i think we can probably speak for both sides, the automated and the authored, to say that there simply aren’t a lot of really great ways to put 1,000 content atoms in a visual representation. you can follow one of these algorithms – they’re effectively still the optimal strategies for this http://cs.brown.edu/people/rtamassi/gdhandbook/chapters/trees.pdf – or you can make a twine. but either, at a large enough scale, results in a lot of small moments of your game dev spent looking for your own content. that may or may not be better than a spreadsheet for some people, but it’s certainly not easy work. in fact, as far as i know, this is a type of thinking which people struggle deeply with. that’s one of the big problems in interactive TV, is the people who sign checks don’t really know how to understand a spec script that’s also a mathematical graph.
How would such a split even work, considering how many hybrid games there are?
Was Skybreak! a parser game or a CYOA?
I’m not even sure how a filter would work in that case, although I wouldn’t object to the idea if it’s author defined. But then I’m also not sure how the OP has gotten so many replies when he admitted from the beginning he was talking out of his ass. How can someone who never plays a CYOA know how much work went into the individual games? I’ve read plenty with a word count of over half a million, and use of variables and state tracking are common. Not sure what about that makes it ‘easy’ or not count as fiction that is interactive.
It’s of course already been pointed out that parser games, and a specific kind, always win. And the OP and anyone who cares have complete freedom to ignore any game for any reason they like. This is a complete non issue.
There are certain types of games I prefer not to play too, but I don’t make threads demanding changes to a long running comp based on my personal preference.
edit: lol, I see a wild The0didactus appeared while I was typing and brought up Skybreak already. That came has really become my favorite one to confound people who see a significant difference in parser and CYOA though.
Skybreak! definitely blurs the line between parser and CYOA, and I’m working on a game now that will utterly shatter it.
In all seriousness, I favor a filter for interface, but maybe the categories should be mouse, keyboard, and touch. That’s actually useful information and avoids the politics of choice vs. parser.
to me, speaking from the user experience perspective, the generic option here is probably best. rather than “judges should be able to filter games by X,” perhaps the answer is “judges should be able to filter or organize their ballot according to their druthers.”
and i think this is probably a better answer for why this is the most commented-on thread so far – clearly, regardless of reasoning, there is significant desire for the ability to tend one’s own garden here.
Is it an aesthetic distinction? What do you see as the source of the inbetween-ness?
I see a parser game as usually consisting of a world model and a parser. Whatever the screen looks like, or whether I can click things too, or there are pictures, etc., that’s still a parser game to me.
The world model is tied pretty significantly. It’s not that a CYOA can’t have one, but there’s a reason they lean simple if they’re there, and aren’t typical. The combinatorial explosion. All elements have to be presented as discrete things, in lists or links, to be chosen. That’s why it’s best if there aren’t a lot. A human is great at pulling any one element out with a word. In the parser approach, the word no longer has to be on the screen, it only has to have been encountered, or remembered, or potentially logical (or potentially illogical) and be typed. To say, ‘Okay, the English language (or some other language) is your basic command/item list’ eliminates the visually crippling graphs/lists that form a challenge for both choice-based design in general, and presentation if seeking to use a highly combinatorial world model. The parser game leverages the mind’s traffic with language itself.
Most discussions I’ve had with myself in reviews about parser-ish’games have been at cosmetic levels, I think, when underneath I was usually dealing with what I’d unambiguously call a parser game. I don’t know that I’ve seen much of the reverse because it’s usually apparent if a game is using a world model with many basic planes or not.
-Wade
I spent a happy hour or so the other day going through all the entry listings, making notes of which ones I wanted to try soon. It’s one of my favourite things of IFComp too. That left me with 20 games for starters. When I’ve exhausted those it will be a case of going back through the list at length soon. Not everyone enjoys doing this process.
I’m in favour of more filtering options because the number of entries is so extremely high that as a judge I find it overwhelming to pick through. I know many are in favour of judging in a purely random order, but that doesn’t suit me personally. I prefer to pick things that appeal to me, especially if I am then going to be spending an hour or two on a game.
However I am hugely sensitive to the problems of filtering. I do like the idea of author defined terms though. But again with over 100 entries this year I do think some filtering would be a nice option. Again I stress an option. People who don’t want to use it wouldn’t have to.
I think when people use the phrase “parser-like” they’re pretty much always referring to the world model. It’s to contradict the automatic assumption that hypertext games must be CYOAs. That’s not the case at all, just like an Inform game isn’t considered a CYOA just because it uses a conversation interface instead of requiring you to type “ask about” for every item or idea in the game.
