I keep trying to build games where you can explore things. In the end, once I've built it, there won't be anything for me to explore because I had to build it in the first place. So I guess what I really want is for someone else to build a game with stuff for me to explore in it.
The coolest game(s) ever would be the one(s) with the ability for players to create their own content for other players to explore. Then everyone can explore - the players, and the creator. Assuming the player's aren't a bunch of dummyheads who only build garbage.
Too bad games like that are a major pain to create.
ID:182444
 Jun 26 2008, 1:18 pm
|
|
|
Jun 26 2008, 1:34 pm
|
|
A.T.H.K
 |
Why not build a random generator to create the areas on load then you can have something to explore and keep the players excitied about the constant new areas.
|
A.T.H.K wrote:
Why not build a random generator to create the areas on load then you can have something to explore and keep the players excitied about the constant new areas. In the end, I still made everything, even if I don't know what order it'll be generated in. I know what's there to discover already, so I never get to discover anything "new". Random generators will never give you anything novel, they only give you the same old things in random order. If you could design something where players could contribute to the database of things to generate, then it might be cool. But you'll still never get anything like what people could create on their own. For example, a Roguelike (using this as an illustration because they're usually run on randomly generated content) with randomly generated dungeons can never have dungeons of such quality as what a human could hand-craft on their own. Random generators can't come up with cool new ideas, concepts or clever tricks. Its like throwing a bunch of random sentences together to get a story versus someone actually writing a story. It just doesn't compare. | |
One of the best things you could do to combat all of the points you've mentioned, is to allow the players to actually code things in-game. This could be a coding language you invented (Foomer Code, etc.) that allows players enough functionality to create something unique if they wanted to. If it's in enough depth, the players could even create things that surprise even you.
| |
Then why don't you have a database setup so people could create lands/worlds/cities (whatever applies to your game) with an actually story?
| |
Also, I didn't mean anything advanced, it would be very simple stuff for people to understand and such.
| |
Popisfizzy wrote:
Then why don't you have a database setup so people could create lands/worlds/cities (whatever applies to your game) with an actually story? i like this idea. | |
Elation wrote:
Tom and Dan had this idea aaaaaages ago. Actually Elation, that's technically incorrect. Tom and Dan had this idea first. | |
I get where you're coming from. Even using a Roguelike method of procedurally generated content, you still effectively have the problem of being aware of the kinds of things the game can throw at you rather than being genuinely surprised by clever design.
Maybe the answer lies in a synthesis of the open design approach, where users create content, and procedural content. One of Sean Howard's game mechanics, #37, dealt with procedural templates--essentially, a language structure for procedural content. Perhaps if you could come up with a good procedural language and a way to check its output for good "game integrity", users could create pieces of content for the generator to use. One way to imagine this in use is a system that chains puzzles together, like Adventure or Yoda Stories (a very interesting kind of small-scale RPG). Players would contribute puzzles as entire sections of map or just drop-in sections. The puzzles might require certain items to solve, or a certain action, or might be straight-up trades where any number of items would be permissible for the puzzle itself but the engine picks one in particular it will need. Nonlinear puzzles could be created by having multiple solutions; some solutions might be used up (like an item to trade or burn) while others might remain open for other puzzles (like a key). Using something like this you can come up with a very interesting game, and it can be augmented by player contributions as long as they're vetted by the engine. In a Roguelike game, content could be player-expandable by allowing new types of spells to be created with an interpreted language. (I'm actually working on an idea for a freeform magic system that is more about discovering spells than authoring them.) New room types could be developed by players as well, which could be plugged into the engine. This is still of course a compromise. Chances are very high that unless your system was fairly bulletproof, players could simply toss in random badly designed foo that would have to be expunged from the list of acceptable modules or never let into the engine in the first place. So you'd know roughly what elements to expect even if you hadn't made them yourself. Shadowdarke avoids this somewhat with Darke Dungeon by (with the exception of maps, spells, some items) letting GMs but not players do the design work. That allows for some rather interesting explorability if he joins a world that someone has designed more or less from scratch. Still, to at least some extent the partial loss of surprise is a curse the game designer must bear--we often write games because we want to play something similar to what we imagine, but designed by an artist much more gifted than ourselves who can astound us with the result. Lummox JR | |
Well, I would argue that most contributed content would suck. Badly. And would probably be used as a way to create uber stuff, harass others, etc. Procedural content may not be novel, but neither is most human product, either. At least procedural content can be made consistently good.
Finding talented contributors is the exception, annoying parasites the rule. | |
I was assuming people could choose what content would be included. Obviously downloading everything that everyone makes would be a terrible idea.
| |
To see Jmurphs description in action, one only needs to look at the Spore creatures people are creating. I swear, I saw one that was a tribute to tub girl, and another consisting of what appeared to be two creatures, but was only clever design of a single creature, in the act of fellatio. I suppose if that's what you're going for, player-based content would work. I doubt that's what Foomer meant by exploration, though. XP
| |
I watched a talk on TED from some guy (can't recall his name right off the top of my head, but I can look it up later if you want) who described using website visitors to do work through games. Basically, he'd want something done, like categorizing videos or labeling image content based on the objects in the image (things computers are notoriously ill-equipped to handle), then he'd create an online multi-player (or in some cases single-player) game who's end result would be the workload, created as people play the game. It was a very fascinating concept, something I really haven't seen people attempt before. He gave techniques for ensuring the output was valid, usually by taking the sum of many people's output and merging it to get a more accurate result, or feeding the results back through the game as an AI player to verify it again. This kept people from really recking the data by inputing false information. He went into other methods as well, but I think you get the idea.
You could create a procedural system that can be modulated through player interaction. From the player's perspective, the interaction would encompass the entirety of the game, being fed into the core system and creating a vast database of content you can then pool at a later time, or pick at random to fill in new game content as needed. By not simply accepting a single input from a single player, but incorporating all input from all players into a larger whole, you should be able to ensure the quality of player content. But then again, this might go down that "too much trouble" path you want to avoid. | |
Yes, but from what I can tell, those ones are the vast minority. Most people are creating non-phallic, non-disgusting things.
| |
Sounds like you're assuming all of those will necessarily fit your game, its systems, environment or theme properly, or even have a place in your game.
Don't. | |
Sounds like you failed to read my other post. You should.
| |
Yeah, I already have, however unfitting-but-not-disgusting things are no exception in that they would be stopped from being included by a good approval system in place just like 'phallic' ones, so I was assuming your reply was counting a game without the burden of an approval system like suggested in your previous post, since otherwise what it said wouldn't really matter at all. My bad. =P
| |
Google is already doing something like this with their image labeler. I have to say the system is fairly inept in some ways because it only shows small images that are often hard to make out in any detail, and they can't guarantee you won't get partnered with a moron even when you happen to know for sure what the image is. The system is also prone to false matches, if two users think an image is something that it's not, but then that's a common danger because it's entirely possible for a minority to have the right answer and a vast majority to have the wrong one--like with urban myths.
I've seen a few videos from TED, all of which were very fascinating, though in this case I don't think the distributed workload idea would work for creative content. It would however be an interesting way to score the handiwork of an adaptive AI, like a procedural content system running on a neural net. I'm just not sure how you'd make game design something that would be easily scored in that case, though. Lummox JR | |