Abstracting Dialog

[Hobbit Lord of Mordor]

So in this thread we brainstorm how to transform dialog into gameplay similar to combat.

Starting points: The actual content of the dialog is irrelevant. No 'dialog trees'. Contextual meaning of phrases is ignored. Outcome is based purely on stats and strategy/player judgment.

The purpose this kind dialog in this case is not to gain information, or clues or to socialize with some NPC, or to 'learn the lore' or 'roleplay'. However you could gain information, quest flags or disposition changes of key NPCs as an outcome of winning/failing/(or your level of success) in the mini-game. Also the 'mini-game' doesn't have to necessarily replace standard dialog trees, which could be used alongside it (for routine conversations or 'flavor' dialog)

This would be a 'mini-game'/sub-system used for example to convince some NPC to trust you, or for example to convince the king that he is being led into a trap in the presence evil mage/advisor present who is trying to paint you as a spy or traitor. Game context would come into play through bonuses or new abilities, for instance finding a letter suggesting the mage is up to no good would supply an extra 'attack' or strategy to be used at the right time in the conversation.

I once played a game, it's a political simulator (election-day.info) which has a debate mini-game. The debate goes on for 4-5 rounds, each candidate (up to 5) is given one action/choice in each round (+ choices if attacked and allowed to respond).

Either the topic of the round is chosen or a random candidate chooses it (each candidate gets to start one round)... foreign policy, city planning, family values, etc. You also choose stances: explain your policy, make a joke, speak in generalities, attack an opponent's policies. If you attack an opponent you get to choose which opponent, and the kind of attack you make, of which there are four or five (indignant, aggressive, sarcastic, subtle dig, serious business point, etc). A character who is attacked gets extra choices. He can attack back personally, turn the criticism to attack the original candidate's policies (on another issue), or he can simply ignore him. You can also attempt to change the topic being discussed in the round to another issue.

The connection between all these choices is a meter for each candidate, which shows their cool/poise/body language/how they are perceived by the audience. It goes from "total wimp" to "bully", and is changed by your actions, and reactions. If you attack another candidate aggressively your meter is pushed toward aggression, and he is made to look like a wimp, especially if he doesn't fight back. A more successful candidate, who is better at attacks suffers less aggression change relative to the "damage" he inflicts (he can afford to be more aggressive and keep his 'cool'. There are stats to do with how effective you are, how telegenic, how charismatic/likable, your policy understanding, even your mental health.

Each round is rated and after the last round rated again. The average cool/balance of each candidate + final state determines what people think of their performance. In addition, if you managed to push your issue position (less effective if you are ridiculed/get in a fight), that can increase your support with some demographic.

So your two conflicting goals are: 1/ push your message (that you are the only person with the leadership to get the school bus running on time/that the sitting mayor is an moronic jackass and probable traitor), 2/ to look normal and sane, and non-partisan.

The trick in multiple candidate debates is to avoid confrontation, and get an attack in at the last minute on your primary rival, without giving him a chance to respond.

If you are weak on policy you try to move the debate onto 'non-issues' (demographics, populism, 'corruption', family values - which you get bonuses on based on your background, most effective against gay, divorced, Afro-American, democrat, academics from another state, in rural towns), or turn it into a general brawl of shit flinging and sarcastic snipes, at the risk of looking like a douche.

Anyway not directly relevant to RPGs but it is one example of abstracting dialog. Another way would be to use stats, or 'card deck' type gameplay (you have a randomly distributed hand of different kinds of cards, attack, change topic, reflect 'damage' back, which you 'spend').

Obviously numbers/quantification need to come in somewhere, the dialog equivalents of hit points, saving throws, and armor and damage rolls.

Any existing RPGs have this?

 

[DraQ]

tl;dr

Also:

So in this thread we brainstorm how to transform dialog into gameplay similar to combat.

 

[Hobbit Lord of Mordor]

You're hilarious, now fuck off

 

[mondblut]

Check the "debate" minigame in Romance of 3 Kingdoms 11. Basically an Arcomage Lite with cards like "insult", "reason", "agree" etc.

Or the monstrous Fate: Gates of Dawn, with a bunch of common options like "talk", "ask about", etc each having a bunch of common suboptions like "tell fibs" or "(ask about) magic" or "join (party)" ad infinitum, and the random responses you'll be getting depend on your stats, quest flags, NPCs (they are all randomly generated) and a lot of luck. Also, each suboption can only be used once, and an NPC disappears once parted with, so if you failed to get what you wanted, you have to look for next one.

