always imagine the amount of time something takes is inherent within the type of action (unless the game shows passing time)”
Jacqueline says (to Busta), “Agreed.”
zarf says, “hm, sort of”
baf says, “Hm. i was kind of thinking of Planetfall as an early experiment in varying the amount of time actions take, but even Planetfall was action-to-action in that sense.”
maga says, “well, yeah, the turn count is a highly wobbly measure of time unless the game does something that specifically makes it more concrete”
zarf says, “there’s no reason you couldn’t have a huge machine that you set up and then type ‘pull lever’ and watch the machine crank away for eight hours”
Jacqueline says, “Like, in The Fire Tower, I had some actions that didn’t take time at all, others took part of a minute, but if you hiked a stretch of trail the time advanced an amount corresponding to how long it would take you to cover that mileage. That was years ago, in I6. Parser IF is more flexible than people make it out to be.”
maga asks (of zarf), “but that’s very much a disconnect from the usual pace of play, no?”
zarf says, “I don’t think it’s a disconnect, it’s just soemthign you have to set up”
maga says, “like SLEEP, if it works, tends to be a chapter-break kind of thing”
baf says, “Fire Tower is similar to Planetfall in that respect, I guess”
Adam says, “In Varicella there is at least a little bit of variation in how much time things take – it’s one of the reasons that the time is updated in five-minute intervals”
zarf says, “it’s not weird when ‘enter castle’ or ‘talk to guard’ ends a chapter”
Adam says, “I didn’t want people to see that, ah, that took five seconds, but oh, that took three minutes”
zarf says, “(it would be weird if ‘take inventory’ ended a chapter and caused an eight-hour break)”
Emily asks (of Adam), “so you wanted to have the effect, but not draw attention to it?”
Emily says, “or at least not have players try to game it”
eu says, “We should probably clarify whether we mean variation in absolute diegetic time or variation in time as measured in meaningful events or meaningful choices.”
jmac says (to eu), “Right. I was wondering whether we are talking about the passage of time in the (in-world) subjective sense, or in the objective “the sun is moving across the sky in the game world while you do this” sense.”
maga says (to Adam), “and, right, even in a timed game like Varicella a player would be kind of missing the point to focus really hard on how action A is the same time-cost as action B”
Jacqueline asks (of Adam), “Interesting. Because you thought if they could see that it would drive what sort of actions they took or avoided taking?”
Adam says (to Jacq), “Actually, it’s more that it just seemed weird for the time to update itself in an obviously lurching way!”
maga says, “because it’s very rarely *interesting* to simulate time in a precise simulation-y way”
Busta says, “is seeing how much time has passed that important? Perhaps for some games with timed puzzled, but for the most part, I don’t really find it necessary.”
Gunther says (to maga), “Border Zone confirms rather than refutes that point”
maga says, “well, I think the central point we’re working up to is scale of action”
maga says, “(or ‘granularity’ as it gets called sometimes)”
Jacqueline says (to Busta), “I’m not sure The Fire Tower listed the time, but it modeled the sun rising, being high in the sky, and eventually setting, which is why time was measured.”
baf says, “Which is why most games are loosey-goosey about it and just track ‘turns’, where different actions can be described as taking very different amounts of time.”
Beluki asks, “Mmmm. How about a game like Aisle? does it count as timed actions?”
Beluki asks, “In particular, do actions such as “remember…” count?”
Emily says (to Beluki), “often actions like remember and look are made instant in games”
Emily says, “where time counts at all, that is”
nm says, “it does seem that it matters if the game has a particular attitude toward time — whether something must get done by a deadline, whether it’s a free non-lethal world in which little changes that one can run around in, whether the cycle of day and night is important”
Emily says, “okay — to back up a bit, the context for this question is not “how can we precisely model one-minute vs 30-second intervals” or even “how can we have scene breaks in IF”, since obviously those are doable”
Emily says, “the original contention was that the scope of action in parser IF is limited, generally, to things that take a small amount of time and whose ramifications take place on a small scale, and that this is limiting to the kinds of stories that parser IF can tell”
Jacqueline says (to Emily), “I would disagree. That’s what we’ve generally done, but it’s not impossible to take a very different approach.”

Library Requirements for a parser game with long-scale time

maga says, “okay, let’s make it a bit more concrete”
maga asks, “if you were going to design a standard library for parser games that took place at a longer time scale, what would it look like?”
eu says (to maga), “Because larger-scale affordances are hard to communicate, I would probably include a way for the player to preview the outcomes of their actions, to make it clear what verb X means in this context.”
Adam says, “Well, being an unfrozen caveman from the ’90s, I find that calls to mind one of Brandon Van Every’s posts”
Adam says, “Where he talked about how he wanted to type ‘I’M HUNGRY’ and have a whole scene unfold of looking for food”
Gunther says (to Adam), “400 words or more? I know a company just for that.”
nm says, “Sven!”
Emily says, “but okay, that suggests controlling the game by specifying motivations rather than by specifying actions”
maga says, “I started working on a game once that was mostly about exploring a city, and every action was a chunk of a day – like afternoon, evening, night”
zarf says (to maga), “My straw-man scenario is a game with actions like ‘hire Fred’, ‘fire Steve’, ‘put Amy on Pepsi project’ and the turn-scale is like a week”
baf says, “There was a game I remember from the Comp a year or two ago that was basically just a walk to a distant place, but the rooms along the way varied hugely in scale. Like, one was a shack and another was a desert, sort of thing.”
maga says (to zarf), “mrm, and that’s a lot more repeatable than the example I was thinking of”
Emily says, “I think part of the challenge is that once you get into this scale of actions, there’s probably not one simple standard library that would really fit it”
zarf says (to emily), “right”
maga says (to Emily), “well, no”
Urbatain says, “in Spanish, we have a system called AGE, Aetheria Game Engine or something, it is text engine for MUDs, but with a higly modeled world”
Urbatain says, “so the author uses it to create conventional adventures, but the system supports multiplayer”
Urbatain says, “So, a solution for time, was builded, for example”
Urbatain says, “if a player start to break a vault, the action would start, continue, and finish, in several amounts of time”
Urbatain says, “so other player casting actions in discrete amounts of time, would see the thief, breaking the vault”
Jacqueline says (to zarf), “Please never write the souless corporation sim game. kthxbai”
zarf says, “My example has an implicit scenario”
maga says, “but you could have a bunch of libraries for different things”
Jacqueline says (to zard), “It was great in that regard. :)”
Emily says (to zarf), “sure — I was more answering maga’s suggestion that we talk about a standard library for larger-scale actions”
Jacqueline asks (of Emily), “Do we need a standard library for that?”
zarf says, “the question of whether any one set of verbs could get as popular as the Colossal Cave standard is… probably out of scope”
maga asks, “which I suppose makes the question: what is it about the action-to-action scale that makes it appropriate for so *many* stories, or is that even the case at all?”

