Temple of The Roguelike Forums
Development => Design => Topic started by: mushroom patch on June 17, 2016, 05:41:54 PM
-
Friends,
I come to speak with you about corridors, you know, this kind of shit:
### ##
### ##
### ##
##
######
I see so much discussion of corridors, how to generate them, and so on. The answer in regards to corridors is simple: They are bad, don't generate them.
There are many alternatives to corridors, so I will not bore you with an exhaustive account. If you are still writing a roguelike game that uses corridors, though, I suggest to you that it is time to rethink and reevaluate.
-
Is this real life?
-
No, corridors are fine in real life. I only mean in roguelike games.
-
So how do you connect rooms? Do you think wider corridors are okay?
-
This is a good question. Wider corridors are better, but unfortunately they are still bad.
-
I had to use increasingly wider corridors in DUMUZID, so the player can more easily fit in the later levels.
The most basic roguelike tactic is to use corridors to your advantage. It's perhaps too good of a strategy, so it's nice when the higher level monsters have some way to disrupt your plans in corridors. Monsters can be made to destroy walls or to summon monsters that surround you on the opposite side.
I have to assume this is a troll thread but even so it's still an interesting topic.
-
Still better if all monsters, big and small, are able to avoid corridor cheez through nonexistence of corridors.
-
You should let the RNG decide wideness of corridors... Nothing will go wrong right?
-
This is a good question. Wider corridors are better, but unfortunately they are still bad.
Before I threw out the dungeon gen system on my project, I had corridors generate like long rectangular rooms.
What exactly is the reason not to have corridors?
-
Why are corridors bad?
They are overly uniform. If full exploration is important in a game, corridors offer required, one-dimensional exploration. This is the sort of tedium autoexplore features are designed to remove, but they generally result in worse play than manual exploration so that if you want to win consistently or you want to win fast in turn count, you're stuck traversing them at least semi-manually.
They lend themselves to cheezy combat tactics around which games tend to be balanced, meaning that the player has to use them or lose. Those tactics involve luring monsters in every nontrivial encounter, e.g. whenever there's more than one monster after you, sometimes a long way.
Baking the assumptions of corridors, with their abundance of chokepoints and all the rest, into your game limits the tactical situation right out of the gate. Don't do it.
-
Breaking a game down into more abstract parts corridors are a tool to deal with certain situations. If you have more interesting (and therefore better in terms of gameplay) ways to deal with a given situation, then its true, corridors are a redundant and antiquated part of game play.
I have often thought RLs in essence are just a series of encounters and a bunch of actions and re-actions. I have pondered on a game which has one room at any given time. You deal with the room and you can then progress to the next room. Obviously rooms can be much more intricate and detailed than a normal RL and therefore open up to more tactical situations.
That being said, because this is a troll thread, I like RLs because I enjoy the dungeon layout and exploration aspect. You just never know what will be around that next bend. So long live the corridor (just keep them short ;D).
-
Well, Angband has corridors, but they're slightly less subject to abuse because the monsters have AI. This is how I plan to solve the corridors problem - have the monsters refuse to fight you single file in a corridor.
-
Angband has some of the worst dungeon generation in the genre. Huge levels with corridors that stretch all the way across them. On the other hand, it has so many other broken mechanics, e.g. more or less unlimited consumables that allow you to reshape the dungeon, instantly teleport off a level, instantly kill any monster, stealth that renders the vast majority of monsters entirely immobile, brain dead AI etc., that just starting to address the problems of angband doesn't even get you to thinking about layouts.
I guarantee you will not solve the problems of corridors through changes to monster behavior.
-
"Save scumming considered harmful.
Don't allow saving in your roguelikes because this can be abused by save scummers."
Compare the above with the following:
"Corridors abuse is considered harmful.
Don't allow corridors in your roguelikes because they can be abused by corridor campers."
If the pinch points are exploited it's not the pinch point's fault. Eliminating corridors won't remove pinch points, just make them less prevalent, and they'll still be abused by abusers.
Likewise, if the gameplay breaks in the presence of corridors it's not the corridor's fault. Gameplay needn't break due to corridors. A more intelligent adversary can exploit pinch points as well or better than most players can. Dumb AI isn't the fault of corridors.
Bad dungeon gen that has shitty corridors doesn't mean all corridors are evil.
Translation: Don't do bad dungeon gen, and also don't use shitty AI.
-
"Save scumming considered harmful.
Don't allow saving in your roguelikes because this can be abused by save scummers."
Bro, if this is your idea of wit, I got bad news... But I'll tell you, it's true: Any local game can be cheated, savefile or no. This is why roguelike games are best played on a secure third party system.
If the pinch points are exploited it's not the pinch point's fault. Eliminating corridors won't remove pinch points, just make them less prevalent, and they'll still be abused by abusers.
This is pretty obviously false. There is no requirement that "pinch points" exist anywhere. In any case, the blame for any problems with choke points in a game lies with the game's designer.
Likewise, if the gameplay breaks in the presence of corridors it's not the corridor's fault. Gameplay needn't break due to corridors. A more intelligent adversary can exploit pinch points as well or better than most players can. Dumb AI isn't the fault of corridors.
