TIGSource Forums

I presume this little family was procedurally generated? Intriguing. I'm very interested to see how this will translate into gameplay ;D

Man, that family has it rough. One of their kids commited suicide, and all the others die of cancer? Such a grim game!

Maybe you play as an insurance salesman.

Eh?

Looks interesting though. I hope all the families aren't this morbid!

Ooops; I messed up my print statements; it was giving the day-of-the-month of his birth day instead of his wedding  ::)

(Even from the smallest snippit bugs flow forth in abundance :'(  )
Nope.

Well dying of cancer/heart disease at an old age can hardly be considered an awful fate, surely?!

Well, I did pick that one because it seem particularly sad.  I'm trying to say as little as possible until I have the code to back it up. Everything threatens to fall apart at the seams at every moment.

But: phase one is almost complete.

Doesn't look paticularly sad or morbid. Everybody dies of course but they die at a rather advanced age, so nothing unexpected.
Concept is intriguing and interesting, waiting for more...

I sort of was burnt-out most of today and yesterday, essentially getting nowhere.  I'm back on track now, but as a result, I'm a day behind.  I do have something to show for my hours of toil over the past two days, though:

Barbara Sweeney née Boland
        Father: Joseph Boland
        Mother: Gillian Boland
        Born on Friday 25 of July, 2003.
        Worked in: Education.
        Mother of 3
        Married to Barry Sweeney at the age of 38, on Sunday 28 of July, 2041.
        Sexually assaulted by boyfriend, Barry Sweeney, in 2024.
        Died on Friday 5 of October, 2068, aged 65.
                cause of death: Cancer

In the future, there is only crime, and still no cure for cancer.

Also, she married him EVEN after he raped her?  ???

It is possible.

17 years is a looooooooong time. Perhaps he changed.

I read that in Balrog's shiny voice. :D Huzzah!

Kenneth Hickey
   born on Wednesday 25 of July, 1951.
   worked in: Education.
   Father of 1
   Married to Pauline Behan at the age of 43, on Wednesday 21 of December, 1994.
   Sexually abused as a child by babysitter, Peadar Farrelly in 1962.
   died on Wednesday 10 of September, 2036, aged 85.
      cause of death: Other


So, I've finally, more or less, gotten stage 1 finished now.  Before I collapse with exhaustion, I'd like to explain what my game is actually going to be about: it's going to be a lacanian psychoanalysis game with, for the competition anyway, the emphasis on psychoanalytical trauma. 

The basic idea is that you have this patient who was, say, sexually assaulted in their past.  This is termed trauma if the event itself is repressed: that is to say, the person is no longer able to consciously recall it.  What is repressed from the conscious mind still emerges in other ways.  So the person will be coming to you, the psychoanalyst with certain odd symptoms (say, a numb arm).  And you have to figure out the root of this problem.  Of course, they're never going to consciously talk about what actually caused it; the only way you're going to be able to find out anything about the source of their problems is by catching any freudian slips they might make.  I'm going to implement this by constructing a set of words that are associated with the trauma and, whenever the patient is talking about anything, if any of the words they use are close (using the Damerau–Levenshtein distance (http://en.wikipedia.org/wiki/Damerau-Levenshtein_distance) formula), they will with some probability be substituted.

So if somebody is pissed because their mother died of septicaemia, they might accidentally get it mixed up with 'September' in a sentence.

So I've divided the project into 3 parts

1: give a person a basic life-background (enough to talk about) - essentially done
2: give a person the ability to talk (about anything except their traumatic incident) - to do
3: implement the Freudian slips (to do, but should be easy).

So, I'm totally pissed off with it at the moment, but slogging on anyway.  If things go well (I don't think it particularly likely), there's a whole swathe of Lacanian formulae for various psychological mechanisms and mindsets that would be highly amenable to being programmed in.  Another thing that I would like to do, though it would call for me to add another level of structure to the people's lexicon, would be to generate dreams (in accordance with Freud's theories of dream-formation).

For now, I have to do some bug-testing, and, hopefully tomorrow, get my patients to utter their first words.

(FTR: so much as I could, I used census reports and other governmental publications as a basis for my people-generation algorithm).

Wow, that sounds like a great concept! Is it going to be purely text-based, or have visuals too?

In any case, this makes me think a bit about Facade. If you can get a full sentence-recognizing system going, combined with your implementation of Freudian slips, this will turn out even better.

I really hope to see this working for the competition. Not many gamestep into fields like this, which offer many unexplored opportunities!

That is an amazing concept!

I am intrigued.

I must have it.

Your unconventional concept had me stroking my dibeard for a while , I confess. But I'm pretty sure you'll manage to pull out your bigthis one out and that it will be great.  :-*

Sounds like a really interesting concept.  I have thought it would be cool to make a game where you could engage in relatively realistic conversations with pseudo-intelligent NPCs...  Creating a system of free-form questions and responses is a tough problem to dissect.  Best of luck to you!
Cheers,
--Zack

Freudian Slips, The Game? Awesome!

This sounds incredible.  :o

Wow.

I want to see this working

This is going to be so much win.

