Well, for this complex stuff, we can use Seduction as a benchmark, since if you can do that you can do almost anything else too! So what are the various steps needed to execute a convincing seduction, and how do we make them intuitive?

(Isn''t that what the How to Get Women 101 post was about, too? =)

Seduction? And what would that lead to? Virtual Love Making? (Simulate that! ...or get a life)

I don''t think its possible really to seduce anyone in a game. No matter how good an interface or how great the graphic is. It''s still just a game and we all know that there may be a big ugly guy roleplaying that lovely virgin

My point is that you don''t need the seduction to be convincing. Just recognisable. You can''t really seduce anyone but if you roleplay good you might get someone to roleplay along and let themself be seduced.

Maybe even a good old text interface would be enough to achive this. For example the one roleplaying the woman may type:

''smile seducting to the man''
''say "Oh, Landfish!"''

And Landfish may see something along the following on his screen:

''The woman smiles seducting to you and says "Oh, Landfish!"''

If Landfish is a good roleplayer and the situation fits he may then decide to roleplay along.

Of course the main game could still be graphical. Perhaps a beautifull rendered 3D world. The text boxes just need to be layered on top of the graphic. Or maybe in seperate boxes below or besides the main playing area.

Voice support, lib syncronization, animated body language and all that would of course spice up the situation and help the not-so-imaginative players but I don''t think its really necersarry to achive roleplaying in a conversation.

Regards

nicba

What you need here is a set of body language components that can be combined to make behavior macros, with hotkeys.

You know, you have certain choices for eyes, head angle, posture, mouth, whatever. Give the player a fun little area to mix and match these to get the desired effect. They then save them into a behavior macro with hotkeys. Maybe they can even name them.

If you add the ability to change the components along a timeline (ie, create short animations), this would take off like crazy. People would create all kinds of actions, and download them off of the web.

I'm not a programmer, but I imagine the work wouldn't be in the code, but rather in creating so much art for each character. What we need is to get a look at the code for The Sims...

- gollumgollum



Edited by - Gollum on July 11, 2000 9:00:48 AM

If the player is going to have to keep track of all these variables the game will stop being fun. Which means it will stop being a good game. Which means nobody will play it.

I don''t mean to sound rude, but I really don''t like your ideas. Was that rude? No seriously, no matter how hard you try to design the interface, there is a certain number of variables you need to keep track of simultaniously, and you won''t be able to get around that. Nobody''s going to want to play something like this.

First and foremost - Warren Spector isn''t evil - and secondly, comparing him to you is liking comparing god to a lowly freak of evolution (landfish).

However - it''s good that you raise this point. For a long time have we been limited by the stereo-typical interface. One hand on the keyboard one hand on the mouse. Mouse means point, keyboard means select gun, mouse means kill. It''s tedious, it''s tried and true - but it was proved in games of the past, not games of the future.

Not only is there are need for a major revamp in software interface, but also the interface in terms of hardware. I''m not here to start a hardware revolution, I''m just a programmer and a designer, which means I''m going to focus on possibilities where we can improve the software interface with what we already have.

Simulating seduction and other such emotive skills is an easy accomplishment in single player, the NPCs are easily wavered by your pseudo-charisma with the role of a device, but in a multi-player situation you can''t tell a player they are being seduced. However to convince someone in a multiplayer situation that you are telling the truth and other such things is far more complicated than just lips, eyes and body language.

But I''ll assume that the computer can simulate an environment that can accomodate this level of playing (role playing actually I s''pose). So given the environment is out of the way (perhaps I''ll leave it to you guys how you''d make up a game that supports strong communication lines between players) then focussing on how you can convey this to another player is the only task left.

So this means we are left with just body language, lip movement and eyes. To be relatively honest, i think it''d be somewhat necessary to have an entirely different mode of play for wehen conversing with other players. Out go the old which sword have u got equipped and where you are walking and what you are bludgeoning - those are out of the way and it''s time to focus on one of those three things.