The concept of shoe-leather and meaningful choice density

Adam says, “One of the reasons this topic really interested me so much is that for seven years now I’ve been mostly working on screenplays, in which pretty much all time is real time (unlike in a novel, in which you can summarize), but you’re supposed to eliminate any and all ‘shoe leather‘”
Adam says, “And IF has a hell of a lot of shoe leather in it”
Gunther asks, “for those of us not in Hollywood, what’s shoe leather?”
Emily says (to Adam), “one of the explicit things in the Choice of Games instructions for new ChoiceScript writers is to leave out any decisions that aren’t big/meaningful choices”
maga says (to Gunther), “dross, deadweight, I assume”
Adam says (to Gunther), “Things like this”
Emily says, “it’s the way in movies people hang up without agreeing what time and location to meet later”
Adam says, “Oh, heh – that pretty much sums it up”
zarf asks, “can I get examples of IF bits that are not action-to-action in the McCloud sense?”
maga says (to zarf), “the bit at the start of Blue Lacuna is the obvious one”
Urbatain says, “this is the unique special system done in Spanish, that handles time in a simulated way, with actions during diferent times”
Gunther says, “ok, thanks”
Beluki says, “maga: perhaps the fact that we always tend to think step by step, logically advancing towards a bigger goal, rather than the inverse.”
Adam says, “I was going to say that in a lot of MST3K movies, there are whole scenes of people driving up driveways and getting out of the car and walking to the door”
Adam says, “As if viewers will be like ‘HOW DID THEY GET INTO THE HOUSE?!?’ if you just cut to the living room”
Urbatain says, “Other cases are conventional uses of time, with examine taking no time, or actions ending a chapter or producing a scene.”
Gunther imagines 2001 without shoe leather
nm says, “that was space shoe leather, though”
FloatingInfo says, “One of the prison scenes in All Roads may be relevant.”
eu asks (of FloatingInfo), “Which do you mean?”
FloatingInfo says (to eu), “At one point after you did an action while in prison it would establish that a lot of time passed.”

The aesthetics of large and small-scale actions

baf says, “See, okay. What Emily just said about Choice of Games explains what I didn’t like about the first Choice of Games game I tried. It eliminated all but the biggest of decisions, which meant I never got the opportunity to explore the texture of the gameworld.”
Beluki says, “Small actions feel more natural in a sense. It would be interesting to try the opposite.”
maga says (to Beluki), “yes, but even then we don’t *think* in terms of individual physical actions, even if that’s the atoms that make up our plans”
Beluki says, “maga: mmmm, I think a lot of times we actually do.”
maga says (to Beluki), “sometimes, sure, but I don’t think it’s our primary mode of planning”
baf says, “So I posit that the moment-to-moment, little-actions model is to some degree an aesthetic choice.”
Busta says (to baf), “That’s what I actually like about the ChoiceScript games.”
Jacqueline says (to baf), “Sure, but one born of ease due to the tools. I guess this is why people are saying that to do something different you’d need libraries, but it doesn’t really seem that difficult to just code this case by case because every approach will be innovative and creative and that’s part of crafting a good experience for the player.”
maga says (to Jacq), “in practice almost everyone in parser goes, hey, the innovative and creative approach will put an extra year of work onto the end of this project, to hell with that”
Jacqueline says (to maga), “Possibly, but libraries are a double-edged sword. They lower the time-to-release bar but create a more homogeneous approach to something.”
maga says (to Jacq), “sure, but more, different libraries would be less homogenous than having one standard approach that everyone always uses”
Jacqueline says (to maga), “True.”
Emily says, “something that Blood and Laurels did was to have explicitly scene-ending affordances alongside the moment-to-moment actions — so you could pick a conversation quip or have something to eat, or you could choose to go visit someone else (say)”
Emily says, “so it was trying to have its cake and eat it too, in this respect”
jmac says (to baf), “Yeah… it’s interesting to argue that CYOA replaces the parser’s sandbox-and-gateway player-experience pattern with, er, gateway-and-gateway.”
Roger says, “Anchorhead also had some big act breaks like that”
Roger says, “I think some games with hand-to-hand combat sort of switch into a combat-time mode when it arises.”
zarf says (to baf), “That kind of exploratory action is implicitly ‘no time’, or not enough time to matter. So in a way it’s orthogonal to time scale”
zarf says, “although I can’t point at examples that do it differently”
Urbatain says, “we have other work, Ocaso Mortal, it is a humor piece making fun of the medium conventions. There’s a situation where the player is a worm, and the scenery is very huge, with a hill upon the east. A single movement would get you there. It was funny.”
Ellison says (to Urbatain), “that kind of joke was also done very well in the game ‘The President of the United States of America‘”
Jacqueline asks (of Urbatain), “But when the worm reached the top of the hill, how many days had passed?”
Urbatain says, “yeah, the funny was in the absurd of the situation, the description talked aboyt a long long walk for a worm. there’s nothing more on it, just, a joke”
jmac says, “In many ways, my ideal IF is written around exploration, and signals to the player ‘Go ahead and poke around as much as you want, and let me know when you’re ready to move the scene along. (Or when you figure out how to do that…)'”
Ellison says (to jmac), “I agree”
baf says, “One of the things that made Howling Dogs so interesting was the way that it varied the scale and scope of the actions. The hub area was very moment-to-moment and adventure-gamey, which might have helped the work as a whole be more palatable to adventure-game fans despite its more daring aspects.”