Wow...this is something that I find (http://www.ludomancy.com/blog/2007/01/29/dealing-with-shrink-the-game/) really interesting :)

I got more or less to the point you are, tough my approach was a little more symbolic (I wanted to avoid making the game too text-heavy)...

Good luck with it!

Whoa, awesome. Like, REALLY awesome. Didn't see this thread before.

It's pretty much impossible to avoid making it text-heavy when it comes to Lacanian approaches, alas.  I was originally planning on implementing a lot of semantic structure in the background, but there's no way I'll be able to fit that in in time.

And yeah, the 'level design' was a big pain.  But I more or less got it out of the way (at least enough material has been included to get a prototype up and working).  I've had BIG troubles motivating myself to get on to the 'implementing speech' part of things.  Because it's not particularly interesting for me to do :/   I took two weeks off, and am going to have to cram the final two stages into this last remaining week  :(

I take it that you haven't gotten the prototype of Shrink up and running yet?

I'd certainly be very interested in seeing how your game progresses (given that we're using very different models:  I should be able to get by for the time being without any explicit emotional content -there might be some psychosomatic symptoms included if I have time, but that will be about it).

mmm...how about using a wordcloud? I was thinking of using "Free association" as the source of information (though without text)...perhaps the slips can be seen as an unfitting word in a group of semantically-related words. I know it's almost as hard as doing the speech itself...maybe you can have a "social" biography, and a "real" biography...you would pull out words from the social one, and slip some from the real one every now an then, as slippages.


Not really...I have a dumb proto, which is winnable but stupidly simple and experimental:

http://ludomancy.com/games/shrink/shrink.html

(I won't go over the mechanics here, but if you confront the fear of darkness with her cousin, you will get the true reason for the trauma).

Heh; sounds like apples and oranges (http://forums.tigsource.com/index.php?topic=1731.0) ;)

Ah; it works.  But yeah: pretty majorly prototypish!

Okay.  This isn't going to be done for the competition, possibly ever.  I got some extra stuff done, but I'm still a zillion miles away from a working prototype.  I'm pretty much happy to abandon it now, but if ever I think in the future of a way of getting a prototype to work that won't cost me my sanity and two solid week's work, I will.  Not too late for a joke entry, but I don't feel terribly motivated to do anything like that quite at the moment.

 :'(

I haven't played a procedural psychology game since Dr. Sbaitso (http://en.wikipedia.org/wiki/Dr._Sbaitso) all the way back on DOS. And that just rephrased whatever I typed as questions along with snide insinuations about my upbringing. Don't give up any post-competition plans!

I was looking forward to this... :'(.

It´s a very tough concept...

Where did you get stuck?

The language engine.  Having firstly to code in the various possible sentence structures would be a pain.  Then having to code in a way to naturally navigate from one possible sentence to another would be a pain.  Giving the character some sort of memory so that they wouldn't say the same thing twice wouldn't be too hard, but would be difficult.  Then giving some control to the player to guide the person to talk about certain topics would be a pain.  In short: pain pain pain.  Also, despite not-inconsiderable efforts on my part, all the information about the characters past was stored in a messy way.

I was thinking while walking home there that, if I were to go back to it, I'd structure everything a lot more modularly, and code the speech in at the same time as I was coding in the data for a particular event (wedding, death, job, &c.), so as to 1: make adding more content easier, and 2: to get a working prototype up quickly.  I might do it as well, but I think I'll need to mull it over to get things absolutely straight in my head before I even think about coding it.

Sounds familiar... ;)

Even though you disregarded it, maybe the trick is to avoid the speech. When you read summaries of real psychological therapies that lasted years, you realize there are only some key moments of intervention during the treatment, and the actually important information from the patient can be explained in a short text (of course, you're leaving out many factors, but making a human simulator is out of our reach right now, I think :) ). That means most of the speech is actually not interesting...

Of course, I have no applicable ideas on how to solve that problem either.

True, but the problem in terms of procedurally generating stuff, for a computer to automatically generate interesting scenarios, you (I think) need the possibility that the key episodes in therapy could occur at any moment, when they were talking about anything.  So, in some sense, it seems to me necessary that they have the ability to blab on and on about a wide range of things, because ideally I want the possibility that any one of those could present a clue.  One might reverse-engineer one, I guess (starting from two words that are closely related, associating one with a traumatic incident, another with some relatively harmless activity some years later), but I'm not sure how well that would work. But also: without giving your person a large range of things they *might* say, the possibilities for interesting slips will be tiny.

Hmm: okay, I'm thinking about it again.  To construct the trauma, and the possible slips I could do in an hour or two's programming (on top of what I already have): it requires no coding effort whatsoever.  That would come out with something like "man raped by father.  father wore blue jeans. he keeps on slipping the word 'blue' into his sentences by accident".  But what could I do with that?  It's very much a single event.

I guess I could just have the computer output a gigantic file of all information about the patient's life (in tree form), but with the traumatic event left out, and with all the substitutions put in.  But that doesn't exactly have the same dynamism to me   :D

(on a not unrelated topic, I started working on the planning of a basic Freudian dream simulator today, which I might be able to do without too much effort).

The answer is simple: You want to bang your mother.

That's right.

Psychology works on how you deal with the fact that you can't.

And before you wonder, if you actually do, it's even worse.