How married are you to perfect mazes? From a gameplay perspective, they're not that interesting. When you come to a crossroads, the results are either going to be:
1) You pick the wrong path, and hit a dead end, possibly having to backtrack a long way, wasting turns and getting no rewards because you already got them.
2) You pick the right path.
In the end there's only one "right" answer, and if you pick a wrong one you just backtrack and pick the next one. There's a little bit of strategy involved with which path you take, but picking the right path isn't much of an achievement, it's inevitable if you keep going at it.
If you're not 100% dedicated to perfect mazes, one thing that I think makes mazes a lot more interesting is tearing down a small number of "walls" within the maze. Either random, or specifically at dead ends, if you want to eliminate dead ends. This potentially means there are multiple right paths, and some are longer or shorter than the others. It gives more of a sense of exploration, makes the level feel more like a "real" place. It also makes it feel like your choices matter, because although you were guaranteed to eventually pick the right path, which path you take matters because you might pick a shorter right path!
If you don't want to do that, an incentive to cushion the blow of coming on a dead end might be guaranteeing treasure (or a special fight...hall minotaur?) at dead ends. Other ideas I had were a special "100% completion" bonus for the maze, or multiple mini-goal spots within the maze, but I don't like those as much. In PBN, I made the entrance and goal spots to the maze random (and the goal spot invisible until you reach it), but I don't think that's a great idea here.
Anyway, don't know what your philosophy behind the mazes is, so maybe none of that makes sense, but just some ideas.
hmmn, not super-married to them. The space maze has random dropped walls (except around planets, where they're cleared completely for 2 sectors), so I suppose I could do a few random dropped walls in mazes, but I'm not sure I could guarantee a lack of dead-ends - the wall drops are dice rolls in the spacegen code, and I usually have to run it a couple times to get a universe I like.
Maybe I'll fiddle with it in awhile and see what some settings look like in testing. I'll split this off to a new thread for that discussion..
Another idea: have the intermediate rewards (items that drop randomly from a newly explored square) scale up with number of already explored squares. So if I "waste" turns exploring dead ends, at least I will get something out of it later.
(rationale: 1) I wouldn't want to write the code to detect dead ends b) don't let players increase rewards arbitrarily by going in circles iii) don't give lower rewards-per-turns just because I had to spend extra turns backtracking)
You could guarantee no dead ends by iterating through all the "rooms" in the maze after it's made, and if it has only one exit, adding a random exit that's not already open. That might make it too open, though.
How about making each non-solution-corridor only go a maximum of four spaces before it ends. There's nothing worse than going 22 spaces with dead ends here and there before finding out it ends close to the destination square without getting there. Of course, that takes away the usefulness of the Genie garage door opener thingy ;)
That is one consideration that should be discussed. Any major changes to mazes would basically render an IOTM irrelevant. maybe nobody would mind - you guys think you got your 5 or 10 peens worth out of it yet?
Mazes in RPG have always been frustrating. That has always been the point. I remember playing Might & Magic and getting lost in the forest before I realized there might be a pattern.
However, I can think of two compromises:
Make mazes smaller - current 250 cell rectangle could be reduced to a more manageable number... maybe 125 *shrug*
Make movement cost no turns, only debit players 1 turn for encounters
The nice thing is that mazes are generated "per-day-per-player" so the opportunity exists to try a couple of things when the cycle finishes, with no particular worry that we can't dial it back the next day.
Ads and I are poking at the degree quests now, formulating strategies for some potential re-writes. Thus, we hope to have a few things ready to rip through in the next Anarchy period, and there's a good chance we'll at least try some new stuff with the mazes then and see what people think. At that point, I'd love to palaver with Rickton to discuss various mechanical solutions. I'm certain there's some stuff I can learn there. :D
smaller mazes would be cool, and the idea of some walls taken down would be awesome. The mazes that Rickton had in PBN were much smaller but even with some walls taken down you still needed to search around a bit. kudos to you guys for all the work you've been doing!
I dunno if frustration for frustration's sake is a great design goal. If something's frustrating as a side-effect that's one thing, but it doesn't seem like a very good "point."
I personally would like smaller mazes, but that'd mean less encounters for people who want to farm in them (although since you're moving against farming that might be what you want).
Making mazes cost no turns might make them too good, because you'd be getting several free good drops per maze.
Sure. I don't claim to be an expert, but I've played around with some different ideas. I dunno if you remember or not, but you're actually the one who gave me the maze code in the first place!
Yeah, I think PBN mazes worked really well for PBN, but the games are different enough they wouldn't be a great idea here. PBN mazes didn't have daily rewards of valuable items for completing a maze, plus you also had to return to an exit to leave.