Announcing the "Text Adventure Literacy Jam" - starts Feb 25th

Looks like a fun jam! :slight_smile:

A few questions:

Rule #3: Are we required to reject any input that isn’t two words? For instance as long as “TALK TROLL” is accepted, may we also accept “TALK TO THE UGLY TROLL”.

Rules #6,7,8: Does the sentence count include tutorial text?

Rule #9: Are IN and OUT ok?

Rule #22: I’m not sure I understand what a “complex” container is, from the examples (get thingamabob from box, put banana in bowl) it seems like any container would be banned (they would already be hamstrung by the VERB NOUN parser constraint anyway)

1 Like

Rule 3 -. It’s a hard two word limit. The reason for this is to keep it super simple. If every time the player types more than two words it tells the player, the game only accepts (a maximum of) two words, then the player knows the constraints and isn’t trying extra words to see if something will work. The moment you break the rule and allow three words, kids are going to try extra words “just in case”.

6,7,8 - Tutorial text should be a maximum of two sentences, and isn’t included in the response text or location text limits. I’ll clarify this.

9 - Yes, IN / OUT / ENTER / EXIT are OK.

22 - Containers are allowed, just not complex container implementations. e.g. EXAMINE BOX (creates some objects that are contained inside the box - fine). Being able to put things in and out of a box is not allowed. Again, I’ll clarify.

1 Like

Just to be absolutely sure: You wrote that Excalibur could be used as a reference. In Excalibur, I noticed that when you include the optional word “the” then three words are allowed, e.g. GET THE LADDER.

So should the parser be stricter than Excalibur?

I am asking since in ADRIFT, the word “it” (e.g. GET IT) will only work if the optional word “the” is allowed to be the third word, i.e. 3 words are only allowed when the word “the” is included. This is due to something hardcoded in ADRIFT that I cannot change.

2 Likes

Good spot.

I’ll update Excalibur/Adventuron shortly to remove that loophole, but I understand that a lot of engines can’t be so specific in how their parser handles these things therefore I have updated rule 3 to allow connective words if the engine requires it, and instead required that if it is possible to detect a second noun, that the game reports to the player that it is unnecessary.

3 Likes

On the subect of containers, I don’t see why complex containers can’t be used, so long as all interactions with the containers can be accomplished with two words. For example, EXAMINE BOX can tell you the contents of the box. Let’s assume one of the items is a comb. You can GET COMB using two words and you can PUT COMB or INSERT COMB using two words (and the box is implied), whereas DROP COMB does a normal drop. Would that be acceptable?

1 Like

So with the new rule, would a game with the following two transcripts be allowed?


You are in a purple room. There is a brass key here. An exit leads to the north.
> GET KEY
You pick up the brass key.
> N
You are in a taupe room. There is a chest here. An exit leads to the south.
> UNLOCK CHEST
(with the brass key)
You hear a *click* and the chest’s lid pops open, inside is a Fabergé egg!
> GET EGG
You take the Fabergé egg from the chest


You are in a purple room. There is a brass key here. An exit leads to the north.
> PICK UP THE BRASS KEY
[By the way, you only need to use VERB NOUN commands in this game!]
You pick up the brass key.
> WALK TO THE NORTH
[No really, two word commands suffice to win this game.]
You are in a taupe room. There is a chest here. An exit leads to the south.
> UNLOCK THE CHEST WITH THE KEY
[Note: No commands in this game require explicit indirect objects]
You hear a *click* and the chest’s lid pops open, inside is a Fabergé egg!
> GET EGG FROM CHEST
You take the Fabergé egg from the chest

I don’t mind adding messages letting the player know that two word commands are enough, but I’d find writing code to make my parser less user friendly rather distasteful.

Another question, from the additional guidance section:

All nouns should have a custom response to an EXAMINE noun (action or custom text).

By noun can we understand objects you can interact with or do you really mean every noun that appears in the game? If it’s the latter we’ll have to be very careful with our descriptions lest every game turn into Lime Ergot. :smiley:

2 Likes

Regarding the two-word input, this is very hard to achieve in most authoring systems, as they normally allow multi-word input by default and their standard libraries are designed to cater for this. I use Inform 6 and I would basically have to rewrite all the standard grammar and action routines to restrict it to two-word input.

I think I know the rationalé for this restriction (make it easier for kids), but it does not encourage good English. For example, why should I force a little kid to say TALK TROLL, when the normal English would be TALK TO TROLL? Why should I force him/her to play guess-the-verb for a simple verb like GET, when the more natural expectation from a child’s point of view may be PICK UP.

Similarly, it means that my game can’t have something like a brass key and a silver key or I get into disambiguation hell.

>DROP KEY
Which do you mean, the brass key or the silver key?

>DROP BRASS KEY
You can only use two words.

>DROP KEY
Which do you mean, the brass key or the silver key?

At this point, the little kid runs for the hills, never to play another text adventure in his/her life.

I really think it should be restricted to a verb phrase and an optional noun phrase for the direct object, but no second noun phrase for an indirect object. In your example, you can use UNLOCK THE CHEST (and the brass key is implied) and GET EGG (and from the chest is implied).