Yeah. Perhaps a simpler definition is “typed commands” vs “choices” although there can be hybrids. A more technical distinction is “can the player create commands the author perhaps did not intend” which is one of the difficulties in troubleshooting and beta-testing parser.
Choice narratives can certainly have a world model and locations and inventory, it’s just up to the author to create it from the ground up, whereas parser engines have an existing world-model for discrete locations and objects and inventory already accounted for.
The thing that made it click for me, originally coming from the PARSER OR NOTHING mindset, is using the Storynexus engine which followed the deck/card-drawing model of Fallen London and mechanics were possible, I just had to think more in terms of doing it like a board game.
Funny that. Seems everyone that wants to see a bias will see it regardless of what system they’re using to make games.
There was a comment on a choice based IF forum relatively recently where the poster said that they were under the impression IFComp discriminated against games that weren’t parser when it was brought up that people should consider entering. They were corrected that this was a thing of the past… and then posts like this one pop up. I’m glad to see the view that choice based games are less worthwhile than parser isn’t shared by the majority, but I’m not sure why it persists since the IFComp is almost always won by parser games. If anything choice games are the ones being disadvantaged here from a selection of players that automatically judge them inferior to parser, sometimes without even playing them first.
I also disagree that it’s very easy to make a choice game compared to a parser one. The difficulty all depends on what is done with it, and regardless of how easy or hard it may be, when marked it should come down to how well a game is made and works with the medium that has been chosen for it. A well made game will shine given the chance, regardless of what system is used to program it.
You can simulate parser to a degree on some types of choice-based programing. (Even choicescript which allows typed commands and variable tracking if you feel like programming it that way. That doesn’t play to its strengths, but you can use it that way.) If you start dividing it into parser and not, it could cause issues for categorising the parser mimics.
As much as I love parser games and text adventures, I actually like point-and-click adventures just a hair more. Maybe it’s because I prefer to struggle with “pixel hunting” than “guess the verb”. Who knows. So it’s probably more accurate to say the type of game I like to make is a text version of a point-and-click game, which looks similar to a parser game because of its roots. But classification aside, it’s a game in text with a story and objects to manipulate (and I’m adding a few puzzles to it too).
I guess if I were to go on a tangent about my thoughts on the subject, I think that the true difference about games with a parser and games without is who is in control. The parser makes the player feel like they’re in control, but hypertext makes the player feel like the author is in control. But the truth is that the author is always in control in both formats. It’s no coincidence that any time story is doled out to the player, the parser game steals the player’s control completely, sometimes even ignoring or overriding what they typed. It’s all the same in the end. Only the interface has changed.
Anyway, that’s my thought.
And that’s the fundamental lesson I’ve learned over years of doing this: The author is in complete control, and the best situation is when you can make the player feel like they have control. The term for this is agency and the best games are “high player-agency”.
Along with that, the other secret is the author needs to be on the player’s team - if you’re from square one fighting against the player to SOLVE MY DIABOLICAL RIDDLES AND GUESS WHAT I WANT YOU TO DO muahahah then the player will become frustrated more quickly and quit and not even encounter everything you planned later that makes it all worthwhile. The more you can make the player feel intelligent and like they’ve “outsmarted” you and are making progress, the more they are likely to hang in and see the whole game.
This happened to me with Arthur DiBianca’s Sage Sanctum Scramble - I had no intention of playing much past the first few screens to see what it was about, but the first few puzzles are easy and it’s like potato chips where you can’t stop putting your hand in the bag. I got about 15 puzzles in before they got hard and I had to skip. Now I can’t stop thinking about it and want to go back. If his first three puzzles were as hard as number fifteen, he would have lost me as a player.
The idea that the player has more freedom in parser games has always been overstated. The only real freedom comes in being able to read the author determined story elements in a slightly different order and pace.
Since we’ve reached that point in the discussion where we’re giving our personal hot takes on the difference, to me it’s always been the fact that standard IF is traditionally very linear. There’s usually only one plot to discover and the interactive elements come in puzzling out exactly what you must type to be drip fed the next bit of the story. CYOAs meanwhile tend to eschew the microscopic details to offer more breadth. The player experience is less about precisely steering their character through making a pot of tea, but navigating them through dozens of alternate lives.
But words supplied by the author are still the medium this happens in and their effectiveness is main thing being judged in this comp or any other.
this breadth is also, to be fair, one of the things that made it tough for twines to find broad appeal with video game players. one of the most consistent critiques i saw, and something i kept in mind when designing an IF system, is the “analysis paralysis” induced by the breadth of routes from any particular page. some work can be done to diminish this labor by highlighting the “navigational” or “important” choices, like “with those we love alive.” in general, though, video games as a broader field seems to prefer having a clear understanding of where they’re invoking agency, and where they’re just getting lore.