Of course, nothing of it is applicable for those who want to be told a Sophisticated Story in form of Narrative Dialogues. Those mechanics, they sound like some lowly "game", not High Art. You can't even say same shit in "good" and "evil" flavour!

 

[Emotional Vampire]

Consider the fact that if you had to "battle" every time you engage into a dialogue you'd get sick of talking to people pretty quick.

Could work with major NPCs though.

 

[JarlFrank]

Check the "debate" minigame in Romance of 3 Kingdoms 11. Basically an Arcomage Lite with cards like "insult", "reason", "agree" etc.

Yeah I actually liked that debate minigame and thought it could be used in RPGs, too. Definitely better than the shitty shit that was Oblivion's persuasion wheel. Oh gods.

 

[janjetina]

Sounds like oblivion. Why would anyone want that?

Anyway, an idea worthy of an illiterate ESF refugee and a

 

[Serious_Business]

Codex at its best when shitting on a potentially serious topic

I think this is interesting, RPGs definitively need some revolution regarding dialog, if we want to step away from combat. Say, in a rpg you play a role - meaning playing an identity - and how do identities form themselves? In dialog. Dialog with others, with ourselves, with authors, God (!?!?!), what have you. So it makes sense for a rpg to have a lot of dialog, but you can't have a game clicking on dialog nodes having a pre-determined virtual conversation, this is just silly.

I like the debate idea, but it's not appropriate in all cases. Especially in shit like Bioware game where 90% of the dialog will be the characters speaking about their feelings and past. Anywho, I don't have much to add here, except that the problem is huge. We want to step away from combat, again, because it makes no sense for a rpg to have so much combat, because a rpg is not a strategy game. How can dialog be made into gameplay? Is it even a possibility? Are rpgs doomed to be a tained genre, never reaching its full potential, being in the end a mix of adventure, action and strategy games? Because that's what rpgs are : variation of a design ideal (the tabletop gaming session I imagine) that can't be done in a PC environment.

Disgression aside, maybe an all exclusive debate rpg would be fun... not sure how you'd create the characters thought (they'd all be debaters, so have stats like charisma and so on, so you'd choose what is most appropriate regarding your character's stats... very linear in a way). But yeah I don't think it works that well in a larger picture, well, it's worth looking into.

You should post this on Iron Tower it would be taken more seriously, and you'd have better answers.

 

[spectre]

Revolution for the sake of revolution, I don't think so. But the idea is worthy of digging into.

I piss on your juvenile scorn on Oblibian's wheel minigame, and instead propose to see what it does right and what it does wrong.

(o) It's repetitive, banal, shit, boring. K, we're done here, but on the other hand:
(o) Forces you to find out what are the most effective ways of persuading the npc.
(o) It combines player skill and character skill - the player need to plan how to spin the wheel, but higher stat allows for more options and better degree of success.
(o) No real consequence for failing it.
(o) No challenge there.
(o) Generates inane responses.

Now, just looking at this, I see two major issues when designing the system: balance it so that it isn't tedious, but a fun challenge, and make it so, that it feels, at least vaguely, as a conversation.
The latter part may actually make it not worth it in the end, cause while I believe it can be done, serious implementation will eat up man-hours like crazy.

I think the minigame should be split into phases - at first you probe the npc, how does he behave in a discussion, is he agressive, avoids direct confrontation, does he use reason, intimidation, how does he respond to different kinds of persuasion, finally, what can you gain from the conversation, learn a secret, a clue, help, complete a quest, reduce prices in a shop?
What you learn during this step could depend on stats such as Empathy, Sense Motive, Perception, stuff like this.

Then comes the actual minigame. I would envision it as a chess board, with startiong pieces and positions depending on each contestant's stats.
The objectives on the board depend on what you learned during the previous steps - give a check, a checkmate, capture a field, capture X pieces, win in under Y turns - all these will unlock different results of this conversational battle.

For example, you are caught by city guard, he demands to know what are you doing out in the night during a curfew. The guard will try to intimidate you, you choose not to confront him directly in the game of intimidate, but rather to avoid. This brings up a chess board, the guard has two rooks, you have two knights. The result of this battle depends on how you do:

If you avoid taking any losses for 12 turns - he lets you go.
If you escaped for 12 turns but lost a knight, he lets you off with a small bribe.
If you managed to stand on a D4 square for two consecutive moves, you learn some interesting information about local situation.
If you captured both of the rooks in the process, you win over the guard completetly with your story about a sick grandma, and he escorts you to wherever you wanted to go (plus some bonuses).

