What do you like/dislike about roguelikes

[SkepticsClaw]

Roguelikes are awesome. Obviously the exploration of both environment and mechanics are what make them really fun; trying new things and seeing what happens, resource management and optimal decision making, the fact they reward careful planning and thought, their depth and complexity which gives me a boner.

Another RPG-like one is Legerdemain, which has an actual plot and some pretty decent writing (if you like Vancian-style purple prose). Mechanically, I haven't played it enough to get a handle on how deep it is, but it's got an interesting magic system at the least and the world building is not generic fantasy.

Going back into a corridor you came from and THEN holding an arrow key as AI goes to his slaughter one at a time is about as old as DND and Rogue.

Wow, you totally have roguelikes down bro! I look forward to you listing your Nethack ascensions.

Sorry, but I really don't get this 'it's boring because you just press an arrow key a lot' argument. You could pare anything down to the bare physical mechanics and make it sound boring if you deliberately omit everything interesting - but that's simply poor reasoning. Behold!

Driving? that's just turning a wheel. boring
Sex? that's just thrusting your crotch, how is that fun
Drinking? all you do is put alcohol in your mouth and swallow it, how fucking interesting i don't think

and so on.

 

[mondblut]

Driving? that's just turning a wheel. boring
Sex? that's just thrusting your crotch, how is that fun
Drinking? all you do is put alcohol in your mouth and swallow it, how fucking interesting i don't think

That's why smart people invented chaffeurs, blowjobs and drugs.

And proper RPGs too

 

[Elwro]

One reason I lke good rogulikes is that they always encourage thinking and discovering your own ways of dealing with obstacles, while still being fast games you can play for a few minutes in your free time.

 

[Awor Szurkrarz]

Especially with various spells and special abilities available both to the PC and the enemies.
Then there's ranged combat. So no, you are completely wrong.

How many levels have to be gained and false starts have to be abandoned until you get those "exciting special abilities" and start encountering opposition that requires their use?

Until then, a roguelike is basically a turn-based Pacman. At least a proper RPG *is* interesting on low levels.

Generally, the PC in ADOM is more like a level 4 AD&D character, so it's fun since the beginning - there's no period when the character sucks.

Wizards, Priests, Necromancers and Elementalists have spells since the beginning and fight mainly with them. Mindcrafters start out with ability to psionically confuse the enemies - they are most difficult class to play.
Assassins start out with poison and alchemy skill which allows them to manufacture more poison.
Necromancers can animate corpses since the beginning and use them as followers.

Then there are also weird classes like weaponsmiths, farmers and merchants.

Then some classes have healing skill and some don't. Healing skill gives slow health regeneration - so there are classes that naturally regenerate slowly and classes that regenerate very very very slowly. Since the amount of healing potions is limited and they are usually unidentified, that skill makes a very big difference - classes that don't have that skill must be careful with every monster because every wounding hit means a long-lasting set-back.

Many classes start out with ranged weapons. Usually ranged weapons are the main weapon - be it thrown knives or rocks or bows or crossbows as they allow hitting many types of animals and monsters without allowing them to fight back. Enemies that use ranged weapons appear since the starting dungeons. Take in account that getting to next dungeon level takes a few minutes since ADOM has one screen dungeons, so you quickly encounter jellies, floating eyes, spiders, summoning monsters, dark sages that use blasts of energy, etc. etc. etc.

And how do you imagine defeating a group of relatively strong enemies just by pressing arrow key? You have to maneuvre to make sure that you don't get surrounded (each additional attacker gets a bonus to hit), often you have to run past enemies in coward mode to get into better position.

Going back into a corridor you came from and THEN holding an arrow key as AI goes to his slaughter one at a time is about as old as DND and Rogue.

There's the wilderness and there are some big open caves. And you can't just hold the arrow against relatively strong enemies as you can get critically hit, start bleeding, some other monsters can come from the other end of the corridor, etc. Also, you have to get to that corridor first.
Also, you completely ignore spellcasters and ranged combat which is very important.