Bad dungeon gen that has shitty corridors doesn't mean all corridors are evil.
Translation: Don't do bad dungeon gen, and also don't use shitty AI.
So your idea is that with some kind of magical AI algorithm, the basic asymmetry between a group of many monsters and a single player character with respect to choke points can be resolved? Please elaborate on that.
It seems to me that as a practical matter, you want your AI to be as dumb as it can get away with being. If not generating certain features that are problematic for various reasons helps you do more with less, this is a good thing.
-
I would play a roguelike where you tackled one room and then got whisked off to a new one, no corridors. It's all about what game mechanical action they serve. If the only thing they do is carry the player from one challenge to the next with there only choice being which door you kick open, that can be handled with a dialog box and an ASCII cut scene. What other purpose could they serve though? I'm not asking rhetorically, what are some purposes beyond "transition from room to room that demonstrates that these rooms stand in some relation to each other geographically"?
-
Why the fuck would you remove corridors, it's just stupid.
-
"Save scumming considered harmful.
Don't allow saving in your roguelikes because this can be abused by save scummers."
That comparison doesn't hold, though. Save scumming is cheating, whilst corridor fighting is a viable (and in fact the optimal) strategy in most RLs. So in a classic rooms+corridors dungeon, you need to balance the game around the assumption that the player will draw back whenever possible.
Not to say that corridors are bad, or that it's trivial to come up with optional map generation concepts. I still think it's interesting to create a game that focuses more on, say, open spaces with differing sizes and layouts. You can have the occasional classic maze in a game that mainly takes place outdoors or in houses. That might even feel fresh, if fighting is designed around something other than trading melee attacks in a hallway. Then there's the option of wide corridors, which would offset the effect of corridor fighting a bit. One could also put in areas with "soft" walls, eg. that block sight but not movement. Trodden paths lead between points of interest, but the player can still elect to wade through the high grass (or whatever your "walls" represent), and can never be sure that some foe won't just jump out of nowhere (Legerdemain did this with corn fields, and I'm definitely planning something along those lines for Land of Strangers).
As always,
Minotauros
-
I would play a roguelike where you tackled one room and then got whisked off to a new one, no corridors. It's all about what game mechanical action they serve. If the only thing they do is carry the player from one challenge to the next with there only choice being which door you kick open, that can be handled with a dialog box and an ASCII cut scene. What other purpose could they serve though? I'm not asking rhetorically, what are some purposes beyond "transition from room to room that demonstrates that these rooms stand in some relation to each other geographically"?
That interpretation is too simplistic. It might work for Rogue or the first few Nethack levels, but more advanced terrain gen exists, you know.
Rather, the challenges that you are whisked between are the floors, and room/corridor/cave/city/forest structure is just a terrain feature in that challenge.
That said, the main purpose of a corridor is generally a chokepoint area. Complicate gameplay by restricting space.
And mushroom patch's argument is that chokepoints help the player too much and force reliance on their presence in order to adequately deal with enemies.
-
That interpretation is too simplistic. It might work for Rogue or the first few Nethack levels, but more advanced terrain gen exists, you know.
Right, I understand that advanced terrain gen exists. If you'll read what I said again, I didn't say that corridors don't serve a useful purpose, I asked what purposes they serve. Here are two fairly obvious purposes that they serve.
1) They are the route through which players and monsters move from room to room within a dungeon level.
2) They demonstrate the geographic relation of the rooms within a dungeon level.
I can think of at least 2 more:
3) They support/develop the exploration aspect of Roguelikes.
4) They provide safe(ish) places for players to recover health in roguelikes with a resting recovery mechanic.
Rather, the challenges that you are whisked between are the floors, and room/corridor/cave/city/forest structure is just a terrain feature in that challenge.
Yes, but why? Why couldn't you create largish, complicated rooms and have them move from room to room via a dialogue box? I'm using the term 'room' loosely here. What I really mean is a compact, contiguous segment of dungeon. Compact here means no long corridors.
That said, the main purpose of a corridor is generally a chokepoint area. Complicate gameplay by restricting space.
And mushroom patch's argument is that chokepoints help the player too much and force reliance on their presence in order to adequately deal with enemies.
That's what I was asking.
5)The provide chokepoints which players can use tactically to control conflicts.
Are there any other things they contribute? Because the only thing listed in this post that looks hard to implement using just rooms is exploration, and that's largely down to room design.
-
"Save scumming considered harmful.
Don't allow saving in your roguelikes because this can be abused by save scummers."
That comparison doesn't hold, though. Save scumming is cheating, whilst corridor fighting is a viable (and in fact the optimal) strategy in most RLs. So in a classic rooms+corridors dungeon, you need to balance the game around the assumption that the player will draw back whenever possible.
So, you agree... The comparison was not meant to hold. It was an extreme example meant to illuminate the exact conclusion you've drawn. I felt my point was clear: In a classic roguelike "you need to balance the game" assuming that valid player actions such as "draw back whenever possible" don't break the game, if that means smarter AI, so be it.