The problem is with making it look like an actual conversation, writing up randomised dialogue for each move (with special dialogues for captures and other moves) is an option, but is tedious as fuck.

 

[Phelot]

I was about to post a new topic about an idea for dialogue but figured I'd post it here since this is similar.

Rather then a mini-game, I was thinking of having a personality system for the main character to determine what dialogue choices you have and more importantly what your character does based on their personality.

Essentially, you pick either traits or have a slider bar between two opposite personality points such as (but not limited to):

Brave - Cowardly
Charitable - Greedy
Temperate - Hot-Headed

And your dialogue is decided for you so for example having a hot-headed character makes your character jump into combat at the slightest provocation, a cowardly character will give up money or information to someone intimidating them, greedy characters will not give you an option for turning down a reward, so on and so on. During most dialogue, there would be chances to reshape your personality slowly so a cowardly character could eventually become more brave. Having a witty character will allow for an appropriate joke at a appropriate time, perhaps impressing an important figure.

Each opposite would have both good and bad sides to them. Being cowardly to a power hungry city guard would typically be better then getting into a fight with him, thus getting into a fight with all his buddies.

The point of all this is to avoid the classic dialogue skills such as persuasion, intimidate, etc, which seem to only do good for the character. Basing dialogue more on personality should hopefully make replays more unpredictable though I realise how much work this would entail. It's just a fantasy :D

 

[AzraelCC]

Diplomacy is directly tied with charisma and perception. When a person tries to take advantage of you through words, it's not just because he's a persuasive talker--it's because he perceives that you might fall for his ruse. This is where I think a new dialogue system should be based on.

For example, in an RPG with a skill based system, you have two main skills to use for diplomacy that is available for all characters: A "Judge Person" (JP) Skill and a "Basic Talk" (BT) Skill.

The JP skill would be passive and related to one's wisdom/perception, activating every time you enter an area with NPCs. Each NPC would have a Challenge Rating, with important quest-giving NPCs having fixed CRs, and filler NPCs (townsfolk) and merchants having random CRs (though of a range that is not too large). These CRs after being generated or set are influenced also by the quests you've achieved, or your reputation in the area. Dice are rolled and your JP skill is added. If you meet the CR of the NPC, you will see an icon above their heads that could be green if they are willing to talk to you. There would also be a number that represents the number of dialog actions (DA) you can make (more on that in the BT skill). Yellow icons always have a 1 dialog option rating, and red means that for some reason, that person does not like you and will ignore you or even use you. So a character who invests in this skill can quickly see NPCs he can manipulate, haggle or get quests from and receive good rewards from. A player who ignores this skill can potentially talk to an NPC that want to harm him or her.

The BT skill activates when you engage in a conversation with a person. The dialogue doesn't involve dialogue trees, but presents you with Dialog Actions (DAs). There are three DAs in the BT skill: Gain Information, Ask Favor and Influence. You can perform as many actions depending on the number mentioned above in the JP skill. Once you consume these "tries," the person will no longer talk to you or give you significant returns. So you can ask one NPC purely to gain info, another to milk for favors, or combine the options depending on the number of DAs you have.

Gain information is pretty obvious--it gives you information. If done on a filler NPC or important NPC, dice are rolled and the BT skill is added. This is compared to the CR of the NPC, and depending on how high your success or failure is, the NPC could give you basic things to know in a town, secret quests, or if you fail, give you false information. If done on a merchant, you can have items identified immediately depending on the rarity of the item.

Ask Favor asks you for further options of what favor you want to ask. It maybe asking for a filler NPC for some copper to help you, a merchant to give you spare items he doesn't need, or a quest giver to give you more gold for a successful quest. The favor options available depending on your BT skill rating and each favor has a corresponding CR depending on how big the favor is.

Influence is the most powerful option for characters who want to focus on diplomacy in an RPG. A low BT rating would simply raise the number of DAs you can perform on that NPC. Once you reach a specific BT rating, though, you can start gathering influence points for that person, which is a more powerful form of ask favor. You can ask a filler NPC to let you rest in his house every time you come to town, or an adventurer to actually do some quests for you. The thing with influence, though, is that you use influence points, there are also material costs like gold or specific items. That is, unless you have the advanced talk skill, which I'll probably post later.