More like the more thought and work is put into designing a roguelike, the better it becomes.

Hardcore RL fans out there tend to disagree with your choice of an example of thought and work. I wonder what does that have to do with ADOM not being "pure" enough?

If they just want combat encounters in dungeons then obviously a game that also features much more stuff isn't for them. Note how they praise streamlining.
Personally, I play wargames when I want just combat.

 

[DakaSha]

Just to hype incursion a bit more: Im playing it again and it really is this awesome.. its the only RPG/RL i find myself playing waarriors just as much as spellcasters.. if not even more. they are just as fun and tactically involved. Also using light or even no armor is viable.. armor protects you but causes quite a bit of disadvantages as well. I actually think i prefer lightly armored chars.

Gods are also implemented in an awesome way (although following the rules of some actually requires some tedius learning )

And playing humans is also fun

 

[SkeleTony]

Dislikes:

1)Ascii(which is why I stick to graphical RLs).

2) single-character(not a game breaker but I prefer Tower of Darkness for being a party-based roguelike)-based.

3)D&D-based game mechanics. My fucking DOG end this shit! of all the P&P systems to borrow/emulate RL developers pick the WORST one!?

4)Non-static dungeon levels(ala the Angbands). This is not an absolute game-breaker as I still play ZangbandTK but man is it stupid!


Likes:

1)Permadeth(though I can live without it as well...not a game-breaker).

2)Tiled graphics.

3)Tactical options(see Tower of Darkness which allows for aimed attacks, dual-wielding, etc.)

4)Randomized dungeons(but STATIC! Once generated the levels do not change when you exit and return.

5)LOTS of races and/or classes to choose from, each having measurable advantages and disadvantages .


Things I do not care about either way:

1)Originality - I do not care if you are doing yet another generic fantasy world, Skyrealms of Jorune, superheroes, cyberpunk or what have you. Just make the game good and I will be there.

2)Story

3)Absurd numbers of options for eating corpses and shit. I just do not care enough to bother engaging in eating corpses most of the time.

 

[7hm]

yeah non-static levels are annoying

shiren does that. I dont get why.

 

[DakaSha]

easier to program

 

[7hm]

easier to program

well, yeah.

And for a small indie game, that is a fine excuse. For a professional game not so much.

 

[Damned Registrations]

Non static levels allow infinite scumming for items/xp. Makes the game easier when you want to accommodate crappy players. (Or shitty balance.... *cough*ADOM*cough*)

Although, at least with one way stairs like in the mystery dungeon series, one can't simply walk back and forth on the same set of stairs until items/enemies you want show up at your feet.

 

[Castanova]

+ Randomization
+ Difficulty
+ Permadeath
+ Loot
+ Your character's unique story evolves from the ups and downs of the dungeon run (I still remember parts of all my ascensions)

- Unavoidable insta-death via bad luck
- Spoiler reliance (even DCSS)
- Weak combat mechanics
- Resistances: Collect Em' All!

I'd love to see a party-based roguelike (PBRL) and I've even started programming one. I stopped probably for a similar reason we don't have more available already: the interface is a lot more troublesome to program. Doing a PBRL using only the keyboard is cumbersome and ASCII seems to be less ideal (how do you know who is who?). So we're talking, at a minimum, a point-and-click interface along with at least basic tiles. Already, this is a bigger barrier to entry than most RL programmers can bare.

On top of that, party combat exposes poor combat design much more easily. Most roguelikes including the most popular ones have bad combat mechanics but that's covered up by the fact that you control one guy and the fact that you're so vulnerable that you need to be constantly considering damage mitigation and escape plans.