Thus, if you assume that saving the game is not a legitimate option and design a system that has resuming from save breaking the game then you've badly designed the game. This doesn't indicate that saving itself is bad as a feature at all, which is what the comparison was for: To have you conclude, "That's preposterous, saving the game is a legit feature; reloading doesn't make a game crappy but other issues might". And then go on to conclude, "corridors are a legit feature, something else must be the problem rather than simply having corridors themselves". Broken game designs are not the result of the use of corridors, but the failure to design gameplay which addresses those who would obsessively exploit corridors. The core point is that features such as saving or having corridors between rooms are not inherently shit, and that there are other factors such as faulty mechanics or exploitive behaviors at play.
Even considering a primarily cave-formed or room adjoined generation system, the exclusion of all corridors would be an oddity. Corridors can be effective at creating a precise scenario that should be available to the (procedural) level designer, otherwise if one does appear your game might break. Not focusing on such pinch points doesn't mean they won't exist. Corridors themselves are not "considered harmful", but games which suck due to poor consideration of gamplay in corridors do exist. Likewise, saving a game is not considered harmful, and games should not break on save/load (even if via external VM) but rather take consideration that saving and loading behavior does exist (just as corridor use exists).
-
Ugh, I'm not going to split any hairs over your roundabout argument. Although I think my point still stands ;) Anyway, I wholeheartedly agree that corridors have their place in the genre as a cool feature, though I wouldn't mind if more games deviated from tried and true hack'n'back tactics.
As always,
Minotauros
-
I'm in the 'better AI' camp.
Corridors are essential in the generic dungeon theme. Dungeons are literally rooms and corridors in various proportions. If you remove the corridors, it is not dungeon anymore but something else (ruins, garden, town, etc.). Obviously there is nothing wrong with it, you can always have a game without dungeons. But when you do have dungeons, corridors are not at fault, no more than rooms.
Are corridors abused? Do they give the player too big of advantage? Probably you are doing it wrong. Try to imagine yourself in a similar situation. Would you really flee to the fairly narrow passage in case of danger? No, that would be one of the most dangerous places instead, badly illuminated space without freedom of movement. In the best case scenario your foe will walk away and you will waste time and supplies. If they will circle around and surround, you're screwed. If they barricade the exits and throw/breathe something in, you're screwed. If they lay some trap-like mech at the exit, then you're now trapped in and, well, screwed. One of them may go and bring a huge spear you can't do much about. Call the archer or even the caster. From the time you went into the corridor you've given up the initiative. Nope, an open, well-lit place with obstacles is the best from a tactical perspective.
If something is abused, it is because it is not balanced. Add a reliable gun to the primarily sword-fighting game and everyone will use the gun. Same with corridors, players expected to abuse them if there are no disadvantages. You just need to either give dungeons up or fix the source of the problem, depending on your priorities.
-
If any of you are looking for a roguelike with no corridors, Ananias (http://ananiasgame.com/) does this, which also fits with it being designed for mobile.
-
Are corridors abused? Do they give the player too big of advantage? Probably you are doing it wrong.
Indeed, you are doing it wrong. You're generating corridors.
It is interesting that you bring up the concept of a dungeon as a fixed kind of thing. Of course, the notion of dungeon one sees in most roguelike games comes directly from dungeons and dragons and the dungeon generation algorithms described in their various rule books. I don't think there was any prior conception of a "dungeon" being a thing with characteristics lending themselves to exploration or good turn based combat. Indeed, I doubt the geometric considerations that become important in evaluating roguelike game mechanics were very carefully considered by the designers of the dungeons and dragons algorithms, since the dungeon map is often more of a guideline than a hard fast rule in dungeons and dragons combat tactics -- many players play without a physical board and map, for example.
Turn by turn movement and combat mechanics make the details of the map's geometry far more important than they are in dungeons and dragons. At the same time, the kind of dungeons favored in dungeons and dragons are in no way representative of a real life concept of a dungeon (they are much more like catacombs or underground cities) nor do they offer a particularly compelling model for granular combat and movement. This is just another case where early roguelikes have uncritically imported dungeons and dragons mechanics. This should not bind us to a particular concept of what a dungeon is. Similarly, we should not uncritically accept corridors as a necessity.
As an additional point against corridors, they are a poor use of space. This is related to the issue of being one-dimensional and boring to explore, but also independent of it. What is the best way to use map space in a roguelike game? Corridors as they are usually implemented involve a lot of negative space -- inaccessible wall interiors with few or no interesting features. If you ask what is the appropriate density of map features, I think you'll inevitably conclude that corridors will consistently fail to deliver. Corridors merely inflate your map without providing significant value. To name a specific example, Angband, again, is a serious offender here. Huge maps with long corridors, tons of empty space that is not accessible to the player (outside of the tedious but sometimes necessary business of digging). What's going on here? Well, moria, the predecessor of angband, started with someone's neato dungeon generator that made levels that are big and have complex corridors compared to rogue. This must be good, right? Well, no, but that's the thinking.
-
Even though I love corridors, here's a good example of space-efficient, kinda corridor-less dungeons:
http://worldofortix.blogspot.ch/2012/02/mordor-11-maps.html
-
It is interesting that you bring up the concept of a dungeon as a fixed kind of thing.
More like 'established' kind of thing. Dungeon is rooms plus passages between them. This is close to what dungeons would be IRL (or believably close to that). Without corridors you end up with paper-thin walls like in reaver's example: looks nice but I'll argue it is not a dungeon. Basement of a castle or level of a magically constructed tower maybe?
