Archived

This topic is now archived and is closed to further replies.

Diodor

what does the I in AI stand for ?

Recommended Posts

We carbon-based units have kept laughing at AI for being stupid, repetitive, unable to adapt, etc. whereas we are the exact oposite. But how intelligent are we ? True, we tend to learn stuff like what units / build paths / weapons / order to solve quests / places to camp will be the best for achieving our game goals, but how much intelligence does that take ? Human actions in a lot of games are just as repetitive as the computer''s. It''s incredible how many times a player can only improve his performance by simply improving its mouse click rate and accuracy or storing in his head a big database of everything that can happen in that game. So, my question is : "what is the smartest, thought intensive thing you ever did whilst playing a game ?" I placed this question in the Game Design forum because I want to know what game design allows/requires human players thinking.

Share this post


Link to post
Share on other sites
Interesting play on acronyms

Instead of concentrating on AI, we should, perhaps, concentrate on allowing the human players to make intelligent decisions within the game by finding interesting paths to goals and around obstacles. Basically, allow the players to use intuition and thought to aquire their goals.




Need help? Well, go FAQ yourself.
What a plight we who try to make a story-based game have...writers of conventional media have words, we have but binary numbers


Edited by - Nazrix on March 18, 2001 7:55:21 PM

Share this post


Link to post
Share on other sites
Adventure games require the most thought. Games like Space Quest, King''s Quest, ect. The cost of quality voice acting doesn''t it make it a popular genre for developers. The cost of video killed the FMV Adventure Game quite a few years ago.

Ben
http://therabbithole.redback.inficad.com

Share this post


Link to post
Share on other sites
yes, but those adventure games generally made it it so that you could only chose the 1 option the game designers decided.

What would be better is allowing for many ways to achieve goals (preferrably using a strategy game approach as discussed here)





Need help? Well, go FAQ yourself.
What a plight we who try to make a story-based game have...writers of conventional media have words, we have but binary numbers


Edited by - Nazrix on March 18, 2001 9:47:39 PM

Share this post


Link to post
Share on other sites
What about Sim City?

A designer doesnt need to know everything about code, they just have to have an appreciation for its limitations and how those limitations affect features they may wish to include in their design. - Drew

Share this post


Link to post
Share on other sites
Yeah, I think sim games in general give a good example of PI (player intelligence )



Need help? Well, go FAQ yourself.
What a plight we who try to make a story-based game have...writers of conventional media have words, we have but binary numbers

Share this post


Link to post
Share on other sites
I think the most interesting are non-linear adventure games (bladerunner for instance).

Hmmm the smartest? I dunno... games dont exactly encourage smart behaviour.... they just make you feel smart by rewarding the behaviour that takes you along the road to the inevitable outcome.

Share this post


Link to post
Share on other sites
I think Thief rewarded the players'' brains...



Need help? Well, go FAQ yourself.
What a plight we who try to make a story-based game have...writers of conventional media have words, we have but binary numbers

Share this post


Link to post
Share on other sites
UrrrrrghghgkKK?

Adventure games? First person sneakers? Sims? Yeah, well, maybe.

But hands down, when you look at all the factors that have to be considered, ya gotta give it to strategy games.

Take a game like Starcraft. Incredibly complex. Many factors depend on other factors in a circular fashion. Who are you fighting? What are your strengths? What are their weaknesses? What is the terrain? When do you move? Scout? Defend, regroup, withdraw, feint, counter, build?

And where?!?!

And how!?!?!?!?!?

Even worse (omigod!!!!) why???? Every action creates a tumbling cascade of consequences. So much shifts and changes depending on this fluid and changing tree of possibilities. If this is static and you''re playing other people, either the ruleset is bad, or the players are unimaginative.

I don''t pull the same moves as when I first started playing. Just as I''ve learned not to send battlecruisers out by themselves, for example, I''ve learned not to bunch them up. But sometimes I still do, and why and when is a matter of creativity, risk assessment, and strategy.