Zooming between levels of time representation

Emily says (to baf), “yeah — and I think the line I’d take on this myself is that parser IF can do large scale actions; that what it has trouble with is shifting gears and doing small-scope stuff and large-scope stuff in the same game”
Emily says, “because all the craft is about teaching the player a consistent verb set”
zarf says, “I buy that line, re shifting gears”
Emily says, “and so you have to somehow communicate the idea “okay, now you’re not opening boxes any more or picking dialogue, now you’re in a montage sequence and your options are >TRAIN AS THE SLAYER or >DRIFT AWAY FROM GILES””
Adam says, “Hmm”
Roger says, “hmmm we’ve had a goodly number of changing-the-PC games; did any of them include a wildly-different sense of time? Seems like an obvious way to approach it.”
zarf says, “again, examining and taking inventory can be swept aside as ‘zero-time’ but the rest of your verb set tends to have a steady pace”
Adam asks, “Other game types tend to involve a lot of zooming in and out between levels of action, right?”
maga says, “or two standard approaches, I guess, of ‘do a Adventure-style parser game’ or ‘do a choicey game with minimal world-model and little repeatable action'”
Jacqueline says (to Emily), “Never drift away from Giles.”
maga says (to Adam), “yeah”
Beluki says, “Emily: perhaps it would be easier in games where say… a player performs small actions that have a big impact over time or viceversa (so we don’t have to mix both, they are already mixed).”
maga says, “and here inevitably I bring up Microscope and only a handful of people know what I mean”
Adam says, “Like, in a war game, you might be looking at a map of Europe and moving armies around, or you might be in a battle scene and actually just being a guy shooting a gun”
Urbatain asks, “”don’t you think those changes in pace from action to another, would kill the flow?”
zarf says (to adam), “almost always exactly two levels, I’d think”
maga says, “but the multiple-level approach is most interesting when the player can *choose* where to zoom in, which is impractical in a hand-crafty medium like IF”
Busta says (to Emily), “You could allow the game keep track of your actions or dialogue choices, and make the decision for you during the montage sequence. For example, if your conversation with Giles didn’t go so well, the next scene would not be training with him.”
Emily says, “the recent Telltale stuff is interesting in this regard, because there are passages where you’re wandering around looking for food or trying to find clues, and others that are much more dramatic, and this mostly works, though occasionally I think “I’m tired of inspecting this sidewalk for bloodstains and I want to go back to talking to Snow White” or whatever”
Jacqueline says (to Adam), “That is a brilliant game idea. How do your decisions as a general in where you move your troops translate into the field experiences of your soldiers? Or any other leader / led scenario.”
eu asks, “So, is it accurate to say that the challenge is not so much communicating variation in scale, but in getting variation in scale to play nicely with player input?”
Adam says, “One thing that gets me about this discussion is that in writing a book, you seamlessly ‘zoom in and out’ of different levels of action all the time”
zarf says, “sure”
Adam says, “The idea of being consistent about the way you treat time is bizarre in that medium”
Emily says (to Adam), “right — and that’s part of where this conversation is coming from, the question of whether there are aspects of the high-level experience in novels and the like that are just not doable in IF”
Emily says, “or parser IF, anyway”
zarf says, “the problem is not the zoom, the problem is getting the player to naturally engage with the proper zoom level at the right point”
zarf says, “(at any point)”
Jacqueline nods.
maga says (to eu), “yeah, it’s about building a) a world model that works well at your given scale, and b) setting up a bunch of verbs which play nicely with that model”
Emily says, “you could do what Busta suggests and use the player’s small-scale input to determine what happens in the more zoomed-out parts, but I’d kind of like to allow for large-scale character decisions sometimes too”
Adam says, “Hmm”
FloatingInfo asks, “Could a > vs >> vs >>> be relevant in establishing when you switch?”
Jacqueline says (to FI), “Intersting idea.”
zarf says, “(‘Don’t tell me about the Prague Spring getting crushed, I want to mix paints on my palette here’)”
Emily says, “like, you have the close-in scene of your first date with Fred, and then you zoom out and control the relationship at a more macro level for a while, and then you drop back in six weeks later to play out how it’s going”
Adam asks, “So the idea is that the system needs to be robust enough to know how to deal with both ‘>TAKE STAPLER FROM DESK’ and ‘>QUIT JOB THEN JOIN CIRCUS’ at any given point?”
Adam says, “‘needs’ is the wrong word”
eu says (to Adam), “I don’t think that’s the contention.”
eu says (to Adam), “Mostly regarding “at any given point”.”
maga says, “well, my basic reaction is ‘that is insane'”
nm says, “seems to me that there are problems in the underlying simulation, too, not just in cueing the player”
nm says, “if you really want to model the sorts of actions/responses that can happen moment-to-moment, hour-to-hour, day-to-day, etc.”
baf says, “Seems to me that a certain homogeneity is part of the promise of parser-driven IF. You expect to be able to issue any command in any context where it makes sense.”
maga says, “I think players need modes of interaction in which they can feel comfortable”
maga says, “and switching scales arbitrarily between actions will make that really hard”
maga says, “(not to mention the trouble it makes for authors, jeez)”
DavidW says, “The > vs >> trick has been used in games where >> means to answer from a limited set of responses, usually YES or NO.”
Adam says (to DW), “Yeah, E,N does the >> trick to specify different ranges of action”
Emily says (to maga), “I think the alternative is to have different explicit modalities at different times”
maga says (to Emily), “that is a lot more interesting”
Busta says, “The easiest way for large-scale actions might just be a switch to a CYOA list format”
maga says, “or just to have more games that aren’t action-to-action verbs, medium-sized-dry-goods world models”
Adam says, “I am a firm believer in limiting verb sets – i.e., if people know that they can do one of three micro-actions or three macro-actions at any point, that’s less insane”
eu nods.
baf says, “Hm. E,N was unusual in that it had explicitly separate world models, divided by geography.”
maga says, “I mean, let’s not worry about the switching part until we can get different-scaled games working in the first place”
Emily says (to maga), “even if your close-in verb set is about dialogue and social action, though, you still have a challenge varying that with the larger scale”