My preference in puzzles doesn’t come from odd situations and how to overcome them like MacGyver. I don’t care about trying to unstick the gear in the clock by using baby oil. I care about the fact that I can now go confront Tom with new information since Sarah’s statement blows his alibi to bits. Characters are infinitely more interesting than inanimate objects. This is just my opinion. But it’s also the reason I personally find it difficult to care about puzzles in text adventures. If anything, I’m only adding puzzles to my game because I feel a bit obligated and that maybe it’ll get parser players to enjoy it more.
And that’s the beauty of having choice narratives and parser narrative - each has a strength they can cater to. Parser is good for object-based manipulation and exploration (The Impossible Bottle is a great example with hundreds of moving parts and interlocking systems), Choice is often better for story and conversation-focused plots where it doesn’t matter as much whether you’re in the Church Courtyard or a Coffee House and the author is more interested in just telling a story rather than building a universe. Conversation can actually be a weakness in parser since it’s not inherently built-in and authors usually fall into ask/tell verb schemes or install a menu-plugin if they want to offer dialogue options. Hybrids can be any mix of the two.
In my experience it’s all about seating the player in the character and offering choices that speak to tension of the scene. I’ve had issues with a few choice games this year that just give you a grab bag of choices at each juncture that seem to have been decided just to be as diverse as possible. They haven’t told me who the character is yet and honestly I have no stake in where they go.
Someone asked me recently if coming up with choices is hard because I have to imagine everything the player might want to do. I don’t really care what any random player might want to do. I care about what my games character is experiencing and what their available choices say about who they are and how they speak to their current struggle. Saying: You’re at a dinner party, you can do anything! What do you want to do? That’s boring to me. Saying you’re at a dinner party with the family you were astranged from since yesterday. You’re here because you’ve felt guilty for leaving for years. How do you deal with the uncle who’s acting like there isn’t a giant elephant in the room? That’s a lot more interesting and if you do your job setting the scene and character right so the player is in their shoes, they tend to have an immediate visceral reaction of how they’d respond, and only have choice paralysis in moments when the character is feeling the same thing.
this is pretty close to how i’ve felt about it! but i’ve also found that, both because of the history of CRPGs, and the history of commercial mostly-IF, some more “mainstream” players also expect that there’s little to no dissonance between “player” and “protagonist.” a lot of the most interesting narrative difficulty in IF comes from the opposite of that – producing strange and compelling intersections between who the conceptual player is, who the protagonist is, and who the narrator is. those don’t necessarily have to be explicit, but i find they’re fairly unique complications to IF writing as opposed to broad games writing. part of that could well be losing access to the “cinematic” gaze provided by 2D and 3D graphics, which i think calls on a lot of visual-artistic history that IF doesn’t tend to evoke.
i think this was also what people found so strange and compelling about facade – you start knowing nothing about yourself other than that you’re about to go to dinner, and all the boundaries are formed through testing the limits of the NPCs. it would be insufficient to suggest that either player and protagonist are separate, but they’re also very clearly not the same. the protagonist is imbued with far more history than the player is given access to, and hardly only in an in medias res sense, given there’s a huge amount that simply must be imagined.
similarly, for a totally non-graphical parser game, spider and web gains nearly all its puzzling, and its narrative, from a protagonist-narrator who fully intends to lie to both their listeners. that would be extremely difficult in visual terms, but textually, and mechanically, it works!
This is something I’ve been struggling with in my own game. I tried to get around the faceless player persona by introducing a sister character that walked around with you and said all of the stuff that I was afraid would distance the PC from the player if they said it, but then I realized that once she was gone, the PC was just an emotionless robot. Which is exactly what 99% of parser game PCs are, I guess, but it just feels wrong in this environment. So I need to think it over some.
Agreed completely here, well said. Although this is touching on another schism in the approach to playing/writing IF that I’ve seen lead to a lot of arguing in the past, and to me is more significant that the authoring system or world model. Somehow it’s become a not uncommon idea that every character should 100% represent the player…somehow…even if they’re a magical assassin in a fantasy world with deep lore. Available love interests need to represent their exact preference, and the character has to have their own gender and beliefs or else it’s not “relatable”. So from the get go the author is dealing with limitations, having to offer instead of interesting and meaningful choices and character nuance, a billion useless cosmetic options. They always affect nothing in the plot and for me, force this generic plastic feeling to the simulation, but are vital to appeasing these more casual readers.