I think last night was the first night I actually got to get home, feed the cats, read the inbox and then just rip into UDS. As mentioned by the hamster reference previously, I'm this spot of trying to figure out what to compromise in design and keep getting things done. Random landscapes remain, but I reduced their frequency so that they would stop being so obstructive to the AI. Now that the bots more or less zerg right for the player, I need to get the playable levels ... well, playable. Not just testable. Like I said before, it's time to stop with a tech demo and try to make it a game.
The play world of Defense Squad has changed drastically - even since the decision to go turn-based. The first map was a very small block with three start points on either side. Friendly and enemy people spawned in randomly and the enemies never moved. Eventually when friendly pawns were movable, the enemy AI was added so that the "other" side will take a turn - but they would just stand and shoot. When movement finally happened, I need obstacles. I already had some random junk generators from the FPS version, so I first added in a random landscape generator. Then I modified the existing "junk" generators to place trees and rocks around. The first few passes were pretty laughable. At one point the trees were so dense that it was impossible to move and for the longest time rocks stubbornly floated in the air.
Now it's smooth but rolling landscape with the occasional tree and rock for coverage. Last night I added the ability for the level to randomly determine a starting spot for both the players and the enemies. Once I have a few bugs in that ironed out (like people being created inside a hill, and then falling to the empty world below the landscape) - I'll expand the size of the level and create random building and larger obstructions.
One might ask why creating this largely random world is such a high priority. It might be easy to assume that making all the weapons work, or figuring out the skill system, might be more vital to the experience than the potential placement of a stump.
One should first take a quick break to read GameDevBlog's take on variance in games. There's little there I can disagree with, but I would also add that variance in games gives a greater rise to replayability. It's a trend games have largely moved away from - outside a limited since in the RTS genre. The FPS genre relies on modmakers and mapmakers to elongate replayability. This is essentially variance via quantity. Variance via randomness, though, can force players to stop assuming what their game world will throw at them. Classic games like Elite or Seven Cities of Gold were excellent examples of this - as well as my favorite - X-Com.
So before I'm worried about how a flak cannon might be appropiately fired by an enemy pawn (my guess - look up, press trigger) ... my main concern is making a simple, but varied game environment which allows for players to have fun in ... regardless of what their pawns might have in their pack.
Tuesday, February 22, 2005
A Rundown of a Random World
Subscribe to:
Post Comments (Atom)
No comments:
Post a Comment