Combat and scene-specific action scales

Urbatain says, “oh! I remember that one of my games has a bullet time moment a-la Matrix”
zarf says (to urbatain), “hey, I like that notion”
Urbatain says, “it was indeed a shooting with pistols”
jmac says (to zarf), “I’ve seen you do a demo of that in parser, in fact. With your kung-fu WiP”
maga says, “The bullet is slightly closer to your face. Your trenchcoat becomes marginally swooshier.”
eu says, “Heh.”
maga says, “>ENGAGE LEFT BUTTOCK MUSCLES”
Jacqueline says, “But we digress.”
Adam says, “Aieee”
Adam says, “Text QWOP”
Urbatain says, “the thing is that the game has problems with the pace, because, when a certain action or scene has a high pace, a normal action, like examine, would kill the flow”
Emily says, “I’m not sure this is enormously different from a lot of AIF”
zarf says (to jmac), “Hm. I guess. It’s more fun with bullets.”
Adam says, “Hmmm”
Urbatain says, “but I think I’ve solved this by capturing the moment and writing down a lot of responses for that particular moment..”
DavidW says, “hm. I seem to remember Rans had a fight sequence where the verb list shrank down to STAB, PARRY, THRUST, DUCK, or something like that until the fight scene ended.”
Adam says (to Emily), “That’s an interesting point”
maga says (to Emily), “I think most AIF is action-to-action, and the stuff that isn’t is more about repeating actions”
Adam says, “It also reminds me of Galatea, oddly enough”
Emily says, “so something I’ve argued in the past is that if you want a high-intensity scene in IF, you need to pare down the possibility space of player action — e.g. trapping them in a room with few irrelevant items, ruling out most verbs, and so on”
maga says, “sure”
Adam says, “I wrote an AIF thing during the 2004-2011 period that I was releasing everything anonymously”
Adam says, “It is interesting that there are two levels on which that genre functions”
Adam says, “I.e., you have your explicit sex verbs, but you can also use higher-level verbs like leaving the room, etc.”
Emily asks (of Adam), “are you just going to put that out there mysteriously so we all have to study the AIF back catalog to find out what it was?”
maga says, “aha, the mystery of GoblinBoy revealed”
Jacqueline says (to maga), “yay”
Busta says, “Creating a high-intensity moment in IF might be a good topic for a future theorything. It’s something I’m interested in learning about and finding good examples.”
Urbatain says, “Busta my was build on that, but I simply just copied the structure of certains actions in The Awakening”
Busta says (to Urbatain), “I’ll have to check out The Awakening then.”
Urbatain asks (of Busta) “did you play it?”
Urbatain asks (of Busta,), “did you play The Awakening long time ago?”
Busta says (to Urbatain), “No I haven’t. Then again, I haven’t actually played much :/”
Urbatain says (to Busta,), “then I will not spoil that, it’s the ending ;)”
Busta says (to Urbatain), “heh ok”
Adam says, “And Galatea is similar in that you have the ask/tell stuff, but you can also back out a level and do regular IF stuff”
Emily says (to Adam), “I don’t really think of those as different *scales* of action, though”
Emily says, “I mean, maybe they are, in that I suppose it takes longer to say a sentence of dialogue than to touch someone’s elbow”
Emily says, “but it’s not orders of magnitude different”
Adam says, “I guess it’s the minutiae of the primary activity vs. the broader actions of other activities”
maga says, “in AIF walkthroughs you often get a normal walkthrough until it gets to the sex part, whereupon the walkthrough author will say something like (scene) or (…) until normal puzzle-solving resumes”
Emily asks (of maga), “meaning that it doesn’t matter what you do at that point?”
Emily asks, “or that they’re too shy to write it out?”
maga says, “probably a bit of both”
maga says, “(there are games where actions in the sex scenes have more story repercussions, and in that case walkthroughs treat them as part of the normal action)”
Emily says, “so from my point of view, the reason I want to be able to change levels is that I want to give the player the sense of investment and up-close experience that comes from being able to explore a space or do line-by-line dialogue or similar, but I also want to tell stories that have a long enough scope that they allow for significant character development or relationship change”
maga says (to Emily), “and you want (some degree of) player control at both levels, is the trick”
Emily says (to maga), “right”
Busta asks (of Emily), “could you give an example where the larger action could not be inferred by one or series of smaller actions?”
Emily says (to Busta), “oh, no — almost any player character decision could be communicated in dialogue or through thought, but that doesn’t mean that’s the best scale at which to tell that part of the story”
Busta says (to emily), “ahh gotcha. That makes sense.”
Emily says, “like, if I want to move on to the player’s freshman year in college and I want it to matter what she’s put on the roommate selection form, that doesn’t necessarily mean I also want to laboriously have a scene at the breakfast table where she talks through the form with her mom”
maga asks (of Emily), “but that’s kind of a unique event, no?”
Spinster says, “The dialogue comparison is interesting, cause it’s one of the most frequent cases where a player’s range of actions is changed significantly.”