Regarding every noun having a custom response, I wondered about that myself. I normally do this as a matter of course, so it’s no big deal for me, but I’d be interested to hear Chris’ rationalé. I think it’s so that kids can examine anything without getting a response that says, ‘You don’t see any wall.’, when there’s clearly a wall mentioned in the description. However, I can’t see anything wrong with saying, ‘You see nothing special.’

3 Likes

Why try to adjust the rules to yourself rather than adjusting yourself to the rules? That’s most probably the whole point of this Jam.

Simply because it puts everyone at a disadvantage unless they’re using Adventuron. See what Chris thinks when he wakes up.

This is an opportunity to use Adventuron, The Quill, Paws, DAAD and other tools compatible with these rules!
It would be nice if the games could be played but also designed by beginners (whether they are children or not).
Now I keep quiet because I do not plan to join this jam.

1 Like

The point is not to be grammatically correct. The point is to give s simple rule for the player so they know the scope of how they can grapple with the parser. The two word limit makes things less grammatically correct but it makes things actually easier for the player. In Excalibur for example there is only one key, no problem with key adjective.

Connective words are allowed via the jam rules but I do think in many ways this is a mistake as it breaks s simple.rule (maximum of two word inputs). Being able to explain something easily to a child is the highest priority. The rule anyways is relaxed for those that wish to enter the competition with systems that do not support the rule. The rule isn’t there to put other systems (other than adventuron) at disadvantage, it’s there to be able to be able to be an easy to explain the rule and increase the pace of the game. Anyways, connective words are allowed but should be discarded.

Aureas already made a very playable game within these rules and there are many playable verb noun games in existence.

The real frustration with parade games is not a game that expects bad grammar. The real frustration is not knowing what the game understands.

About the complex transcript, all the two word commands are permitted, even with containers. When you open a container its contents should become the contents of the room and can no longer be placed back in the container.

Using adjectives with nouns should not be required, adjust the design accordingly. Try not to have too many keys coexisting. The rules do allow connective words (that can be discarded) but the parser should not understand adjectives for the purpose of disambiguation. The design itself should not require disambigustion. Authors have control over the design, but these are the constraints.

I know it’s not easy, and making the parser less friendly in some ways counter intuitive to the goals, but the jam.ks opinionated that a rule known to the player is more important than parser flexibility for beginners. If that assumption is incorrect we will discover.

Chris

2 Likes

I’m entirely sure that an 8 year old will indeed type TALK TO THE TROLL and GET THE APPLE first off, rather than TALK TROLL and GET APPLE. But kids are really smart (generally much smarter than adults) so they’ll pretty soon get it, especially if they’re told ‘just say “talk troll” and “get apple” and the computer will understand’ when they try it.

2 Likes

It may be a good idea to have a limited panel to determine some special title and prize, but I really hope that voting will be open to everyone. If not, I am sure much fewer people will play the games. I doubt a lot of IF-players will play most games if they cannot vote.

I can only talk for myself, but I enjoy family-friendly games, also when they can be understood by a nine-year old kid.

Normally, competitions and jams are good places to publish your games, because you will get a lot of feedback, both comments and average scores. but if only a few people can vote, the amount of feedback will be very limited.

1 Like

I agree that it would be preferable to have a public vote of some sort - I think the problem is accommodating both types of vote (a public and a closed one) on itch. I’ll do my best to play and review the games but I’d definitely be more motivated to do so if I could vote on them as well.

1 Like

I really love the idea encouraging more beginner friendly games and games that have built-in tutorials. However it seems like some of the rules (crippled parsers, sentence caps, …) are more focused on pretending we’re in the 1970’s — pre-Infocom — than on actually making games beginner friendly.

Which is perfectly fine, retro-computing can be a lot of fun (I for one am looking forward to the next S.A.L.A.D. game :)). I just misunderstood and thought the jam was about introducing children / new players to parser IF when it in fact is about introducing children to retro parser IF.

1 Like

I disagree with the assertion that a verb noun requirement is regressive/retro.

The purpose of the jam is not to introduce children to more complex forms of text adventures or interactive fiction, but the basic fun of manipulating world state or story via simple textual commands in a known format.

As stated earlier in the thread, a VERB NOUN parser requirement avoids undue stress of fighting with the parser.

If you know everything is two words, there is no stress. Adventuron as with other engines supports parsing many more words than two words, so it’s crippling itself for the purposes of this too.

Verb noun games are very fast paced, and the other rules tries to add a brevity to the storytelling too. Fast pace and fast rewards are more important than parser precision or complexity.

Over Here! by aureas is a good example of VERB NOUN magic. Parser grappling is non existent, and the game is all about the characters and the puzzles. It’s not perfectly suitable as a TALP game due to lack of tutorial but the concept of a minimal game with lots of interactive media attached is demonstrated well. Game in the jam can use a modern or retro aesthetic. It could all be hand drawn art, or no art at all. Animations, mp3 sound effects. It could be any Google font, or ttf. Nothing about the presentation has to be retro.

There are certainly strong limitations in the parser elements of the jam, but you (generally speaking) couldn’t add complex graphics, animations, sound effects, interactive maps, or other innovations to 70s games, because of technical limitations of storage / hardware.