The irony is that I think a PBRL would actually refine the roguelike formula to a better place. More time would be spent in individual encounters so the dungeon could be smaller, there could be less (none at all?) filler combat, less exploring same-y levels just in case there's a ring on the floor. Positioning and geography would be even more important. Permadeath would still be painful but it's not binary anymore - it may even be expected to "ascend" with some dead party members and a sign of accomplishment if all your guys survive. Characters/classes/builds can be more differentiated - no longer necessary that all character builds converge by the time you ascend. You can have character classes that are less self-sufficient (crowd controllers, buff/debuffers, and so on, to put it into cliche terms).

I'm getting a hard-on just writing this right now.

 

[Qwertilot]

Non static levels are just different. Means that depth is used by the player to set the difficulty level they think they're up to. Thats opposed to forcing them down to a given depth with random equipment which could well just kill them. Seems very reasonable to me with permadeath and all.

Lets the devs go a touch crazy with level difficulty at times too as you can just run(or not.....). Of course you can consequently scum horribly - witness the Borg for Angband being capable of nearly(?) winning - but should you wish to inflict that sort of tedium on yourself its only you who suffers

Oh detection/mapping magic is another thing which you don't seem to find often anywhere else. More teleportation stuff than usual too.

 

[Yeesh]

Let me say this about spoilers. I don't know how I feel about them. Are you really going to play a game as difficult and complicated (and requiring such an investment of time) as a roguelike and still want to figure out everything by yourself via trial and error? Have you unlocked the secret of eternal life or something? How many hundred of hours do you want to devote to one game?

On top of that, party combat exposes poor combat design much more easily. Most roguelikes including the most popular ones have bad combat mechanics but that's covered up by the fact that you control one guy and the fact that you're so vulnerable that you need to be constantly considering damage mitigation and escape plans.

The irony is that I think a PBRL would actually refine the roguelike formula to a better place. More time would be spent in individual encounters so the dungeon could be smaller, there could be less (none at all?) filler combat, less exploring same-y levels just in case there's a ring on the floor. Positioning and geography would be even more important. Permadeath would still be painful but it's not binary anymore - it may even be expected to "ascend" with some dead party members and a sign of accomplishment if all your guys survive. Characters/classes/builds can be more differentiated - no longer necessary that all character builds converge by the time you ascend. You can have character classes that are less self-sufficient (crowd controllers, buff/debuffers, and so on, to put it into cliche terms).

I'm getting a hard-on just writing this right now.

First, I don't see why ASCII is any less viable for a PBRL than a normal RL. If the player can be expected to keep track of what 8 different colors of D mean, then he can surely get used to memorizing which color @ symbol is who. I mean, I don't like ASCII anyway, but people seem to be able to play extremely complicated games using ASCII and there's no reason adding more party members would break the system.

Sure I'd get a kick out of trying a PBRL (once I've failed in my Stone Soup naga project), but there's no reason to believe that such a game has any inherent advantages over the classic one character model. Why would a party game be LESS likely to suffer from from build convergence than a single character game? Does the endgame of M&M VI vary more from playthrough to playthrough than that of Gothic II? I'd say the more characters you have, the more you're guaranteed to have every skill covered.

Also, more tactical combat is delightful to talk about, but here are two things that are very hard to balance:

1) a challenge in every fight means there must be a real possibility of character death
2) for the endgame to be challenging to a full party, doesn't it have to be more or less impossible for a party that's been whittled down to half strength?

The balance questions are vast. Traditional roguelikes are notoriously challenging (to most of us), and that comes not from any single fights that are tailor-made to be so tough but from the unforgiving nature of being all alone in a dungeon full of things that can kill you if you make a mistake. It's not so easy to see that the formula inclines by adding more party members to the equation.

 

[DakaSha]

1) a challenge in every fight means there must be a real possibility of character death
2) for the endgame to be challenging to a full party, doesn't it have to be more or less impossible for a party that's been whittled down to half strength?