On superobjectives, objectives, and actions

Adam says, “David Mamet’s ‘On Directing Film’ has a chapter that touches on these different levels”
Adam says, “He’s workshopping a scene with some students”
Adam says, “And they break down what the character is doing into superobjectives, objectives, and actions”
Adam says, “For example”
Adam says, “The superobjective: win the respect of the instructor”
Adam says, “Objective: show up early”
Adam says, “Action: open door”
maga asks (of Adam), “so you’re thinking of an action/objective tree, sounds like?”
Adam says, “Traditionally, parser IF has dealt with the action level in that schema – I’m wondering how well it could deal with the broader ones”
DavidW says, “(I can’t help but think ‘kill lion’ right now; forgive me.)”
nm says (to Adam), “well the higher-level ones are often the player’s goals”
Adam says, “Hmm”
baf says, “Hm. Now you’ve got me imagining a system where you have ‘objective commands’, and only have to drop down to ‘action commands’ if something interferes with them”
baf says, “Superobjective commands would probably be a little weird.”
Adam asks, “I was never big into D&D, but doesn’t roleplaying often involve various levels working simultaneously? Like, don’t players switch back and forth between ‘I pick up the sword’ and ‘I search the village for clues’?”
eu says (to Adam), “In RPGs I’ve played, yes.”
maga says (to Adam), “right, but there you have actual humans directing the story, and that kind of zooming in is easily handled”
baf says, “Adventure games were originally an offshoot from D&D, which is almost entirely played in moment-to-moment actions. But even the oldest D&D modules have notions like a table to roll on if you ask around in town for information about the dungeon.”
zarf says, “this whole superobjective thing turns out to be a big part of Hadean Lands which I can’t talk about much but I figured I should torment you all.”
zarf says, “you’re welcome.”
eu says (to zarf), “So the HL action automation might be worth mentioning anyway.”
zarf says, “What I’m hoping is that it works if the high-level actions become interesting once the low-level actions are familiar and boring”
Emily says, “Eric Eve has this movement library where you can say “GO TO FOOBAR” and it sets Foobar as your objective and remembers to keep moving you towards that location each turn if you choose “Continue” (or C), but you can stop along the way and do other things if you want”
Busta says, “I’d prefer to just arrive at the Foobar with some kind of passage text inserted.”
Emily says (to Busta), “that’s what I prefer myself, but it doesn’t always work if there’s the possibility of timed events or moving people interfering”
Busta says (to emily), “Yeah, hopefully the system would be smart enough to know if you would get interrupted.”
Urbatain says, “there were system that didn’t show nothing in the journey, but if something happening halting the journey”
Urbatain says, “for example, later Level 9 games”
maga says (to Emily), “that seems like a convenience more than a scale thing, really”
maga says, “(travel is definitely something where action tends to scale inobtrusively, though)”
Emily says (to maga), “yes, but it makes me wonder about systems that would allow for having the player have certain registered objectives and then automating those actions unless you belay the order”
maga says (to Emily), “well, hmm”
maga says, “that means that the game still has to work at any point where you halt it, so you now basically have a huge action-to-action game that you can fast-forward parts of”
maga says, “which seems like a dangerous design approach to me”
Adam says, “Narcolepsy had a hierarchy model for traveling – you could move to either one of the micro-locations within your macro-location, or move to a different macro-location to see what micro-locations were available there”
two-star says, “Roguelikes often have a thing where certain actions take a much larger number of game clock ticks than others, so if you are atempting to forge a sword, you will watch your pet move in brownian motion for a bit.”