This becomes a question of terminology. In the end, there are a whole lot of different map/level designs, with and without corridors. There is nothing wrong with using ones without. But corridors are an element of established design (usually going by the name 'dungeon'). It's just some designs have features you'll have to be careful with. Long and boring paths in a maze. Pillars in an abandoned temple you can dance around. Annoying traps in a tomb. I'm sure there are a lot of things one may fail to balance besides corridors.
As an additional point against corridors, they are a poor use of space.
I'll agree on the point that poorly made corridors (long and featureless) provide a questionable element of design. A lone path going across half of the map is just a failure. It might be a good enough reason to throw that generator away.
-
There's no doubt there are other common design flaws in roguelike games. Many established designs consistently produce poor results. This is just the one I am interested in at the moment.
-
That interpretation is too simplistic. It might work for Rogue or the first few Nethack levels, but more advanced terrain gen exists, you know.
If you'll read what I said again, I didn't say that corridors don't serve a useful purpose, I asked what purposes they serve. Here are two fairly obvious purposes that they serve.
1) They are the route through which players and monsters move from room to room within a dungeon level.
2) They demonstrate the geographic relation of the rooms within a dungeon level.
When you said "If the only thing they do is carry the player from one challenge to the next with there only choice being which door you kick open, that can be handled with a dialog box and an ASCII cut scene", I replied with the first sentence in my last post, implying that I believe this is a false sentiment. If that wasn't clear.
"What other purpose could they serve though?" was answered by the third sentence in my last post, which stated that corridors are tactical chokepoints and terrain features.
Here are uses of corridors that can't be replicated with a dialog & cutscene:
+ A +
+ +
+ +
+ +
+ +
+ +
++++++++ +++++++++
A A
++++++++ +++++++++
+ +
+ +
+ +
+ +
+ +
+ A+
You can have two, three, or four intersecting corridors which connect two, three, or four rooms. The effect? The corridor becomes a chokepoint at any of the points labelled 'A', but loses its value as you venture towards the intersection. A player standing at the intersection no longer has the advantage of fighting against a maximum of two enemies, he can now be surrounded and besieged as if the area were a wide open space.
In fact, at the intersection of two corridors, you can be attacked from four sides without the possibility to escape diagonally. This means that the center of this structure is actually more dangerous than an open space; the tight one-square corridors instantly reverse their advantages.
You can line corridors with traps, too. Any intersection can be given a higher percent chance to spawn with a bear trap or monster-spawning trap. Then you could also give the player an alternative route,perhaps an alternative wider corridor, to cause a decision to be made.
How about this:
+ A +
+ +
+ +
+ +
+ +
+ + +++++++++++++++++
++++++++ ++++++++++++ +++
A A B +
++++++++ ++++++++++++ +
+ + + +
+ + + +
+ + + +
+ + + +
+ + + +
++++++ ++++++++++ +
+ +
+ +
+ +
+ C +
+++++++++++++++++++++++++++++++
In this case, the corridor is more of a tactical decision. Say you are at point C, wanting to move towards point B. Do you take the open-space route where you can be attacked by monsters without a chokepoint, or do you drag monsters into the corridor, another potentially risky move? Maybe the monsters can push you back one tile, which in this case would shove you down the corridor into the intersection and possibly into any traps. In an open-space area, the push-back might be less dangerous.
Are there any other things they contribute? Because the only thing listed in this post that looks hard to implement using just rooms is exploration, and that's largely down to room design.
But the real question is: what is your definition of a corridor? How does one use "just rooms"? If you implement a chokepoint, at what point is it just a narrowing of the area, and at what point is it a separate concept?
-
Without getting into the specifics of your analysis of different spaces and routes, which sounds a little fishy to me, the problem with this kind of analysis is that the player will generally have a lot of power to decide where a fight happens, so the fact that some squares in your system of corridors are marginally worse than others becomes meaningless -- the player just elects not to stand there. It is too obvious to be an interesting decision. You have to look at what the best thing for the player to do is and what follows from that.
Also worth noting that you typically empty a level of its monsters as you explore in a roguelike game, which means that corridors tend to acquire a preferred direction when you first explore them. You tend to meet new threats from one direction only. As for open space, as long as there's a corridor remotely nearby, the possibility of being surrounded does not exist without special mechanics to make it happen (monster summons or traps, for example). If there is more than one monster, you can just take them back to the corridor.
-
I take the point, and agree that games where corridors are always the optimal place to fight are boring, but a game where corridors are never the optimal place to fight because there are no corridors would also be pretty dull. And where do you go from there? Fighting with your back to a wall is always optimal so you take out all the walls? I would say that a lot of roguelikes would be improved by having significantly less corridors, but I wouldn't get rid of them entirely.