So far, CPUs (in most games) don''t seem to know how to do this.

--------------------
Just waiting for the mothership...

Share this post


Link to post
Share on other sites
hey, I gave credit to strategy games



Need help? Well, go FAQ yourself.
What a plight we who try to make a story-based game have...writers of conventional media have words, we have but binary numbers

Share this post


Link to post
Share on other sites
I think I can divide the players thinking into three categs

1) Supervised Artificial PI (SAPI): the designer designs what the player should think to go through the game. Quests, adventure games, part of RPG''s (the story, the quests) use this PI a lot. In a RTS type 1 PI would be learning what the buildings do, what the building/fighting rules are, etc. This leads to the player reverse engineering the game design rather than going through the mental process he was supposed to. (i.e. if you find a crate you ask yourself "why did the designer put it here", and start searching using for the crate like climbing). PI1 appears when the designer hides game information from the player; it''s his job to find them out; after he does find them out, going through the game again is trivial.
examples :

2) Unsupervised Game PI (UGPI): this is where the strategy games come in. After the player learns SAPI game rules quite fast (they are not classified, reading tutorials on these rules isn''t cheating), he starts combining these rules in different strategies. Though with some games you can go on for ages doing this and still coming up with new strategies, after awhile it gets boring, as you have tried them all, and found the few strategies wich are working best. After that point, you just keep going using those.
examples : all the games you mentioned are UGPI high powered; VGA Planets (fascinating game; try reading the strategy guides first, coz the game isnt really user friendly), Lemmings, TIM


3) Out of game PI (OOGPI) : it appears when the thought process''s target is not the game itself, but the person at the other end of the multiplayer network. I wont go into defining this one cause it''s not really done in a lot of games. However bits of OOGPI do from time to time rise to the surface. An trivial example of OOGPI was me making a joke in a msn zone hearts game (also has a chat window). All multiplayer games have some form of chat implemented and that is a window to OOGPI (especially in RPG''s). The problem with these is you cant really make a strong connection betwen the game and the OOGPI. Jokes & similar stuff require huge intelligence, beyond the wildest AI dreams, but they don''t have a continous direct effect on the game. Maybe in an RPG you form an alliance with some chaps, but after that is all RPG monster bashing again. Likewise in an RTS. Notice that these alliances are more of a direct consequence of UGPI (the game requires them to achieve some goals), so jokes remain, well, jokes.

My vision is a game where you have to talk alot, lie and find out if you''ve been lied to, threat people into submission or making other people friends, a game of intrigues and secret alliances. This kind of game will be very soon made possible by real-time audio&video images.
examples:
Learning from your spies that a neighbor you just signed an neutrality treaty with is mobilizing his army is UGPI, but telling him that one who would stab another in the back is a big fat rat is OOGPI.
A guy used the following "Terror tactics" in Total Annihilation : he packed a large number of very cheap bots, teamed them with an radar erase (wipes you off the enemie''s minimap), moved them close to the enemy lines, into the enemy radar range, made the bots move towards the enemy and turned off the anti radar. After few seconds, he turned his radar on again and retreated to safety. What the enemy was seeing on his radar was a huge battle group (many dots, cant tell if they are heavy tanks or peewees) allout attacking him out of nowhere. Dont know if this worked, but this is game connected OOGPI at it''s best.


I think that OOGPI has by far the greatest potential in the industry (after all, we all get tired of bragging about headshots or level 5000 characters whith 7941421634 zillions XP to our fellow gamers). Talking about how you''ve fooled some dwarves (player characters) into believing you have real magic powers, or how king playerX is just a corrupt politician and with him the hole of Azeroth will fall in orcish hands would really make gamers able to brag to people who never played the game (_Out Of Game_ PI), making both them and the non-gamers feel better (better inclined to buy the game that is).


What I want from you is fun examples of OOGPI that did happen or you''d like to happen. Perhaps a pattern would come out...

Share this post


Link to post
Share on other sites