I’ve played parser games, and I’ve played hypertext games. I don’t know much about games which try to bring the verb noun experience out of the command-line interface, but I’d like to. I’ve seen the VIBAE demo, which is the kind of thing I’d like to know about more broadly. What are the games to experience in this area?
I know I’m not the only one curious about this so pointers to previous discussions are also appreciated.
(I hope I have this in the right place.)
I don’t know how close it is, but Quest engine games can be set up so that the items in scope have little cards on one side of your screen, and the relevant verbs are listed on each card.
Something else I ponder a lot is an engine idea where you put a command together using clickable keywords in a list or collage. The ambiguity of what verbs work where is preserved but you have less typing now.
Not sure which direction you meant, if any. Kinda surprised that I haven’t seen an engine like the latter before, honestly.
This was a similar question whose answers may be of some interest. List of parser-ish touchscreen games
Thanks @Zed, I try to follow the discussion here but I didn’t see that one. Lots of stuff there to chew on.
@inventor200 I think it’s a UX thing. Easy to implement, hard to make it feel good. I’m looking at Detectiveland (will play it more fully) and I feel a little bombarded with windows and buttons. Not trying to judge the game based on five seconds of experience. Some games thrive on complexity. A blinking cursor communicates a kind of elegance though.
That sounds like the Monkey Island games to me. I believe the text based Magnetic Scrolls games also had a feature like this, but I’ve never played any of those.
Huh… I was thinking of the user side, but you’re probably right.
HOW DID I FORGET THERE WERE WHOLE GAMES LIKE THESE??
I’m not even in my thirties yet; it’s not like it was too long ago for me!
Because interactive fiction is a ridiculously huge space with a lot to remember?
Yeah, the two authors I would recommend here are Robin Johnson, for his *land/Gruescript games, and Agnieszka Trzaska – I actually haven’t played Lux, but Chuk and the Arena and The Bones of Rosalinda fit the bill and are really great. Then I haven’t fully played the Open Sorcery games (Open Sorcery and Sea++), but an excerpt from the latter was entered into IFComp as A Murder In Fairlyand and that piece at least is very very good indeed, with some lovely bureaucracy-based puzzles.
I’ve said this before, but I think the “parsery choice” movement is kind of rediscovering the graphic adventure – except with more focus on the writing, duh, and generally better design chops. Which is great!
These more “parsery” Twine games are pretty cool. I’ve been fooling with Lux and 16 Ways to Kill a Vampire at McDonalds. Although for me, they don’t quite make objects or actions feel like first class citizens, as programmers say.
Legend games such as Eric The Unready also had lists and lists of verbs and nouns and you could assemble a sentence by clicking.
I think the SCUMM interface is probably the closest to perfect version of this. Your nouns came from clicking items in the graphical window, so it actually had a concept of scope that was very cool. Whatever the mouse cursor was over would display in the sentence-construction line.
In Maniac Mansion you had three characters to control, and they would automatically move to be able to touch any interactable nouns.
The cool things this enabled was “multi-character” puzzles where one had to hold a switch to keep a door open for another character to enter, or one could ring the doorbell as a diversion to get an enemy to answer it so another person could search their private room in real-time. You could also pixel hunt a dark room to “feel around” for a light switch or a wall to check for hidden loose bricks or secret panels in a way that was natural and intuitive.
What I especially like about the Legend games (like Gateway etc.) is that they retained the command line and allowed free-form typed input, and the GUI parts were only optional.
So they remain real parser games (not parsery non-parser games) with complete freedom, but also eliminate guess-the-verb and the sorts of mistakes and misunderstandings where people type “WHERE AM I” or “I WANT TO OPEN THE DOOR, PLEASE” or “LOOK CAREFULLY AROUND THE CORNER”.
PICK UP THE SWORD WITH CONVICTION. 
I never played that one, although I did play Day of the Tentacle.
Most of the point and click games seemed more shallow to me and although you didn’t have to “guess the verb” you did often have to “hunt the pixel” instead.
It will be interesting to see how Roberta Williams handles commands in her upcoming Colossal Cave 3D game.
The SCUMM games are some of the least shallow graphic adventures around. In fact, the modern SCUMM successor Thimbleweed Park is an example of how much detail they would have included if they’d had more memory space and resources back then.
I’d venture the reason it worked in Lucasarts games is because 99% of the time the images were straightforward so nothing was extraneous nor stupidly hidden unless it was an actual loose brick that you wouldn’t notice unless you touched/examined it. They also were really big on not letting the player get stuck. Maniac Mansion had a couple of permadeaths but only through gross misadventure (filling a pool when another character was idling in it, turning off the nuclear reactor coolant, Ed discovering you doing horrible things to his pet hamster) but you had three characters and could win the game with two - possibly one depending how far along you were.
Maybe not “shallow” but maybe a game isn’t quite “parsery” until it has a critical mass of verbs in scope? 30 plus? I count TADS at around 80 and Inform at 60, although I’d think a lot of classics come in lower. Scumm games being around 15 and some point-and-click games down in the single digits.
Most BASIC games had far fewer than that, and the Scott Adams format had a hard cap that was very low (though I can’t find where that limit is documented right now). I don’t think you can put a hard cutoff on that without excluding many of the earliest ancestors of the genre.
I thought that might be the case, but maybe parser games became more parsery over time.
I counted 21 in an Adventureland walkthrough FWIW. There’s also a difference between telling the player they can use these specific twenty verbs, versus an unspecified vocabulary. The vocabulary in play is larger than the program’s actual vocabulary, if the player doesn’t know which specific words the game knows.
Somewhere between five and five thousand there has to be a point where being given the verb list doesn’t change the sense of open-endedness. I think the big systems are near that point, because they’re complete enough that a player could feel unfairly treated by a game which secretly introduces unobvious verbs. That is, a player could feel that with 80 verbs or so, knowing the list ahead of time doesn’t prevent the full experience.
I’m curious how many viable verbs get used in a standard parser game. SCUMM had most of them including LOOK AT which serves as EXAMINE and USE (something) [WITH] which conglomerated a bunch of others like PUSH, PULL, PUT SOMETHING ON SOMETHING, PUT SOMETHING INTO…
To be fair I don’t believe I’ve ever used 30 verbs in a parser game (disregarding synonyms) unless it had a very specific system of custom verbs, say for magic or dialogue.
I’m curious how many verbs a typical game uses as well. It’s pretty easy to count up the verbs used in a walkthrough (Zork 1 clocks 30 I think), although you’d miss any nonessential verbs that way, or verbs which are required for figuring out a puzzle even if the solution doesn’t need them. But like I said, a word bank of 30 verbs is different from a word bank of 30 verbs plus the chance any other verb may appear.