In the same way as everything else in a tactical game, you ideally want to make it so that fighting in a corridor is sometimes optimal and sometimes not. Fortunately, there are an absolute shit-tonne of mechanics you can employ to this end, some of which are so common as to be almost ubiquitous in Roguelikes anyway. Such as:
- The risk of getting trapped between two enemies
- Limited time to kite enemies back to corridors (hunger clock)
- Limited vision in corridors (darkness, corners)
- Enemies difficult to kite (ranged enemies, fast enemies, teleporting enemies, smart enemies, fleeing thieves)
- Area-of-effect abilities (spells, Brogue-like weapons)
- Bouncing spells (magic missile)
- Allies
- Different terrain types (water, high ground)
- Mana-/Stamina-limited abilities
- Reflecting explosions
- Dodging mechanics that require an empty space
- Barriers to retreating indefinitely (closing doors)
- And Many More!
Something I've not seen any roguelikes do but that might be quite cool would be to model the way that in Dark Souls and/or real life, you don't want to fight in enclosed spaces with certain weapons because of the difficulty of getting your swing on. An accuracy modifier based on your weapon and the number of wall tiles you're next to would be a pretty straightforward way of doing it, and may make the choice of where to fight a bit more interesting.
-
I take the point, and agree that games where corridors are always the optimal place to fight are boring,
This describes a lot of roguelikes, particularly the older and more successful ones.
but a game where corridors are never the optimal place to fight because there are no corridors would also be pretty dull.
This is a leap to put it mildly and almost certainly false.
In the same way as everything else in a tactical game, you ideally want to make it so that fighting in a corridor is sometimes optimal and sometimes not. Fortunately, there are an absolute shit-tonne of mechanics you can employ to this end, some of which are so common as to be almost ubiquitous in Roguelikes anyway. Such as:
[long list]
You get that half the items in this list either make the situation worse (allies, special terrain) or already are in a lot of the classics and don't change the situation, as addressed elsewhere in thread, right?
Yes, you can come up with complex rules that will not be evident through the interface without explicit tutorials (e.g. dodging and hit rolls affected by number of adjacent walls, "reflecting explosions") and that a typical player will not grasp through simply playing the game just so that you can have a low-density, one-dimensional feature in your dungeons. Or, you know, you could just generate your maps differently.
-
Is there any chance you could write a long essay detailing this confusing notion about corridors?
As it is, it's like that one guy who makes a post alleging some unsupported contentious thing, and then he pops back in to make further unsupported contentious confusications over anyone's responses that try and make sense of it.
Insert picture of someone urinating into the global air flow which happens to be flowing their way at the time, and any time they happen to be standing in that location.
-
I don't know, I usually need a whinier request than this to screw up the motivation to write a careful exposition.
-
Most of the thread boils down to "Guys corridors are bad; trust me I'm right and also all the classic roguelikes are shit." though.
Please offer an alternative.
I'll agree that corridors are almost always the worst element of a map, right up there with the porous cave bullshit often seen in certain *bands.
[edit: most of what I said was already covered, snipetty snip]
Game design aside, aren't corridors more of a by-product of room generation? Certainly in Rogue, if you want to fit 9 (variously shaped) rooms on the screen in a 3x3 grid, you're going to end up with a lot of connecting corridors.
So your options boil down to what?
Generate rooms connected by doorways? People'd just fight in the doorway, so that's straight out.
Generate big rooms connected by 2-3 tile wide "tunnels" (totally NOT corridors!) that curve around? That might help a little...
I don't think generating one big room is the answer. As has been said, people will just fight with their back against a wall (or get swarmed and die). I wouldn't want to play a game where I go down the stairs and find myself standing in a wide open space facing 20 enemies AGAIN and AGAIN.
The core problem might be that the player occupies 1 tile and doesn't require any external tiles to be unoccupied in order to swing his weapon and bring the hurt.
http://www.roguebasin.com/index.php?title=Swift_Swurd
The above tried to do position-based fencing. Judge for yourself how fun/successfull it was.
-
^Swift Swurd looks really cool. I've heard of it before but never tried it. Maybe I'll try later this week.
-
Most of the thread boils down to "Guys corridors are bad; trust me I'm right and also all the classic roguelikes are shit." though.
Please offer an alternative.
I would remind you that the typical level of engagement in this thread has been along the lines of "What?? If you don't have corridors, you won't have any walls at all!!!" See your own post, for example. Like, people are genuinely baffled by the concept of a map in a roguelike game that doesn't have corridors in spite presumably having lived in a society filled with top view, 2-dimensional video game settings for decades. The level of imagination there doesn't inspire one to make a careful exposition accessible to all comers.
I'll agree that corridors are almost always the worst element of a map, right up there with the porous cave bullshit often seen in certain *bands.
[edit: most of what I said was already covered, snipetty snip]
I'm glad we agree.
Game design aside, aren't corridors more of a by-product of room generation? Certainly in Rogue, if you want to fit 9 (variously shaped) rooms on the screen in a 3x3 grid, you're going to end up with a lot of connecting corridors.
No, they are a by-product of a fairly literal-minded transcription of dungeon generation routines from dungeons and dragons. There are many ways a collection of rooms can hang around in an ambient 2-dimensional space. The particular choices made in rogue reflect what it says in the dungeon master's guides of the time.
-
Most of the thread boils down to "Guys corridors are bad; trust me I'm right and also all the classic roguelikes are shit." though.
Please offer an alternative.