Thats a question of designing the game around the fact that there are more party members.. Not of how to add a multi party team concept to a single player character game design.
It's not even an issue.. at all. It is at most an interesting challenge for the game designer

edit: Not even (more of) a challenge necessarily.. Designing a multi character game from scratch with that in mind would be no different then designing a game with the known fact of it being a single character experience

 

[Awor Szurkrarz]

Personally, I don't care about spoilers and random YASDs. I expect roguelikes to have enjoyable mechanics and diverse starting options. And judging by my experience with ADOM, finishing the game is the end of the game for me.
So, I actually like the idea of starting hundreds/thousands of times because it means much more hours of enjoyment for me.

 

[Zomg]

You could do a balanceable roguelike with non-static levels just by making it so you can never go back up once you go down. But non-static levels is not otherwise balanceable, if you let the player choose the danger level then they should tactically choose the most trivial that can still generate progress, which is exactly why Angband can be borged.

Edit - Actually, I guess you could rig it so you always get absolutely nothing from trivial floors, and make it possible to get 3x better stuff at 2x more danger or something, so that danger is a real tradeoff. Sounds like a hell of a design challenge to balance.

 

[Hobo Elf]

Or you could make a Roguelike with level scaling.

 

[FinalSonicX]

I find the idea of party based roguelikes very interesting, and I'm working on some design ideas for one at the moment, but there are some key problems exposed while also revealing some advantages.

Advantages:
+ more accurately emulates the pen and paper model
+ potential for hot-seat multiplayer (depending on how long a turn lasts)
+ potential for more complicated tactics
+ more classes and races can be coded and utilizes in ways that don't all need to be practical (bards and support classes become more viable). In other words, the role of each character can be emphasized.
+ games may start and end with different sized parties (death isn't necessarily the end of the game, as your party members may continue onward to victory)

Disadvantages:
- Movement systems become complicated. The most convenient systems tend to be the blob-crawling model or having the other party members have AI which allows them to follow the party leader (player). Perhaps anyone else has suggestions?
- Anything less than a tactically rewarding combat might become a chore
- A party could potentially "cover all of their bases", meaning the weaknesses of each class would be de-emphasized while the strengths are emphasized
- A lot of set-up required on the player's part before the game begins (building parties)
- Balancing the game may be a bit difficult, though I think it could be resolved by scaling the scale of the dungeon, and the power of the monsters such that larger more powrful groups of monsters are present, and the dungeon is slightly larger/more open to allow swarming to actually take place. I know a lot of people will want to burn me for suggesting any kind of scaling but the scaling would be proportional to the party size, not party power level (the scaling would probably need to be faster than linear because 9 characters are more than 3 times as effective as 3 characters).

That being said, I look forward to seeing someone actually pull off the mechanics and balance right. If not, I'm still working on mine so maybe I'll be the first

 

[Zomg]

You can have character classes that are less self-sufficient (crowd controllers, buff/debuffers, and so on, to put it into cliche terms).

Eh, I think that stuff is a bit cargo cult in a single player game. The point of breaking up roles in P&P or MMORPGs/MUDs is to force players to coordinate. In a single player game coordination is always going to be perfect anyway, and throwing out roles to separate @ signs doesn't matter much. The higher level of squad tactics is about geometry and space, area control, enfilade and defilade, terrain, cover, concealment, overlapping fields of fire, etc. I'd be more interested to see a roguelike where you controlled three very interesting fighters with only minor differentiation than a rehash of D&D.

 

[FinalSonicX]

One of the main reasons to provide multiple party members with specific roles is that it enables the enemy to "disable" the capabilities of the enemy more quickly and without needing to take out whole chunks of the enemy. In essence, the enemy can make a more "surgical" attack on the party.

An example if the classic fighter wall with mages in back. The crowd control poses a problem to the enemy, so they decide to flank and go for the mage, sacrificing a small number on the front as a diversion. With a quick and decisive flanking maneuver to kill the mage, they have effectively knocked out the party's capability to provide crowd control or wield the arcane magic they may rely on. These actions can provide interesting challenges even in a single player game, especially if the player relies heavily on certain approaches or tactics.