Actions representing sub-actions

baf says, “So, there’s the notion of higher-level actions that are basically macros for a bunch of moment-to-moment actions, but it still strikes me that there’s a possibility for a more abstract approach.”
eu nods.
Spinster says, “Actions that encompass a lot of smaller-level actions amount to essentially an AI planning problem, if they need to be generated automatically in any way”
eu says, “Yes, macros still seem fairly at odds with what was in, say, Choice of Broadsides from the reading suggestions.”
maga says, “yeah, I am sceptical about heavily-simulationy approaches to this”
maga says, “(if only because they seem likely to make most of the small-scale stuff bland, and most of the large-scale stuff processor-expensive)”
Emily says, “yeah, Choice of Broadsides is alternating very close-in options with very zoomed-out ones (“say this specific rude thing to the first mate? spend the next six months studying sailing?”)”
Emily says, “and that seemed like one extreme of this idea”
Urbatain says, “I don’t know if this is related to the topic, but yes, Shade by Zarf did great making high level actions, more fluid, so the player would not waste typing time”
baf says, “Like, to take Adam’s example: superobjective ‘arrive early’. This is an atomic action. At a certain point in the story, there is a rule ‘instead of arriving early, go into the arrested-for-speeding scene’.”
JamesPKing says, “I’d like this to happen every time, but avoiding the “KILL LION” thing, unless you killed the lion at least once already.”
JamesPKing says, “I remember one game in which I had to open 2 doors and lock them back every damn time I was going in/out a station. I’d have loved a GO IN/OUT macro once I “solved the puzzle””
Urbatain says, “And yes, The Nemean Lion was quite funny about”
eu says, “ARRIVE EARLY is a good example of a command difficult to squeeze into the parser world.”
Urbatain says, “In a sense, a conventional IF forcing the player to make all atomic actions feels literary weird. without a proper pace”
Adam says, “Hmm!”
Emily says, “I feel a bit jerky repeatedly referring to a game no one can play, but with Blood and Laurels the larger-scale decisions typically did one of two things: make changes to the stats of your character, your relationships, and other characters that would then affect how small-scale scenes played out; or select which scene would happen next out of several possible scenes”
Emily says, “so it wasn’t super-simulationy at the high level but there was enough there to interweave with the lower-level stuff”
Adam says, “When I took John Searle‘s class back in the day, he talked about how there was a difference between commands such as ‘IMITATE EISENHOWER’ and ‘RESEMBLE EISENHOWER’ – maybe ‘ARRIVE EARLY’ falls into the latter group, as well as being hard to implement”
Busta says, “Or you could count the number of actions taken before leaving, and if less than a certain number, you would arrive early.”
eu says (to Busta), “Right, but then it’s not an atomic choice.”
Spinster says, “Very few IF games do adverbs.”
JamesPKing says, “An old Sherlock game had you examine CAREFULLY things to discover evidence, iirc. But that was bland. And pointless, as in the end you carefully examined everything”
maga says, “now I’m thinking of something a bit like Bleach of Etiquette with maybe a bit of Bee thrown in”
maga says, “where the high-level stuff is about selecting (somewhat repeatable?) scenes, and the low-level stuff is about getting specific outcomes out of those scenes”
AdamM says, “I’m wondering what the pay-off of simultaneously available action and objective verbs would be for the player. My experience has been that those two types of interaction have very different rewards.”
AdamM says, “The reward for a sequence of actions is usually the achievement of an objective, which is satisfying because you’ve executed a plan based on some degree of insight into the world model.”
AdamM says, “I think people generally expect objectives to be rewarded with consequence, though. Which is expensive.”
maga says (to AdamM), “well, yeah, Choice of Games handles this by making most consequences generic”
Emily says, “I’m not sure they do need to be simultaneously available; I think the idea of having some in-between-scene passages where you can make the bigger decisions tracks well with how this works in literature and movies”
Emily says, “but yes, you do need to attach some consequences to the larger-scale choices”
maga says, “this is the stage of the conversation where I want to stop talking and go and make a thing”
maga says, “which is inconvenient”
zarf says (to maga), “I will make an announcement on that topic at the end”
Urbatain asks, “what about a Proust’s muffin?”
eu asks (of Does), “anyone know if there are choice-based games where choices of different scale are presented in the same menu?”
eu says, “Er, Does anyone…”
zarf says, “like ‘fuss with paperwork’ versus ‘go home for the night?’ surely yes.”
Emily says, “yes”
Adam says, “I thought that a little bit above here someone was talking about ‘say [a thing]’ vs. ‘do [something] for six months'”
Emily says (to adam), “I was talking about Choice of Broadsides, and in that particular case those are the sorts of options that appear in different nodes, not in the same node”
Adam says (to Emily), “Oh, I see”
two-star says, “It might be interesting to allow setting objectives to change what low-level actions are hinted in a scene.”
eu says (to zarf), “Hmm, I mean more in the sense of crossing the objective/action boundary that’s been discussed.”
zarf says, “that I don’t know.”
maga says, “to me that sounds like something you’d mostly use for comic effect”
Adam says, “Well, I’m sure I have seen radically different scales in the same menu in amateur pieces”
maga says, “‘Eat the cupcake’ vs. ‘Give up and become a nun instead'”
baf says, “‘To eat a muffin, turn to page 6. To talk to the waitress, turn to page 38. To overthrow the government and institute democratic reforms, turn to page 12.'”
Emily says (to eu), “I started working on something like that in Twine, where you could tweak your intention and then it would pop up different options for what to do next in response to that; but I got bogged down messing with macros and didn’t get far”
eu says, “Hmm. Yeah, that’s more what I meant.”
Emily says, “but the idea was to try to make a CYOA in which it wasn’t immediately apparent what all the options in a given node might be, because you’d see only a subset of them; and part of it was about experimenting with who your protagonist was”
Emily says, “in order to open up what options your protagonist would then have to act on”
Gunther says, “[Punch clown] [Train for six months then punch clown] [Become a master of insults then insult clown]”
maga says (to Gunther), “now I am reminded of the CYOA that Edward Gorey wrote”
maga says, “which is really the best thing ever”