 

[The Ticktockman]

The Sims dialog system.

 

[zenbitz]

I am not sure it's relevant but I have a great board game that I love called "Tales of the Arabian Knights".

The gist of it is that you encounter a noun - person, place, or (sometimes) thing. That noun often has an adjective associated with it "i.e, "Imprisoned Princess" "Old Man", "All-Powerful Efreet", "Glittering Artifact".

For each "class" of things (there were about 10) you could pick from a list of about 10 verbs "attack, enter, aid, seek help, court, converse ,etc." and there was a big matrix modified by your skills and a +/- die roll to see your "story".

One could probably use a similar thing as an extension of dialogue trees with various "verbs" being only talky-talk options (attack and rob being handled by a different engine) - although perhaps this looks too much like FO3 [Persuasion].

Instead of a flat adjective you could of course use a whole personality for the NPC (as others have suggested above).

A list of "talking" verbs:

Orate (persuade by emotional argument)
Lecture (persuade by logical argument)
Intimidate
Flirt (or Seduce)
Swindle
Grovel / Brownnose
Plead ( for aid )
Demand
Befriend

To make whole - customized by NPC - dialog trees for each of these options multplied by all possible reaction matrices (PC vs NPC)... would get out of hand. You would have to some how generify it.

 

[Hory]

Exactly. There's no system more abstract, flexible and intuitive for communication than natural language. It has, after all, evolved through thousands of years of use. Combinations of verb + object (+ other objects or subclauses) can cover pretty much anything. Even a command such as "TELL Character1 THAT Object2 IS IN Location3" is pretty easy to parse. The difficulty is in covering the large number of combinations that are possible. Tho I have never used such a system for a real game, I don't think that the amount has to be unmanageable. A good writer doesn't add fluff to his story. Each NPC will have certain roles that can be scripted manually. They don't have to be generic walking encyclopedias. Most of the verbs and topics don't have to work on them. That's the challenge for the player, determining what approach could work (rather than plowing through the dialogue tree). Simulated chat bots don't add anything meaningful to the story.

 

[DraQ]

The problem is with making it look like an actual conversation, writing up randomised dialogue for each move (with special dialogues for captures and other moves) is an option, but is tedious as fuck.

No, the problem is using something completely unrelated and not analogous at all to a given activity to simulate it.

Am I the only one here seeing how wrong it is? You can't make dialogue minigame resembling chess right any more than you can make a combat minigame resembling knitting, or a lockpicking drinking game. It's. Just. Too. Fucking. Unrelated.

I was about to post a new topic about an idea for dialogue but figured I'd post it here since this is similar.

Rather then a mini-game, I was thinking of having a personality system for the main character to determine what dialogue choices you have and more importantly what your character does based on their personality.

Essentially, you pick either traits or have a slider bar between two opposite personality points such as (but not limited to):

Brave - Cowardly
Charitable - Greedy
Temperate - Hot-Headed

And your dialogue is decided for you so for example having a hot-headed character makes your character jump into combat at the slightest provocation, a cowardly character will give up money or information to someone intimidating them, greedy characters will not give you an option for turning down a reward, so on and so on. During most dialogue, there would be chances to reshape your personality slowly so a cowardly character could eventually become more brave. Having a witty character will allow for an appropriate joke at a appropriate time, perhaps impressing an important figure.

Each opposite would have both good and bad sides to them. Being cowardly to a power hungry city guard would typically be better then getting into a fight with him, thus getting into a fight with all his buddies.

The point of all this is to avoid the classic dialogue skills such as persuasion, intimidate, etc, which seem to only do good for the character. Basing dialogue more on personality should hopefully make replays more unpredictable though I realise how much work this would entail. It's just a fantasy :D

This, on the other hand, seems at least interesting.

 

[spectre]

Am I the only one here seeing how wrong it is? You can't make dialogue minigame resembling chess right any more than you can make a combat minigame resembling knitting, or a lockpicking drinking game. It's. Just. Too. Fucking. Unrelated.

Not a big chess player than, are you? With a bit of imagination it resembles discussion a lot.
Chasing the opponent into the corner for a checkmate, launching surprise attacks with long range pieces, overwhelming the opponent with heavy arguments (pieces).