The VERB NOUN requirement is purposeful, and a limitation of this iteration of the jam. Children will understand the rule quickly, and be able to structure their commands accordingly.

The sentence caps are there because of attention span deficits, and being inclusive for an increasing number of children that have ADHD. There are plenty of books and comics that convey their stories with low amounts of text, so that is another design challenge for authors.

I think the modern approach is not to design a regular game with a tutorial, but to design a completely streamlined game with a tutorial. Sentence caps are part of that. There is no sentence cap on game startup text now, but there are still caps on location description and response text. I think this is 100% the right decision for being as inclusive as possible, and appealing to the widest possible range of children, including those that would immediacy switch off when confronted with a screen full of text. In many ways, I wish the sentence cap was just one sentence, and I think the most minimal games will do well in the jam, so allowing two sentences is a good A-B test.

3 Likes

I like the idea of reclaiming simple parser games from being seen as “retro”! A lot of what makes a parser game feel retro isn’t required here: if a blocky/pixely aesthetic and obscure/unfair puzzles are taken away, I’m not convinced that allowing only very short inputs will feel retro at all.

We will have to make worlds that can be satisfyingly modeled in a simpler way; that’s the interesting challenge to me.

6 Likes

I don’t agree with this thesis. What makes fighting a parser frustrating is a guess the verb situation due to an under-implemented set of verbs, not the fact that some of the verbs are phrasal verbs or that you occasionally have to disambiguate a command.

>PICK UP KEY
I don’t understand.
> GET KEY
I don’t understand.
> TAKE KEY
You pick up the key

Is much more frustrating than

> LOOK IN BOX
Which box do you mean, the red box or the blue box?
> BLUE
Inside the blue box is a brass key.

To take Excalibur as an example, I gave up at the end and had to peek in the source to find the solution of the final puzzle (the non-standard verb reflect spell) after trying oodles of commands that weren’t implemented (dodge spell, wave sword, wield sword, fight spell, …). The fact that it was a two word parser didn’t help at all.

I think making games more accessible would be better served by mandating a common verb set (and by requiring the verb set to be enumerated in the tutorial) then by enforcing arbitrary restrictions on the form of the grammar.

That’s fair enough.

4 Likes

Having a blue box and red box is something you would not design in a game for this jam. It’s a design constraint. Having two words is the constraint, and you adjust to it. The player should never be faced with anything that requires disambiguation, and therefore the problem will not occur.

GET and TAKE should reasonably be aliases in any standard engine. PICK UP is usually an alias in Adventuron, but not in this jam, where it would be reported as “The game only accepts up to two words.”

You say that grappling with the parser is mostly to do with “guess the verb”. It’s not only this though. It’s also guess the quality of the parser, guess how specific you have to be, guess if you have to be specific this one time. The VERB NOUN requirement eliminates all that guess work, and all that’s required is a simple message to say you entered too many words when you have.

About Excalibur’s last puzzle, it’s obtuse for sure, but it’s also a port, with the TALP tutorial being added afterwards that takes you to half way though (through to the castle).

I have never presented Excalibur as a perfect adventure, nor as a perfect tutorial, it’s a reference implementation bolted onto a game I enjoyed so much as a child that I bought the rights to for the purpose of teaching about the good and bad aspects of the game. The Adventuron tutorial B actually discusses design flaws in the original game and leaves them in for the purpose of discussion of design.

The jam is an opinionated jam by its own admission. A two word parser is part of that opinion. It’s not presenting itself as the only way to do text adventure games, but rather a way of trying out verb/noun as a way to increase understanding. If ultimately the experiment doesn’t work, we will have new information, and the next experiment will adapt, perhaps allowing an adjective too. There remain plenty of places to innovate within the constraints.

4 Likes

I mostly agree that a two-word parser don’t necessarily make a game more accessible, for the same reason others have pointed.

I think too that that’s is the important part. As far as I understood, Adventuron doesn’t really have a concept of actions like, say, Inform; it only has the concept of commands. While writing my game for the Christmas jam, I found that this encourages the use of one-off verbs and other “non-standard” commands, since there are no common base authors can rely on.

With Inform, I first look at and implement the built-in verbs because I know they can be considered as “standard” and that experienced players are likely to try them first. I think that a game that teaches the genre to newcomers should teach these kind of conventions used in most games instead of forcing something (e.g. the two-words parser) that isn’t used a lot in mainstream games.

Well, today, the most used authoring systems have parsers that can handle complex phrases, so you don’t need to think about it a lot in practice. In an ideal world, it’s the job of the author to think of the sensible commands that a player is likely to type (according to me).


Now, I’ve got nothing against using a two-word parser, and I don’t think it reduces the accessibility that much. Children are resourceful enough to quickly understand that you have to type 2 words at most. And as @adventuron is the organiser, he has the final word, even if I don’t fully agree with him; we may be a bit overreacting on this parser issue.

Also, the discussion now revolves a lot on whether a two-word parser is good for newcomers, and we’re not really discussing about the jam anymore (even if both subjects are linked.)

4 Likes