What do you hate most about Roguelikes?

[domtron]

PG can do infinite detail very easily, I think it has to be something else. Maybe it could be that the "interestingness" generated by PG is extremely sparse, while hand-made art has it very concentrated, because a human discards uninteresting stuff and only keeps the good things in, so that the result is very attractive to other humans.

I think you misunderstand me. I would say a fractal counts as only 1 detail and not a very well done detail either. It keeps going and going sure but it's a repeating pattern that never changes and is thus boring after a while. Even normal terrain generation is more interesting than that. When I say lacking detail is more a breadth issue then a depth one. This is what I was trying to get at via the 3 examples, 1st two focus mostly on PG terrain while the 3rd expands into social systems and character design. Depth is important, but if you leave out breadth it will get boring. A good PG would be more like using 20 fractals, I don't mean literally fractals but mathematical formulas and algorithms in general, and overlay them in a way that affects every aspect of the game.

[mdtrooper]

Someone please make roguelike first person shooter.

There is a software, that it has name of Necklace of the Eye . It is as a "frontend" for roguelikes. And it has a lot of features and yes. You can switch to FPS view some olds roguelikes:

[dulsi]

Necklace of the Eye is pretty neat looking. I'll have to try it out at some point.

[Wuzzy]

What I hate about roguelikes is when they INSIST on shitty text-only interfaces, or when the graphics mode is only secondary with no prime-time support. I'm not against optionally offering a text-only mode, but I don't like it when it's considered the PRIMARY mode or there is no graphics mode whatsoever.

This is just too antiquated and it's not good either: Graphics just are easier to parse (for the human brain) and also you can show much more different things with graphics. With ASCII, you only have so many characters, so many roguelikes just re-use the same character for tons of different items and monsters, over and over again. It not only looks ugly, but is also inefficient because you would have to check every single item on screen by hand, while with graphics, you just see it instantly.

Thankfully, most roguelikes have understood this and offer some graphical tiles. Yes, I know, graphical tiles can be a problem when the tiles are very large, but the solution is simply: “Make a smaller tileset”. :-)

What is especially dumb is that “ASCII” made it into the “Berlin interpretation” but I think ASCII is not actually what defines roguelikes. It's just an artifact of the era when Rogue was invented. Nowadays, this is just nostalgia. It does not even have anything to do with the genre, this is just dumb.

Apart from that, there is nothing else I dislike about roguelikes. It's a great genre.

[dulsi]

Wuzzy would this be acceptable for graphics?

https://www.gamingonlinux.com/articles/stone-story-an-rpg-animated-entirely-in-ascii-symbols-is-releasing-in-early-access-next-month.14658

Obviously it is not a roguelike. I'm very impressed but the art quality using ASCII art.

EDIT: Well that is disappointing. I read the description of the game. You don't directly control the character. All the animation is carefully scripted out then. It's very impressive but not what I expected and not as interesting to me.

[charlie]

I have 3 common problems with roguelikes:

1. Everything tries to kill you, and only you. It's a bit sad this this is such a common gameplay mechanic, and most times you just go scorched earth and kill everything. Yeah, there's occasionally friendly units, but there's very rarely any more complication than that. Why not rival creatures or sides? Why not big animals eating smaller ones? Why aren't animals territorial? Why do they all just swarm you and queue up to kill you? Simplicity of implementation obviously but it's still a very shallow and tedious game mechanic.

2. Only you need to eat. These aren't ecosystems. It's just... a way to stop you taking your time and healing too much. Very few roguelikes (well, none that I know of) provide a replenishing source for food.

3. Permadeath. Actually, not really permadeath, but the fact that it isn't properly accounted for. Having played way too much DCSS (so many hours of life I'd like back), the number of games lost after encountering enemies whose power I just didn't know or understand... there's literally no way to tell other than fighting monsters as to how dangerous they are. You learn by dying and losing your progress, and there's so many creatures in the game of varying powers, that's a lot of learning and ultimately thus a lot of frustration.

[Evropi]

My biggest problem with roguelikes is how unambitious they often are. This is changing, but when I got my start with them in the late 2000s, almost every RL followed the so-called Berlin Interpretation, which sets out the most basic formula for an RL game. The problem is that there are so many games that do exactly this, and too many others that feel beholden to that formula and are afraid to break out from it (or worse, bow to their users' uneducated demands).

I think the most important thing to break out of is the structure and objective. For that matter, I think having an open world instead of a dungeon with levels is a much, much more interesting structure for a game. Cataclysm: Dark Days Ahead absolutely nails this. It has an enormous amount of freedom and systems that feel appropriate in large part because it doesn't force you into proceeding down a dungeon with 26 levels.

Other good things to drop:

• Gimmicks and troll features (items that are solely bad and game ending are hard to define but something to avoid. Sokoban in NetHack is also a key example of a terrible gimmick)
• Extremely unintuitive control schemes and displays - using every letter of the keyboard and being case-sensitive about it, having to constantly go to other screens to see important info...
• Not having a theme, or choosing a very boring one. And no references! I've grabbed the Amulet of Yendor enough times already, stop remaking Rogue

I'm talking about just traditional roguelike RPGs here. In the last decade facets of the genre have entered many others, platformers in particular but also turn-based strategy.



It's worth thinking about how we can go beyond that and into more epic, longer-length games. This is the main notion that RL developers need to break out of in my opinion - that your game should be beatable in 45 minutes, which is how long the longest RLs last in a winning run.

I refer you to one of my all-time favourite games, Liberal Crime Squad. This is by the brothers who went on to make Dwarf Fortress, but they made sure to 'liberate' LCS first (no pun intended) under the GNU GPL. It's a terrorism simulator. By this I mean that most of the time is spent managing your activities, recruiting members and getting these members to do tasks to influence the political system. So it's a management game mixed with some on-the-ground action, which is often just talking to NPCs in different location, sneaking into places (disguise system) and more.

It's dripping with theme and hilarious, if somewhat offensive humour that honestly pokes fun at everyone. The community has worked on it from the late 00s onwards, with only a few people contributing code but countless people contributing ideas to what's already a masterpiece. ASCII, but supremely playable. A talented console programmer might be able to help make the Linux port a bit better

But more than that, you owe it to yourself to play it to see what a genuinely creative take on the RL is, a game that can last for in-game years. I'm not sure of the actual play length but I will say this - you can generally keep going after a severe failure, and this actually makes the game more interesting. So while you have permadeath, you are setting up from the beginning how you can have the LCS carry on even if the founder dies or worse, is turned into a Conservative. This is an interesting way to keep the thrill of permadeath but have more reasonable features around it, and may only really work in a grand strategy/management type game - but worth considering nonetheless, in addition to expanding the game's scope.