 

[DraQ]

Am I the only one here seeing how wrong it is? You can't make dialogue minigame resembling chess right any more than you can make a combat minigame resembling knitting, or a lockpicking drinking game. It's. Just. Too. Fucking. Unrelated.

Not a big chess player than, are you? With a bit of imagination it resembles discussion a lot.
Chasing the opponent into the corner for a checkmate, launching surprise attacks with long range pieces, overwhelming the opponent with heavy arguments (pieces).

I'm more of a Go person actually, but I have a hunch you might want to talk to people IRL a bit more often.

 

[spectre]

Your hunch is wrong then. Next?

Talking =/= getting what you want from people.
The key point is conflicting interests resulting in a confrontation which can be dealt with in a number of ways, depending on each person's personality and knowledge.

 

[DraQ]

Your hunch is wrong then. Next?

Talking =/= getting what you want from people.
The key point is conflicting interests resulting in a confrontation which can be dealt with in a number of ways, depending on each person's personality and knowledge.

And how will you represent personality in chess?

 

[spectre]

Re-read what I wrote a few posts earlier and you'll get an idea for a possible solution.

 

[jagged-jimmy]

How about the dialog "fights" in Republic the Revolution? With some extension and rule design it could work.

 

[spectre]

Didn't play this one, what does it look like?

 

[zenbitz]

What I don't understand is why make dialog like combat at all? Just use combat and change the words.

It's not like combat in RPGs is very deep anyway "spam best attack and heal when damaged"

 

[DraQ]

Re-read what I wrote a few posts earlier and you'll get an idea for a possible solution.

Looks horribly... ...unrelated.

But ok, let's assume that your chess minigame does indeed have some relevance here - so what? The problem with minigames is that they break the flow of the game - I, for one, do not sit down to the Newest-Awesome-aRPeeGee-With-Awesome-Branching-Plot-Ample-Choices-and-Consequences-Brilliant-Combat-And-Ultimate-Stat-Mechanics quivering with anticipation that I will finally get to the heart of the game - chess dialogue minigame. I don't know, I'm just weird this way and find such a BLAM rather annoying when I'm eager to see what intrigue holds when suddenly a chessboard slams me in the face challenging me to beat some fuck in a game of mental chess before I am allowed to continue the game.
It's about as entertaining as this lengthy 15' of ads and trailers at the beginning of a film in a cinema or TV commercials of washing machines right before the climax.

Even if you manage to make a minigame that resembles simulated activity closely enough to not pull the player out of the game, it can still be fucking annoying - see lockpicking in a game we all love to hate - Oblivion (let's ignore the fact that it's horribly broken in regards to stat-wise mechanics, it's not terribly relevant here).

If you absolutely have to put minigames in, make them strongly enough related to the subject to not pull the player out of the game, and make them as short as possible (see skill minigames in Anachronox or Hack/Repair/Modify minigame in SS2) - there is a reason they are called "minigames" - they don't belong in the game proper.

Besides, the minigame for dialogue is unnecessary - at most you can just put in some more fluid range of tones, maybe make time flow count, and tidy-up dialogue trees, so that they flow more naturally and you don't get to walk the whole tree with random NPCs avoiding stupid sounding loops and making dialogue choices matter, for generated content just add something resembling question and answer generation based on topic as seen in Daggerfall. There is no point fixing something which isn't broken.

 

[spectre]

Besides, the minigame for dialogue is unnecessary - at most you can just put in some more fluid range of tones, maybe make time flow count, and tidy-up dialogue trees, so that they flow more naturally and you don't get to walk the whole tree with random NPCs avoiding stupid sounding loops and making dialogue choices matter, for generated content just add something resembling question and answer generation based on topic as seen in Daggerfall. There is no point fixing something which isn't broken.

Yes, it is arguably necessary, and yes, it is preferable to perfect the exisiting formula, but since we're just bullshitting on the internets, and it costs us nothing but free time, the point is to explore what (if anything) can be done to create an alternative to the tried&trusted dialogue trees.
Sure, it is as likely to produce something akin to farting and belching from Fable, but I have hopes that a bunch of some codexians can at least do better then Peter M.

The problem with minigames is that they break the flow of the game

You're right on this one. The main problem with such features is to make them fit.
When abstracting dialogue as combat it would only work if many features from dialogue and standard combat overlap, so that it is overall consistent.

I think this is the crux of the issue, you can alwas say that it's unnecessary, and too much work for little gain but it's till fun as a design exercise.