How much can be accomplished with writing alone

Urbatain says, “This is not related to the topic but I have a point: when remaking Dracula Part Two, it won an award for best NPC in the Spanish awards. The funny thing is that Dracula was not implemented at all as an object, it was all pure literature”
Urbatain says, “So, maybe one can do time changes just with writing”
Urbatain says, “without the need of a system or simulation”
Emily says (to Urbatain), “that’s certainly true, and people certainly have”
Busta says (to urbatain), “yup yup.”
icouldntcomeupwithaname says, “all the best NPCs are at least 90% writing.”
JamesPKing says, “I suppose that kind of thing is easily done in Choice-based, while terribly hard in parser”
Urbatain says, “Yeah, so 90% of pace and interesting work in time could be done just wrtiting”
maga says (to Urbatain), “sure, but only if you ignore the player’s involvement part”
maga says, “which is always what messes everything up”
JamesPKing says, “I can see the “A) jump; B) Get to Nirvana” in Twine. Can’t think of how to do it (without being silly) in Inform”
icouldntcomeupwithaname says, “also depends on how linear the game is. open world probably needs more scripting.”
zarf says, “(‘Nirvana is to the east’)”
maga says, “(which is a signal that the TV is not really very important and you shouldn’t bother with it)”
Urbatain says (to maga), “yeah, I was thinking preciselly in that. As I said before, you need to capture the moment, so system actions, or atomic actions, or out of context time actions break the pace of the secene”
Adam says, “Well, what are we talking about here”
Adam says, “I mean, this sort of thing has been done for ages:”
Adam says, “>E”
Adam says, “You walk into the living room, sit down, and watch TV for a couple of hours. But then you start to get restless again.”
Adam says, “So, sure, you can have time pass, with summarized actions, accomplished solely through writing”
Emily says, “I suppose what I’m talking about (and the reason I’m interested in the parser issue) is that I would like the player some of the time to form intentions and plans concerning higher level/zoomed out behavior, which they are then able to execute because there is enough of a system in place”
Emily says, “and it’s hard for the player to do that if you’re giving him only unique CYOA choices to work with, though you can certainly present world-model-based affordances in a point-and-click way”
Urbatain says, “but if you have a shift of time in a concrete action, if that mood and pace begins and ends in an action, there’s no problem, for example, changing a chapter.”
Urbatain says, “the problems comes when you have a concrete pace and mood for a whole scene”
maga says, “I am thinking about this more and more as an issue of modular design, rather than zoomable simulation”
eu says (to maga), “Yeah, that makes sense.”
zarf says, “Hm. I guess I’m dodging the planning problem in HL by having most plans be pretty well fixed, and the only question being whether you carry out a particular objective or skip it.”
zarf says, “(and there are several level-2 ways to achieve a level-3 goal)”
eu says (to zarf), “Which I bet works well in HL. But in other stories, I could see that leading to the same problems as in Emily’s post about the ring progress meter.”
AdamM says (to Emily), “That makes me wonder about CYOAs with a consistent verb set.”
AdamM says, “Possibly something like Long Live the Queen.”
maga says (to AdamM), “well, or Fighting Fantasy”
Busta says, “I was totally going to mention Long Live the Queen earlier. But a lot of the visual novel stat-raising or dating sim games are similar.”
Adam says (to Emily), “One of the ideas behind Pantheon (which is now looking like a project for the 2020s, if then) is that different gods will become your patrons based on the actions you have taken up to that point – you can try to choose actions hoping that they will win a god’s favor, but it’s not a one-to-one thing”

An idea of selecting scenes within which there is more simulation

Urbatain says, “I remember in Versu, correct me if I’m wrong, where you can build scenes, and those not where tailored to a place.”
maga says, “right, Versu scenes are by nature sort of free-floating”
Urbatain says, “but our IF systems are location based, so that would be difficult to define a concrete pace for a complex scene not dependant of a place”
maga says, “(although you can narrow them down by conditions until they effectively form a tree, and a lot of people do exactly that because trees are easier to grasp)”
Emily says (to Urbatain), “in Versu there was usually a room in which they were taking place, notionally, but the location wasn’t necessarily very heavily simulated. you could do that with Inform too, if you wanted, especially if you stripped down the default verb set”
Adam says, “I think that was pretty much inky’s review of Narcolepsy right there”
Emily says, “Versu also had a notion of when a scene was running out of dramatic content, and could introduce new events or bring the scene to a close, which helped with the pacing issue”
Urbatain says (to Emily), “great”
maga says, “(no, wait, I thought Varytale even though we were talking about Versu)”
Emily says (to maga), “oh, yeah, in Varytale there’s not necessarily any location modeling at all. I guess you could introduce it as a variable the way you might in Twine or something, but there’s no unique concept for it”
Urbatain asks, “do you think any IF system are (sorry my wording) bad for the issue at hand?”
Spinster says, “IF parsers only handle macro-actions with much cajoling”
Emily says, “I think it’s easier to do this specific thing in CYOA systems; I’m not sure among parser systems that one is easier than others, though possibly some of the verb-hinting stuff in Quest might make it easier to communicate possible action scope during zoomed-out scenes”
eu says, “None that I can think of. More that, in some systems, you would have to shed a lot that’s prebuilt and start from scratch.”
icouldntcomeupwithaname asks, “(in story driven games) maybe removing the concept of physical movement (of the player) altogether? or perhaps to make the physical movement connected with story rather than commands?”
maga says, “well, been done”
Urbatain says, “I’m thinking in systems like Storynexus or Varytale, but of course, those systems has other interestings thing, just I feel they could fail sometimes on the pace”
maga says, “the trick is not so much about what you take *out* – it’s what you need to put *in*”
Emily says, “Varytale is a bit easier to control from the pacing perspective because you can have very long individual storylets with a lot of content, whereas StoryNexus gives you a kind of mandatory granularity (unless you do stuff with forced follow-up storylets, I think; I know Zero Summer did a bit with that, but most StoryNexus pieces I’ve played don’t)”
Emily says, “if you merged the Varytale idea with a parser system, I think what you’d get is something where individual scenes played out as normal (or maybe with movement stripped, or somethign like that), but where the player could to some extent control which scene happened next out of multiple options”
maga says, “that seems pretty appealing to me”
maga says, “that’d be a good start, at least, until you could build more interest into the upper-level action”
eu asks, “So big actions become a kind of nonlinear Table of Contents?”
maga says (to eu), “that’s basically what Varytale is, yeah”

