What do you think is IF's strong point?

I think it has one, but I want to bounce some ideas off other people. This is wait raised the question to my mind:

I was playing a graphic adventure game just after having a long discussion with my friend about all the various strengths of widely different artistic mediums: movies, theater, music, visual art, things like that; we also discussed what each of these creative mediums seems to revolve around the most (music being structure, visual art being contrast, and literature being continuity; film was a special case). I was thinking about some of this stuff, because he had done most of the realizations, and since I wasn’t talking to him any longer, I wanted a little more.

When I think of the differences between graphic adventure games and IF, not very much comes to mind. They are both primarily puzzle-solving and storytelling experiences (I think). This point may be a little off-topic, but I would like to share that most of the graphic adventures I’ve been playing seem to have better stories than many of the old IF game’s I’ve experienced. Also, I realized now that death in the adventure game medium is very deterring, and doesn’t work well very well as a casual punishment.

Do you suppose graphic adventures and IF have different strengths, despite their similarities? I would like some of your thoughts on much of this. I’m thinking about calling back my friend and giving him this question, though since he hasn’t played any IF (but knows about it), I don’t suppose there is very much from experience that he could say. I hope you don’t find this a tiring question, because I think the main reason I would like to know this mediums strength is so that I can write IF better than before.

The differences between graphical adventures and text adventures are no greater than the differences between movies and novels.

This essay I wrote last year has some relevance: eblong.com/zarf/essays/if-for-writers.html

I’m aware of the reason I never rated point and click adventure games compared to text ones. With point and click, you can roll the mouse over everything and the answer is somewhere there. Any choices, any hotspots, even a menu of words, they’re always there on the screen or in the interface somewhere. It’s impossible for the game to not display the answer in some form as a button or buttons.

In a text game, the answers aren’t necessarily there in the interface. Whether cued by the game a lot, or not at all, you pull the answers out of your mind. The words that solve the problems may never even have appeared in the game. I call it the ‘illusion of promise’ (at least I did once in an email). Whether illusory or not at different times, the effect is real in terms of… it’s the only one that makes me feel like I’m really in there solving problems and doing things myself. In graphic games, I never escape the feeling I’m just pressing buttons that already exist, in the right order, and you can get lazy and brute force them as well, or painfully drag your mouse over every nook and cranny to pull out the hotspot.

This is just one of lots of reasons why I like text adventures, but it’s always the one that’s real important to me in the context of what they turned into with graphics and mice.

There’s really two features of IF that, I think, should be discussed separately: the input method (the parser) and the output method (no graphics).

You can mix and match the two. For example, the early Sierra games had text input, but graphics on the screen. Multiple-choice IF like “Choose Your Own Adventure” uses nothing but text, but only allows the player to click on pre-defined choices.

You can even combine output methods (with an IF game that has a few illustrations) or combine input methods (with a game where some options are clickable but some must be typed).

What are the strengths of avoiding graphics?

  1. Low cost/barrier to entry. Check the credits on professional AAA games; you’ll see dozens and dozens of graphic artists credited. By avoiding graphics and sticking to text, an extremely small team (even one person) can make an entire game in a reasonable time frame.

Lower barriers to entry make it affordable to explore really unique game ideas, to experiment with material that could fail.

The lower cost of text also means that it’s more affordable to rewrite. Throwing away work is always painful, but rewriting results in a more polished product. Movies, for example, can almost never afford to go back and reshoot, but novelists almost always rewrite part of their novel.

  1. Mental interiority. Only in text can you directly describe what someone is thinking or feeling. (I’m not such a big fan of this one; IMO it’s generally better to show the audience how the character is feeling instead of telling the audience directly.)

  2. Smells, tastes, and textures. Only in text can you directly describe what something smells, tastes, or feels like.

  3. Ambiguity. Visual illustrations leave less to the imagination, which gives the audience less opportunity to decide for themselves how things look.

  4. Large blocks of time. Only text can describe large blocks of time. Even movies/television typically put words on the screen (“a few months later”) to depict the passage of time.

  5. (Un)usual behavior. Only text can state clearly that an event is commonplace, which can be helpful in establishing setting. (“Every night the same dream.”) You can hint that something is unusual by showing characters acting surprised, but commonplace items are harder to show without telling. (“As you know, Bob…”)

What are the strengths of parser input?

  1. The feeling of creativity. When you type in a word not included in the text, and it works, you have the feeling of inventing the idea yourself. This is especially good when the word you typed is the solution to a puzzle; it inspires a feeling of owning the puzzle’s solution.

  2. The feeling of freedom. Sometimes, with very good parsers, you can make players feel like they could type anything and the game would understand it. This is certainly not literally true, but it can sometimes feel like it’s true.

Overall, parser input is a particularly good fit for text adventures, games that give the player the freedom to explore a space while solving puzzles to advance the plot. The feeling of freedom makes exploration feel more fun, and the feeling of creativity makes solving puzzles more rewarding.

I think that there are also unique strengths that emerge from the combination of the two – as an offhand example, the sense of play as conversation with the narrator. (You can get this from other media, but it’s much easier to sustain in IF than in any static media I’m aware of.)

Yes, I can see your point, especially about the parsers. As for the comparison of IF and graphic adventures being no different than novels and movies, I guess I never considered that, but it sounds reasonable. Yes, I think these were some really good responses, so thanks!

I find it odd that you’d mention “conversation with the narrator” as a good example of combining the parser with text-only gaming. In most games, IF and otherwise, authors use multiple-choice menus to sustain dialogue, where I say something and then you say something in reply.