If you had 3 fighters in the same scenario, there would be less positioning involved, as by their nature each of the party members had multiple roles bundled up into them, thus taking out the crowd control is less of a task of coordinating movements and attacks etc. but more of a task of attrition. With 6 targets to choose from, each with their own unique position, status, and role in the party, the question of who to target and how to do so becomes much more complex. 3 targets provides less granularity.

 

[Zomg]

I have no idea why I picked the number 3, I didn't mean that to be part of my point. And I meant "fighters" each with the depth of a normal roguelike main character, not D&D fighters, bad term choice. I'm just thinking of all the squad tactics games I've played that didn't seem like junk and they usually have innate role differentiation tuned way down, with immediate role usually equipment-based, and an enormous number of single player CRPGs have boring rote role differentiations.

 

[Johannes]

In a single encounter I do consider it better if the differnt partymembers have a decent amount of differentiation, it can add a lot more to the positioning if not only the formation you're in matters. But they should be versatile enough so that they can (and must) adapt to different situations differently, instead of just sticking to a samey formation every battle.

One way to go about it could be that at the beginning everyone's pretty generic and versatile, but as you advance they eventually specialize based on how the game goes - who gets to wear the top notch plate mail you find, who learns which spells you find, who carries the teleport scroll or invis potion. Some of which could be switched between battles, while some would be permanent choices.


And in a party roguelike I think losing a guy permanently shouldn't happen too easily, allow for some sort of resurrection/healing from coma. Just make it drain on your resources a lot. Because it wouldn't be too much fun to always have these crippling blows when your partymember dies, but the game still continues - it's much better to die in a single decisive battle than have it a slippery slope of losing partymembers. Or alternatively allow to hire new people in a good manner.


Best way to counterbalance non-static dungeons imo, or any overt scumming in general (like killing regenerating mobs on a static level), is to just have some sort of time mechanic built into the game that overall punishes inefficient grinding. ADOM has the background corruption, though it would need to be much more stronger on low danger levels to achieve that. Or in the mildest just have score heavily based on turn count.


And ADOM has level scaling btw.

 

[Castanova]

I'd say the more characters you have, the more you're guaranteed to have every skill covered.

This isn't an intrinsic flaw of a "PBRL" - rather, it's a design decision. I can just as easily make the decision that you simply don't have access to all skills on every playthrough. Skills/spells could be randomized dungeon drops, or (using skill trees as an example) you don't get enough skill points to get all available skills. Etc, etc. This is especially true if the game doesn't have character classes, or you can't pick your character classes, or your attributes are randomized and you must pick your character class based on which ones your stats support.

1) a challenge in every fight means there must be a real possibility of character death
2) for the endgame to be challenging to a full party, doesn't it have to be more or less impossible for a party that's been whittled down to half strength?

Sure, which is why the game would probably need to involve some sort of resurrection mechanic. Although I could see a short, coffee-break PBRL doing without it.

I'm just thinking of all the squad tactics games I've played that didn't seem like junk and they usually have innate role differentiation tuned way down, with immediate role usually equipment-based

Replace the phrase "equipment-based" with "skill-based" and I don't see why a PBRL wouldn't compete just fine with the better squad tactics games. I'm not thinking of stupid-ass typical RPG skills like "+5% to damage!" or "Heavy Attack: Regular attack but +25% damage!" or whatever. I'm thinking skills would be more active and more situational... like, say, a "Push" skill where you knock the enemy back a square.

 

[Awor Szurkrarz]

You could simply have replaceable party members.

 

[7hm]

Sure, which is why the game would probably need to involve some sort of resurrection mechanic

Get outa here with your un-hardcore roguelike.

your attributes are randomized and you must pick your character class based on which ones your stats support.

DCSS handles this properly. There is no attribute randomization, because it leads people to scum. Anyone who has spent hour after hour in Wizardry rolling characters knows about this. It's more simplistic, but it's also easier for all involved to just have all Fighters start with 15 STR, 5 INT, etc.