I would remind you that the typical level of engagement in this thread has been along the lines of "What?? If you don't have corridors, you won't have any walls at all!!!"
Nah, that post here seems very much more comprehensive:
I take the point, and agree that games where corridors are always the optimal place to fight are boring, but a game where corridors are never the optimal place to fight because there are no corridors would also be pretty dull. And where do you go from there? Fighting with your back to a wall is always optimal so you take out all the walls? I would say that a lot of roguelikes would be improved by having significantly less corridors, but I wouldn't get rid of them entirely.
In the same way as everything else in a tactical game, you ideally want to make it so that fighting in a corridor is sometimes optimal and sometimes not. Fortunately, there are an absolute shit-tonne of mechanics you can employ to this end, some of which are so common as to be almost ubiquitous in Roguelikes anyway. Such as:
- The risk of getting trapped between two enemies
- Limited time to kite enemies back to corridors (hunger clock)
- Limited vision in corridors (darkness, corners)
- Enemies difficult to kite (ranged enemies, fast enemies, teleporting enemies, smart enemies, fleeing thieves)
- Area-of-effect abilities (spells, Brogue-like weapons)
- Bouncing spells (magic missile)
- Allies
- Different terrain types (water, high ground)
- Mana-/Stamina-limited abilities
- Reflecting explosions
- Dodging mechanics that require an empty space
- Barriers to retreating indefinitely (closing doors)
- And Many More!
Something I've not seen any roguelikes do but that might be quite cool would be to model the way that in Dark Souls and/or real life, you don't want to fight in enclosed spaces with certain weapons because of the difficulty of getting your swing on. An accuracy modifier based on your weapon and the number of wall tiles you're next to would be a pretty straightforward way of doing it, and may make the choice of where to fight a bit more interesting.
But you answered that with such an infuriatingly dismissive attitude that the impression is that you just want to rave about some pet peeve instead of making a coherent point.
-
No, they are a by-product of a fairly literal-minded transcription of dungeon generation routines from dungeons and dragons. There are many ways a collection of rooms can hang around in an ambient 2-dimensional space. The particular choices made in rogue reflect what it says in the dungeon master's guides of the time.
I'm not seeing the alternatives (then again, I'm also not very imaginative). There's already been mention of wider corridors as a semi-viable option.
If you just use ONLY doors like in this below monstrosity I just made, you're creating an "optimal fight place" IN the doorways, which is the tactical equivalent of fighting in hallways.
http://imgur.com/VR9pkIB (http://imgur.com/VR9pkIB)
Are you saying that that is STILL preferable to having hallways? It's true that you have more real estate now for "rooms" and the tactical advantage is slightly less compared to a hallway, but you're still soft-forcing the player to fight in a certain location.
Wouldn't the better solution be to give any entity inside of a hallway an across-the-board speed penalty (because of the cramped quarters)? Seeing as how ranged attackers outside of the hallway would get more attacks and the "faster" ticking of the food timer, you'd have given a sufficient trade-off for the safety you get from the tunnel.
The thing with maps from dungeon crawlers like Eye of the Beholder or what have you is that walls don't take up space.
In a roguelike you're dealing with the following
..#..
..#..
......
###
(^ 3x4 grid, yes?)
but in a dungeon crawler that very same architecture would be
..|..
......
___
which is in fact JUST a 2x2 grid, with some squares being non-traversable in certain directions.
If there was a way to leverage THAT in a roguelike, then you'd REALLY open up options for good/creative map generation.
-
Regarding your picture, if the doors are more than one square wide and all of their tiles open together, then that's not terrible. The main bad thing is being able to body block monsters into collecting such that you fight only one at a time. Fighting a group of monsters one at a time should require retreating over explored territory to maneuver them around corners and so on. It is important for combat tactics to interact with dungeon features and map space.
Tight doorways have a problem of funneling groups of monsters, limiting their flow in too facile a way. It's not as extreme as corridors, but I am not convinced about doors either.
As for what I actually favor, amorphous cavern formations with soft walls (vegetation, fog, etc.) as described by AgingMinotaur seems to me to be the most compelling model for connecting rooms.
-
I have to refer back to Angband again in that while it's a broken game which only works by piling broken mechanic on top of broken mechanic, it's still pretty difficult to corridors scum wolves and especially spiders.
I will concede the point that corridors are bad design, but their disadvantages can be mitigated.
-
The thing about angband is that for a significant part of the game, dangerous monsters move way faster than you and for the rest of the game the main danger is being summoned or breathed on. The solution to the second problem is to build zigzag corridors and fight one monster at a time or just use stealth to avoid the fight entirely. In both cases, you just don't move in combat. This is terrible at such an early stage of the analysis that it doesn't even matter what the dungeon layouts are like from a tactical perspective. There are no real tactics in angband anyway. You just have to understand how the consumables work.
That raises a related point: If generating corridors is bad, letting players make corridors is even worse.
-
Is it possible to consider the behaviour of certain pack monsters (notably wolves and spiders) independently of the other idiosyncracies of the game though? I mention it because in other roguelikes I've played enemies tend to just blindly follow you down corridors single file.
