'The key in making strategy more accessible is...'

“The key in making strategy more accessible is to make it something the human brain has a lot easier time parsing.”

I think this Gamasutra article from a couple of months ago has a lot of applicability to IF, and not just in RPG terms, but in terms of simplifying things like conversation trees, examining everything in a room, etc…

I always have a little eye-roll at sentences that turn the brain into a little homunculus, but it makes a good argument for stripping away conventions and trying to get a more satisfying core experience.

What more accessible features in IF would you like to see more of?

One of the interesting parallels is the article’s discussion of the naturally more intuitive analog positioning versus grid-based play, when compared to the way we divide up rooms. And that’s in addition to the more directly relevant statements about avoiding having to go through a lot of menus. I also find interesting the use of ‘parsing’ not to refer to what the computer does but the human mind (I.e. maybe there has been too much preoccupation with the handling of the wrong ‘parser’). I could go on and on but haven’t the time ATM.

Rooms can be used to represent abstract conditions, or vast areas of space, or a state of travelling rapidly through space, without modifying traditional text-adventure mechanics. Maybe that’s a loose analogue to what the company did by abolishing the grid in their turn-based strategy game.

But IF is experienced as a flow of text… maybe a closer analogue would be making real-time IF instead of turn-based. Or somehow subverting the concept of a “turn.”

I’d be interested to hear what you think about that, when you have time. Letting the player do the parsing sounds like pen-and-paper RPG mechanics, to me.

There’s a lot of interesting stuff there but a lot that ISTM would be hard to adapt to IF. (Which isn’t to say it shouldn’t be tried.) And translating some of it to IF seems like it’d require a different style of game than the standard puzzles-with-your-inventory model. (Which definitely isn’t to say that it shouldn’t be tried.)

Ditching rooms (the equivalent of the grid) in favor of a more fluid approach to space would be really cool, I think. Part of the smoother-flowing game experience this yields just comes from not having to type n.n.w.d.se every time you want to try something in a different room, which is maybe the rough equivalent of not having to troop through menus. Pacian’s latest game, Castle of the Red Prince, does that even though the individual locations are still rooms. (And there’s a time or two when you basically have to exploit this system to get around a timed challenge, which I didn’t like.) And big-map games increasingly have a goto command that solves some of these problems. The Stately Gardens example in the documentation has a truly fluid spatial model, though I don’t know if anyone’s done anything like that in a working game, and you’d probably want to vary the text even more than it does.

“bring[ing] the details of the battle to the player’s attention without the need for a long, complicated list of details”: This seems like it’d be hard as hell to do in text. A lot of the cues they describe seem like little eyeball kicks that depend on the non-linear way we process graphics as opposed to words; having five soldiers draw their swords as you come in range can communicate things efficiently if you can see it done, but if you read “Soldier 2 draws his sword! Soldier 5 draws his sword! Soldier 6 draws his sword! Soldier 8 draws his sword!” it’d be stultifying. And unilluminating. You could maybe translate this into some system that communicates things about the player’s capabilities or what-have-you through subtle changes in the text output, but it’d have to be very well done.

Making strategy more accessible and intuitive: That seems like the big cheese. Something that I often see people talk about is how they like to know how their actions are affecting the NPC reactions around them, which often means something like being able to check the stats or state changes that their actions cause. (This may be a bit of a straw characterization.) But ideally you wouldn’t have to do anything so crude – ideally the NPCs’ actions would just make sense in light of what you’ve done. Going gridless in a strategy game allows for a more intuitive gameplay that lets us use our spatial processing abilities, but going gridless in IF to let us use our – what? emotional intelligence? – seems like it’d require something that’s a lot harder to model. As the authors say, they’re about replacing puzzles with strategy; replacing puzzles in IF with a more natural interaction seems like it’d be hard. And you have to want to replace puzzles with something.

Worth doing, though. Of all the games I’ve played Galatea seems like the one that best matches the ideal – Galatea’s posture gives feedback in a natural way, and I never really felt tempted to expose her stats and see how what I’d just done affected them (well, except when I was trying to grind the stats for certain endings). I also had that “The NPC is reacting like a human being” feeling when I played Best of Three. I guess I should look for Versu’s stuff as soon as it comes to a platform I have.

Not necessarily. Time in conventional turn-based IF is space. Therefore the way to compress or undivide time is to compress or undivide textual space. Notice how much space it wastes to move people from room to room. What if we just ditched all of that? Give people a more intuitive way of determining where they are, if not influencing where they are. I am not really saying don’t use compass directions – the article more reminded me of subversion of the way rooms / player location is interpreted and printed. It isn’t just used to determine your position behind the scenes, it’s used to print a whole rigamarole at you of bolded room title, typically fat descriptive paragraph, followed by contents listing. It is these things that I think of as the ‘grid lines’ in IF, not necessarily the compass direction.

To return to the graphic analogy, you don’t have to complicate the eight-point joystick in order to achieve analog positioning: you just have to feedback the player’s movements to them in a less obviously grid-like way. The lowest level of this is just – don’t draw the gridlines. The analogy being, don’t title and describe the rooms so ostentatiously as ‘separate boxes of existence’. A higher level would be to have the player’s compass directions or other movement commands not actually be universally interpreted as a way to switch from one ‘room’ to another but rather as a way to move around within spaces – getting marginally nearer to or farther from something already mentioned, rather than squirting the player into an entirely different, discrete ‘box’ with all different referents to manipulate than the box before. This is an almost universal paradigm which, if you think about it, is entirely unnecessary even for standard puzzle-and-inventory IF. It exists merely by convention, and I am not against it or anything, I just wonder about all the paths not taken, and articles like this help my mind trip out there and think about what could happen.