When IF games don’t use menus, they typically use a topic-based conversation system (“ask snow white about the queen”) where I don’t feel like I own lines of dialogue; I’m just directing the overall course of a conversation.

IMO, most IF authors agree that menus beat parsers for dialogue. Which is to say: sure, a good chatbot can be fun to play with, but it’s so hard to write a good chatbot (in the context of an adventure game) that nobody does it; when they do, the chatbot becomes the centerpiece of the game.

But maybe I missed your point: it’s not about conversations with NPCs, but conversations with the narrator? I dunno; in Violet, I didn’t feel like I was in conversation with Violet; I was doing stuff in my office while Violet watched and kibitzed.

Maybe a bit in Hitchhiker’s Guide to the Galaxy, when the narrator resisted my commands? But “go south. go south. go south. go south” is not a very good conversation. :slight_smile:

Yeah, I don’t mean conversations with NPCs. What I meant is: the narrator of an IF game speaks in text, and you speak back in text. There is a sustained illusion – one that is largely implicit, and only becomes really obvious in games that mess with it – that you are speaking to the narrator, like the conversation between GM and player in an RPG.

Well, most Socratic dialogues aren’t good conversations either, qua conversation. Mostly the interlocutors just chip in at intervals while Socrates delivers his lecture. But their form is still fundamentally conversational, and this form gives them important qualities that they would not otherwise have.

I don’t think that this is straightforwardly true; there’s substantial disagreement on the subject. I think it really depends on the kind of dialogue. Menus tend to make conversation easy, free-flowing and directed; not all real-life conversations are like this.

In terms of gameplay, I think parser-text-gaming has two distinct features:

(A) A convincing illusion of tactical infinity.
(B) Unknowable boundaries.

The two feed each other. Whether or not they’re “strong points” is a matter of debate, but I vote heartily in the affirmative. Beyond that, it’s just back to that “we stick our graphics where the sun don’t shine” advantage of pure text.

My “home” gaming form (pencil and paper RPGs) has related distinguishing features (tactical infinity and dynamic boundaries) which is why I’m attracted to them in similar ways.

If Inform was less object-orientated and more open to simple string parsing, I would not agree. I used to program in LPC for a MUD, and you could parse everything you wanted, which enabled you to generate deep conversations. “NPC, but what about the possibility to zarf the magas with the gravels?” - check for “possibility”, “zarf”, “magas” and “gravels”, check some flags you set on the way, and voila - deep conversation. Stuff like that, ya know.

Inform has always had enough string-matching ability to do simple keyword matches of that sort. (“if (snippet) includes (topic)”, which came in long before indexed text.)

However, going from keyword-matching to “deep conversations” is, how shall we say, not a trivial problem.

[Off Topic]
That’s what I miss. Player input “NPC, aren’t those gazoongas absolutely phantastic?” >>> system string variable containing everything after "npc, " >>> if(member(systemvariable,“gazoongas”)==1) answer_something_about_gazoongas;
Inform is so unnecessarily complicated about strings, at least viewed from my very limited knowledge of it.
[/Off Topic]

I don’t think it’s about the language, but about what we’ve learned from how it works with actual use. It’s a dead horse by this time - we all know and have seen how those systems make it impossible to achieve actual communication, because we have no idea what part of our input will be recognized and what part won’t; also, we have no idea whether we’re on the right topic but phrasing it wrong, or worse, phrasing it right but in a way the program won’t recognize it.

So, quite aside from the programmer’s difficulties with such a system, is the difficulty of interacting with it - and the result will always, always be a broken experience.

Don’t go that way, Grueslayer, that way madness lies. :wink:

To me that sounds more complicated than ‘Instead of answering NPC that when the topic understood includes “gazoongas”: say “Gazoongas are great.”’

Agreed regarding that feeble example. But I’m not after parsing one single buzz phrase, that works fine under Inform. I’m rather talking about stuff like “answer X if string contains buzz phrase 1 but not buzz phrase 2, and buzz phrase 3 if variable Y is 138” - stuff you need if you want deeper conversations.

I had an idea for how to do this: You’re in, say, rural Quebec, you barely speak French, and the NPCs don’t speak English at all. So if you type in “Under no circumstances should you hand me the bucket” the response is something like, “You wave your hands around and say ‘Under no circumstances should vous hand moi le seau.’ The concierge, delightedly recognizing the word ‘seau,’ hands you the bucket.” The experience is broken in all the ways you mention – you don’t know what’ll be recognized, you don’t know whether you’re coming up with the right word (though a kind author would program in lots of English synonyms, and stay far far away from “gazoongas” for several reasons), even if you do come up with a word that’s recognized the parser is likely to rip it out of context – but this is mimetic, because the PC’s experience is broken in exactly the same way.

I have another similar idea involving a rock concert that’s so loud that the NPCs can’t hear more than every third word you’re saying. The point is that you need some sort of justification for why the parser is ripping individual words out of context. (I figure the player would quickly catch on that it’s just parsing keywords, but hopefully it’d be fun.)

I find that first idea delicious. Go and make that game forthwith, I want to play it. :wink:

Me too!

Ha, awesome. That sounds like fun.

You are talking about this sort of construct:

After reading a command:
	let pc be the player's command; [ just an abbreviation, to make the following line easier to read ]
	if pc includes "fizz" and pc includes "buzz" and pc does not include "frog" and YVar is 138:
		now YVar is 236;
		instead say "Hello.";

You could build your whole game that way.