-
Oh right, angband has that behavior where animal packs hang around in open areas as though you'll come out to get them. I haven't played angband in a while, but as I recall there's an easy way to deal with this: The monsters only stay at the mouth of the hallway if you're within a certain range, otherwise they approach. So you retreat farther down the corridor and they'll come after you. You can then go back at them and fight them one at a time. I don't remember whether they retreat again, but I think not.
If they're hounds, you just arrange to meet them at a corner or better one of angband's silly turnback corridor formations.
This is a case that should add to suspicions about monster AI overcoming dungeon geometry. Geometry wins every time.
edit: Okay, I went ahead and tried this. In the version I tried, I saw this pack AI with jackals. When you retreat to a corridor, they just hang around in the room they started in. Maybe that would work if you had a game where it was desirable for monsters to guard places. That game isn't angband obviously. You just walk away.
-
Right, but with persistent floors and finite resources it could be. Monsters that guard stairs, treasure vaults, chokepoints etc. Or just having to deal with it every time to earn experience at all.
Not generating corridors would be an easier, more comprehensive and neater solution to the corridor problem and this thread is motivation to get on with generating forest/outdoor maps but not until I've made a traditional dungeon work.
Throw ranged combat in though and the requirements on the AI to decide whether to rush in, hang back or dive for cover become more complex. I'll post again when I have a working implementation.
-
Corridors are just narrow rooms.
-
I think in a stealth-centric roguelike corridors could be made interesting. The player character should be inferior in combat to almost anything else, so that fighting one-on-one in itself is already dangerous. Corridors become about controlling less visibility at the cost of having less freedom of movement.
With multi-tile creatures corridors become a safe escape, which doesn't necessarily translate into trivially solved problems – have smaller enemies with higher chance to detect you, for instance, or make any route necessarily go through a room with a bigger enemy.
-
Right, but with persistent floors and finite resources it could be. Monsters that guard stairs, treasure vaults, chokepoints etc. Or just having to deal with it every time to earn experience at all.
. . .
Throw ranged combat in though and the requirements on the AI to decide whether to rush in, hang back or dive for cover become more complex. I'll post again when I have a working implementation.
Right, I think you're going to have problems with players picking off monsters at range in the guarding scenario.
-
Why are you still talking about this bullshit?
This is as pointless as arguing wheter permadeath is a good or bad thing! Come on, everything has good and bad sides, you know?
-
Well, while I think corridors are perfectly fine to have in your roguelike I think a discussion about how to make them more interesting is still worth having.
But by my impression maybe not in this forum. What a shame.
-
Wow, got the coauthors of Corridor Quest up at arms here...
-
Why are you still talking about this bullshit?
This is as pointless as arguing wheter permadeath is a good or bad thing! Come on, everything has good and bad sides, you know?
Because corridors are bad and it's worth comparing ideas on what would be better? I shouldn't have to explain this, but good points and bad points don't cancel each other out 1:1, and at some point a thing has so many bad sides it's better to just put it in a box and bury it forever.
Just saying "things have pros and cons, now let's not think about them" is so deviously clever. A really subtle way to post, yet have it be completely without content or meaning or merit.
As an aside, there's no argument about permadeath. If you're against it, roguelikes are not for you.
Well, while I think corridors are perfectly fine to have in your roguelike I think a discussion about how to make them more interesting is still worth having.
But by my impression maybe not in this forum. What a shame.
Don't say such reddit things. You're better than that.
As for what I actually favor, amorphous cavern formations with soft walls (vegetation, fog, etc.) as described by AgingMinotaur seems to me to be the most compelling model for connecting rooms.
As I read this I was reminded of Incursion's forest rooms, chasm rooms, slime pits, graveyards and kobold warrens.
And I definitely see the attraction. If only they got generated more often.
-
Here's another idea to make corridors interesting and dangerous to the player: Add the ability to move past an enemy for giving them an automatic free simple attack. This way players can be way more easily be circled inside corridors. When the game is balanced to mostly generate groups of between 2 and 7 enemies, fighting in open spaces becomes more desirable than fighting in tight corridors, as long as the ability to get away fast is more important than to fight few enemies at once.
The question is how to balance it that freedom of movement is more important than preventing raw damage.
Don't say such reddit things. You're better than that.
I'm kind of sorry. After coming back here after a few years my first impression just was that any topic explodes into fights pretty fast, no matter the topic. I actually did look at reddit shortly after (didn't bother with that before) and at least roguelikedev seems to be a nicer place.
-
Why are you still talking about this bullshit?
This is as pointless as arguing wheter permadeath is a good or bad thing! Come on, everything has good and bad sides, you know?
Because corridors are bad and it's worth comparing ideas on what would be better? I shouldn't have to explain this, but good points and bad points don't cancel each other out 1:1, and at some point a thing has so many bad sides it's better to just put it in a box and bury it forever.
Just saying "things have pros and cons, now let's not think about them" is so deviously clever. A really subtle way to post, yet have it be completely without content or meaning or merit.
As an aside, there's no argument about permadeath. If you're against it, roguelikes are not for you.
it was just an example to show how stupid this is...
And also are you trying to say that roguelikes are not for me?! That's my first post you see on this forum and you fucking say roguelikes are not for me!!!
Are you fucking trying to shit on me?
-
Why the fuck would you remove corridors, it's just stupid.