Conclusions

Emily says, “but it seems like we’re converging on: a) this is hard; b) no existing libraries handle it; c) possibly it’s not even reasonable to think about having a library to handle it because what the high-level action needs to be would vary so much from game to game; d) there are very few parser games that do systematic large-scale action to look at even for an idea of how that would work; e) if one did have different action scales, it would probably work best to present the player with different modalities of action at different times rather than try to make all the verbs work simultaneously;”
eu says (to Emily), “That sounds about right.”
Emily says, “and most of the suggestions about how to present large scale actions we’ve come up with so far are effectively some form of “give the player a menu””
Emily says, “though I think I can imagine some cases where it might be possible to do something else given a sufficiently curtailed context and verb set (the HIRE FRED / FIRE ALICE idea)”
Emily asks, “so I guess the flip side of this question is — is this really an area where the answer is just “CYOA wins”?”
maga says, “mrmf, not really”
FloatingInfo says, “One could use micro-actions to imply large-scale actions like GIVE BRACELET TO RUME in Blue Lacuna.”
Emily asks, “as being easier to design and think about and providing a better player experience?”
Jacqueline says (to Emily), “Great summary.”
maga says, “I think the answer is that it’s difficult to come up with a shiny new verb set/world model off the top of your head in a theoretical discussion”
zarf says, “‘more work needed'”
AdamM says (to maga), “But a number of people have been thinking about this for some time.”
Adam says, “Yeah, I’d also say that it’s hard to know how to respond without an a specific thing I’d be trying to implement”
zarf says, “so now, like maga, I want to transition to writing an example”
Emily says, “well, so it sounds like maybe we’ve hit the “time to SpeedIF” part of the evening”
Emily asks, “unless people have anything else they want to bring up first?”
Busta says, “The verb set is doable. The other problem is people enjoy parser for having to come up with correct actions on your own, whereas a new verb set would require a specific tutorial for the verbs that are available.”
baf says, “Menus are a solution to the problem of communicating to the player what the options are. Curtailed context is another.”
zarf says, “having a well-known set of tropes (cliches) in a situation is another”
baf says, “But even the Infocom-standard moment-to-moment verb set has this problem. It’s just that we here in this room have had decades to habituate to that particular command set.”
zarf says, “(these overlap)”
Emily says, “(yes)”
JamesPKing says, “the difference is that in CYOA the possibilities are given and numbered, while in parser you (usually) have to find them by yourself. I guess that goes for perfectly different audiences”
JamesPKing says, “if you remove the story from the equation.”
baf says, “I guess I’m kind of wondering if there’s a set of larger-scale verbs that could be as generally-applicable as the moment-to-moment ones, or if every work is doomed to making up its own novel scheme.”
zarf says (to baf), “I haven’t come up with one.”
maga says (to baf), “I think that’d require a larger-scale game that was both popular and suggested other games in a similar idiom”
Busta says, “I would never think to try FIRE ALICE on my own. I would be going through the dialogue options trying to find the one that actually fires Alice.”
Emily says, “when I’ve thought about this in the past, I’ve generally wanted to do something that involved machiavellian manipulations where the verbs were on the scale of BLACKMAIL, SEDUCE, BETRAY, and so on”
Emily says, “it is possible that that was biting off more than I needed to”
maga says (to Emily), “right, that’s combining the large-scale problem with the problem of how to simulate highly complex NPCs”
JamesPKing says, “newcomers who are not used to Infocom-style parser games tend to try the most “absurd” things. I’ve found out in scripts from IFComps.”
JamesPKing says, “On the steps of the PC own house in Awakening, somebody typed “GO TO WORK”. Along with “WALK” and “GO”, which were actually far easier to accomodate”
Spinster says (to JamesPKing), “A parser also requires actions to be well-described procedurally. “Get” has a fixed effect. “Disguise” doesn’t.”
Emily says, “I mean, I guess my reduction-to-absurdities argument that a parser game could in theory do long-scale time is that it is notionally possible to implement Diplomacy with a parser”
Busta says (to Emily), “My current WIP has FLATTER or INSULT that can be done anytime during conversation, though I’d still say that’s small scale.”
djfletch says, “the standard verb set might be special in that it mostly uses verbs that are in the list of 1000 commonest English words and therefore easy to guess (n.b. I haven’t actually checked this)”
baf says, “And that’e where hypertext definitely has an advantage: it trivially avoids the need for simulating minds by limiting the player’s actions to things the author has explicitly handled.”
icouldntcomeupwithaname says, “what about a CYOA with a parser? that could either be the best of worlds or the worst.”
icouldntcomeupwithaname says, “CYOA hubs/scenes with a parser to figure out what to do.”
eu asks (of icouldntcomeupwithaname), “What do mean by that?”
Emily says (to icouldntcomeupwithaname), “there are lots of parser games that handle some portion of the game with menus”
zarf says, “I sense we have permanently tangented”
Emily says, “conversation most frequently, but also other stuff, as with motivation-picking in The Baron”
Emily says, “but yes”
baf says, “Usually dialogue.”
maga says (to Emily), “few of them take a really CYOA-y approach to overall structure, though”

Planning for next time, and a SpeedIF

Emily says, “okay, I think we are basically done”
Emily says, “next time: CYOA structures, May 10, same time; initial info at https://emshort.wordpress.com/if-discussion-club”
eu exclaims, “Okay. Thanks everyone for the interesting discussion!”
Emily says, “indeed, thanks for coming!”
zarf says, “Let me interject”
FloatingInfo says, “Thanks and your welcome.”
zarf says, “I would like to have these monthly chats alternate with periods of experiemtantion”
icouldntcomeupwithaname says, “(what I meant was the structure of a CYOA, but the interaction and parser of a IF.)”
Busta exclaims, “thanks guys!”
zarf says, “Therefore, I declare some speedIF”
baf says, “Ooh”
zarf says, “mess with these ideas we’ve been talking about for two weeks”
eu says (to zarf), “Good idea.”
Kelden exclaims, “thanks for the interesting read!”
AdamM says, “Thanks for organising this again, Emily.”
zarf says, “post a link in the intfiction.org forum post about this chat”
zarf says, “and we can (on the forum) discuss in two weeks what we’ve come up with”
maga says (to zarf), “tempting, but good gravy am I overcommitted already”
Emily says (to maga), “maybe we get extra points if we manage to make our shufflecomp games also on a different time scale”
maga says (to Emily), “I endorse this”
zarf says, “yes, as nm says, #practiceclub”
DavidW says, “Thank you, Emily.”
zarf says, “I will do something, probably a tiny little experiment, but hey”
zarf says, “that is all. Thanks all”