In this case I didn’t mean anything so radical. It’s just an important reminder, I think, that the player is parsing too, all the time, and a lot of problems with communication between the game and the player can be solved by giving the player better (i.e. easier) information to parse, rather than trying to make the computer so super-intelligent it can understand dialogue, etc.

This is related to the ‘Just tell the player the damn commands’ principle. Many many great games can be written with very limited sets of commands that can be recombined in interesting ways – and yet most puzzles in most IF games are just another variation on hide-the-key-description-in-an-obscure-place, leading to an Easter Egg hunt that often devolves into guess-the-right-noun and can interact unpleasantly with the parser which players already feel is a guess-the-verb challenge.

But you can design an entire elaborate puzzle only with the commands N,S,E,W. You can design great puzzles with very limited sets of commands, as long as you allow the player to perceive that underlying puzzle structure – make the field of strategy more available to the player, make it easier for THEM to parse what are their options and how can they be combined, and you won’t have to catch so many corner cases caused by players flailing about with no clue as to what kind of a game they are a playing yet.

Just a fancy way of saying, ‘give them all the help they need, don’t make them guess at the tools, only at the solutions’. But I like the idea of ‘making it easier for the player to parse’ because it emphasises concision, whereas ‘help the player’ could end up with reams of boring tutorials they will never get through, and misses the point. PLaying the game should not require a lot of reading either – it should just be easy – it’s solving it that should be hard.

TL;DR – Much fewer commands, tipped much more obviously to the player, combined to make deeper puzzles that don’t depend purely on exhaustive exploration. I have thought of this before but the article really made me consider it from another angle and kind of verified for me that there are important design principles that are the reasons why I lean this way. It’s not just my intuition. 8)

I don’t think it’s as hard as it seems. It’s just not a one-to-one correspondence so there is a layer of creative interpretation there as a designer.

Mmm-hmm, interesting, I did not know about Pacian’s new game. In places, Rogue of the Multiverse came pretty close to the kind of flow I want, without getting rid of the grid, so I’d be interested to see what Pacian does when he actually ditches the obvious ‘room grid’. A better example might Hoist Sail for the Heliopause and Home, a nice little game about a style of navigation that largely ditches overt rooms, that plays and reads so damned smoothly once you figure it, that it ended up being a much bigger influence on me than I expected considering its subject matter.

But why?? Text is so simple and economical – it is only the weight of all the room/object/inventory conventions we have inherited that makes this so. It is not hard as hell to do in text, it is just swimming upstream, or cutting against the grain, or whatever, when you do it in TADS or Inform or any other system that swallows the Infocom paradigm whole.

True here, but those soldiers drawing their swords in one line each are moving along a standard RPG-like grid of weapon-equipping. All three of their actions could be combined in one line, if sword-drawing timing is so important to the strategy (doesn’t sound like that interesting strategically but let’s assume it is a key part of the game) – it is only because we have to cut across the grain of ‘every enemy is an object and every sword is a separate object and therefore every equipping of a sword will be handled by a separate routine that will print its own separate line’. Cutting across the grid in IF, to me would mean cutting across all of the object-separation and combining as much strategy-relevant motion into one simple line as possible. ‘All the soldiers draw their sword.’ Or go even further, and combine that actions with the description of the first round of initiative, or whatever.

I feel like the one-to-one correspondence of character-object-to-inventory-object-to-game-action-to-output-paragraph just produces ooodles of redundant unnecessary text and kills the momentum of almost any action scene. Very rarely have I scene an action scene in IF that didn’t feel like the main character is actually a robot and I almost have to move them piston by piston (I’m exaggerating), so painstakingly overconcerned with irrelevant minutiae is the average IF game. I think of this stuff (and not just the room divisions) as the ‘grid’ in IF. When the grid determines the structure of lines and paragraphs, then the grid is very obvious to the player and it kind of sucks. When the grid has a much looser relationship with the reuslting text (allowing it to be much compacted) then things can go much smoother, but how rarely have I ever seen this done. I will try Castle of the Red Prince.


No I don’t believe puzzles need to be replaced. But you are quite right that an alternative to intuitive spatial processing is required. The wonderful thing about text, though, is that we are not tied down to literal spatial processing, as long as we don’t confine the field of play to the standard room conventions of IF. You can place the player in any field, spatial or conceptual, as long as you make it easy for them to parse that field and you can still have puzzles based on the player’s discovering a better or more effective pattern of ‘movement’ through that ‘field’. Based on conspicuous feedback cues, as always, just not necessarily overtly grid-positioned ones. (Before somebody says, ‘It’s actually impossible to escape the grid in IF’ let me stipulate that this is true, but to point that it is also impossible to escape the grid in a graphical game – the way to achieve ‘gridless’ play in graphics is just to make the gridlines so narrow that you don’t notice there’s a grid – i.e. ‘gridlessness’ is actually smoke and mirrors, but it works, and that is the part that I find interesting.)

Yeah Galatea is a fair example but not a good example, I think. The reason it’s fair is that a pure conversation game is essentially roomless, so it does demonstrate how the flow of things sometimes goes better when you dispense with all the room divisions. The reason it’s not a good example is that conversation systems are not easy to parse for the player – particularly not Galateas! Which I feel is part of the point of Galatea: making the successful parsing of the ‘conversation landscape’ into the actual puzzle. But what I was thinking was more of a quite clearly parsible (while possibly still abstract, and not overtly snap-to-grid-ish) landscape, the successful traversing of which may require the solving of puzzles depending on which ‘direction’ you wish to go.

I told you I could go on and on about this. XD


Some time ago I tried to enumerate the different kinds of IF granularity in a number of posts.