That about covers it. Maybe take this one to the DCSS forums?
-
Interesting opinion.
I have to admit Nethack's heavy use of corridors is one of the reasons I always preferred Adom to it (which is basically Nethack with a moustache as they say).
Corridors force indeed bidimensional tedious uninteresting play, and should be kept at the bare minimum.
I'd just put some perhaps in fixed content, but wouldn't make it possible for the map generator to place them randomly on any map.
-
Seems like a non issue to me, I don't design maps based on "generation." Or rather, I didn't use to.
-
I don't design maps based on "generation." Or rather, I didn't use to.
That make no sense. Please explain.
-
I don't design maps based on "generation." Or rather, I didn't use to.
That make no sense. Please explain.
As in I hard code it, there isn't anything random about it. It was more important when I was using booleans instead of variables.
-
No procedural content generation => not a roguelike.
-
No procedural content generation => not a roguelike.
Not necessarily, it depends on how it's done. There is a wide area between having something have static dungeons, and having something be "truly random". By definition, truly random ends up looking like a mess.
What you want in something in between. You have a limited number of dungeon designs, but the game shuffles between these different designs. You might have only three designs, or you might have over 2,000 designs.
The way I figured it out is to create mini-levels, and have a state machine shuffle randomly between different levels. There seems be a lack of tutorials on it, is how to specifically do procedural randomness. Most of the few out there is either behind a paywall, or it's in some other language besides Ruby.
On the other hand, badly written random can look like this.
* * * * # * * * # * # *
* * * * * # * # * * * *
* * # * # * * * * # * *
* * * * * # * * # * * #
* * # * # * * # * * * #
* * # # * * # * * * # *
# * * * * * * * # * * *
* * # * * # * # * * # *
* * * # * * # * * * # *
* * # # * * * * * * # *
* # * * # * * * # # * *
* * # # * * * * # * # *
* * # * # * * * # * * *
* * # * * * * * * * * #
* # * * # * * # * * * #
* * # # * * # * * * * #
* * * # * * * * * * * *
# * * * * # # * * # * *
Not the kind of level I really get anything out of playing. And it creates more work for the programmer for people that is just beginning to code.
-
I see; semi-random generation through randomly selected predetermined chunks. That's an interesting concept indeed, as long as there is plenty of chunks available so that dungeons never feel samey.
-
I see; semi-random generation through randomly selected predetermined chunks. That's an interesting concept indeed, as long as there is plenty of chunks available so that dungeons never feel samey.
Just finished it in a couple of days. It's not going to be perfect. This one is called Nihilist.^ ^
-
I'll go a step farther and say dungeons are harmful.
I'm sick of being underground all the time, I want to do something else.
Procedural generation can be applied to a lot more than just dungeons.
-
I've been thinking about that actually, would a roguelike even work inside the context of a spaceship?
Historically I'd make a spaceship be represented by letter P, and an enemy spaceship called E.
But increasingly, I'm considering having the Spaceship be the dungeon itself. But that in itself would largely change how one would go about doing procedural generation. Something like, the spaceship design reminds constant, but the surrounded space is procedurally generated. The spaceship itself would be something like a "home" state.
-
I've been thinking about that actually, would a roguelike even work inside the context of a spaceship?
Historically I'd make a spaceship be represented by letter P, and an enemy spaceship called E.
But increasingly, I'm considering having the Spaceship be the dungeon itself. But that in itself would largely change how one would go about doing procedural generation. Something like, the spaceship design reminds constant, but the surrounded space is procedurally generated. The spaceship itself would be something like a "home" state.
I've seen it done a few times before, both inside a spaceship dungeon and piloting a spaceship.
But usually when it is a dungeon, the spaceship just kind of works as a generic dungeon with a theme.
Prospector had space exploration in a spaceship, and Gearhead 2 had space combat with some simple physics.
I think another game by the guy who made gearhead called deadcold had a unique dungeon design for a space station.
But then it was pretty simplistic and not complete. Though its layout seemed pretty modular like what you describe.
-
I'll go a step farther and say dungeons are harmful.
I'm sick of being underground all the time, I want to do something else.
Procedural generation can be applied to a lot more than just dungeons.
I was just thinking of this post recently while playing Omega.
Take a look at this randomly generated outdoors section made for an encounter, it is exactly what I am thinking of.
(https://i.imgur.com/QrDj06N.png)
If you stretched this out some it would make for a really interesting overworld. This is what Cataclysm's outdoors should look more like.
There are too many roguelikes where "wilderness" just means an open plain with a few trees you can easily walk around.
Unlike the typical dungeon there are both hard and soft barriers here. The green hedges and grey rocks can snag you to slow you down, and rivers just drown you.
-
I was thinking about the scale recently and in "real life" rooms are (often) small compared to corridors which are at least 2 tiles wide and mostly straight pieces. If you think about the layout of real life house it's quite different than a roguelike dungeon. Then again dungeons are supposed to be more cave-like, but yet another again many caves are a lot like regular houses in that they don't have extremely long narrow corridors with large rooms such as classical roguelikes. This is something to think about if you want to create more "realistic" layouts I guess. I think reducing the overall length and amount of corridors would be something to try at least.