Lip synching should be the easiest. It really is a matter of getting a processor powerful enough to recognise what sounds your mouth is making and then interpolating between the previous mentioned mouth positions. Sure it''s processor intensive, but as mentioned - these are ideas, not game design docs in progress.

TOne is controlled by the players input into the microphone, it''d be almost impossible (almost ) to actually try to geta computer to simulate tone.

COntrol over the eyes can''t be too difficult? Why not just have the player hold down a mouse button and move around the eyes from the eyes point of view - that can''t be too big a deal? But what about body language? Body language could stuff up the whole idea. Sure the microphone and lip synching is still taken into account, but how would you control body language whilst moving the eyes?

There are several solutions, make it so that the body language is controlled by a series of confusing buttons which make present poses and allow the player to move the eyes around still with the mouse. Personally that kinda suits me best, sure the players may hate it to begin with, but if they get used to it then it''s better than breaking up the action with a screen before or during conversation to choose your next actions.

Possibly you could have the player click and drag on body parts to move them around and an invberse kinematics system takes care of the rest - but how would you move two arms away from each other at the same time - this would require that the player gets drawn out from the game to a pause point and then makes these planned actions and then executes them. Again, hardly plausible especially in real-time - plus it has the problem of not allowing the player to control the eyes whilst moving the body even if it were somehow implemented in real time.

So perhaps I''m thinking a little too narrow here? Why not just shove the player in a full body suit and motion capture - sure that''d solve all our problems, but i''m realyl at a loss as to how this could be done without further enhancements in what the manufacturers of hardware give us! Can only work with what'' we''re given and we weren''t given the equipment to translate a persons actions and emotions into a game. But give it a few years.

Ok, I''m gonna plug my idea again.

Have you ever watched somebody play Starcraft? Or someone who is very good at Unreal or Quake? If you are into a game, hotkeys are no big deal. If you let the player decide what keys to use, it''s even easier.

An area where the player could set up and test new moves would not be an extra annoyance, especially since they wouldn''t _have_ to do it. It''d be a hoot. Imagine trying to script your character to do the macarena, or limbo, or strike an Austin Powers pose.

And the scripting would be easy, right? You''d just set up variables (or an array, or linked list, or whatever, like I said, I''m a newbie) for each body setting, then change them at certain intervals. Just come up with an interface to let the player assign the variables.

Maybe I should let it go. If no one thinks it''s a good idea, then when I _am_ good enough to do it, watch out!

- gollugollum

See? I said in my last post that I fully expected people to not like this idea, or not like the game if it were realized. So fine. Ok, you''ve given your input (naysayers), but the idea isn''t changing. Games for combat will always exist.

As for the more de/constructive responses, I have to say I like Gollum''s the best. This way, if you actually created your own facial structure, you could choose a body language and distinct facial expressions to suit it. You could base these off of real people you know (my brother does this eyebrow thing), hence creating characters with real personality. And hell, seduction is 90% personality (and 90% physical.)

The main benefit is that macros and such tend to limit the system to one, or at most two button presses. A beginner might start with a set of ten (the top row number keys) that are constantly shown on a bar at the bottom of the screen. A pro user migh have every key on the board used, with some kind of hot key (If text based). Probably one for every event. This is important, because one of the detractors up there was right on the mark, too many variables and the characters won''t feel human. The system would be unmanageable and no one would play it.

Body language should have two states: Inert and active. Generally the active movement doesn''t kick in until the character is actually talking. Example: If angry is the mood, character doesn''t start flailing his fist in the air UNTIL he''s yelling also. The body language would be triggered only when speech was also. Inert mode would generally be an ambient thing. You can look at someone, and if they are sitting close in a corner with thier head down looking at the floor, they might be funming or sad, but certainly not happy!

I have to go, I''m looking forward to more responses.

Humm.. ok landfish.. time to get some Niphty! ;p

As it seems to me.. you want a graphical game in which you can do unlimited number of actions with your body.. animated on screen. Well.. ok.

unless everyone is running broadband, no voice capture is possible. I played several early on games with voice capability.. and they sucked. I knew people who worked on voice-over-IP codecs and whatnot for all those internet phone companies. It''s all really basic when you think about bandwidth.. there''s never enough.

Now.. imagine you tone this down a bit. Think.. the Sims, multiplayer.. text-based entry system with graphical representation of the character.
You have a text box which shows you past conversation in your area and what you''ve done recently.. consider this short term memory. You''ve got the screen, which is live-action. And you''ve got a text box to enter in commands. You can move via the mouse or arrow keys, at your discretion. You can use the mouse like in the Sims and click on people to talk to them. you then walk to them and begin to talk. The Sims has all of this down pat and is working on expanding the database of actions and Sims Sounds. Now, instead of having meters for friendship level with other people in game, you could have something like that which serves no actual purpose.. or something. The friendship meter in the Sims determines what actions you can do with other players. If you have this as a mere meter in the game, but not a determining factor, that''d be cool. if someone hugs you then, you could have a hotkey.. say esc key.. that means "i don''t like this person and what they''re doing." Your character would then do like in the sims and push the person away or something
Now, if you allowed this friendship meter to exist in it''s form that it''s in now.. then you could simply choose who you are and aren''t friends with.. and what level of friendship you are with them. Then your character would be partly automated, based on the choices you''d made in respect to other people. You could always override these automated controls, like in the Sims.

Now, you could also type in the text box something like "look at landfish" and you''d look at them. The text for things would appear in the short-term memory box: "Niphty looks at landfish." The character on the screen would also turn to look at landfish. Now, this gives TOTAL flexability based on if you perfer graphical things or what.
Now, you have verbs in the game:
kick, kiss, tickle.. etc.
kick would show up as "you kick landfish/niphty kicks landfish/niphty kicks you" just like in a MUD. On the screen, however, my character would walk to landfish''s and kick.. literally. And landfish''s character would grab his leg and hop for a second.
kiss and tickle work much the same. Now.. suppose you added this into an RPG environment. Let''s say landfish has more dexterity than i do.. and he dodges my kick. This becomes true interaction
now, the game would have to take into account things like "is niphty in landfish''s line of sight?" "is landfish too busy to move?" "does niphty have such a big foot that landfish cannot possibly avoid it?" "does landfish want to avoid it?" hehe I''m assuming that as friends, landfish wouldn''t nessicarily avoid it.. as long as i didn''t kick hard but once i kicked hard, would landfish still really be my friend? hehe.

So it''s a big debate from there.. you could have hundreds of levels of friendship alone.. and all of those dictate things like "would so-and-so allow such-and-such from whoever?". I like the Sims for having a few generic levels, but i don''t know if you could get away with a game like this with less than 10 levels of friendship

Oh, landfish.. bet you didn''t think i''d been thinking about a game like this, eh? ;p Ever since i got the demo disc for the Sims i''ve been wanting to turn that kind of AI and such into a game. I impart my knowledge on to you in hopes you will help me refine it and join me in making something crazy as hell as a project like that! Cause it''s one i can''t do alone.. i need a dork like you to fill in some details for me ;p

J

I liked the idea.

I think the important thing is that most of the actions should be automated. (i.e., you automatically do something). You could change settings for this in some sort of player management screen where you might type in some sort of script or choose from a list box what you want to do in certain situations. So if your playing the "charming gentleman character", your guy will automatically smile and wink at pretty female players . And, of course, if you playing the evil goblin character you might start growling at human players.

In other words, you should be able to set most of the variables before hand and then have the computer automate it. (in other words, you''re telling the computer what you want your character to be like).

----------------------------------------
Whenever I see an old lady slip and fall on a wet sidewalk, my first instinct is to laugh. But then I think, what if I was an ant and she fell on me? Then it wouldn't seem quite so funny.