Results for the 2013 Roguelike World Survey has been posted

[AgingMinotaur]

shroom, I'm not trying to aggress you, although I do enjoy the hyperbole of the discussion and I think some interesting points have come up. However, Slash's bullet points doesn't include something like Game of Life, and by reading them as if they did, you just make yourself out to be stupider than you probably are. Sure, something like "turn-based play" should be taken to imply "turn-based play, with continuous player agency", which does exclude GoL; but there are yet innumerable implied meanings in such a simple statement, that you could nitpick at will, so it would be impossible to cover all of them (for instance, permafailure doesn't include keeping a pet, and procedural space doesn't include a game of Go). I'll grant you that GoL takes place on a grid, but the other points frankly do not apply. In this context, we should allow ourselves to assume that people know the first thing about what is being discussed, and constrain ourselves to not pretending as if we don't; filling in blanks is part of language, human, and stubbornly failing to do so just to score rhetorical points, isn't very productive.

You obviously put a lot of weight on the technological origin of Rogue et al. While that's interesting, I don't see how it constitutes the actual gameplay of the genre, much like how modern novels are considered epic literature, even though the earliest epics were part of an oral, rather than a written, tradition. This history of literature's origin is hugely significant, but still, the sorry sods who most assuredly claimed that Gilmgamesh or The Iliad stopped being "real" literature once they were written down, have been forgotten (probably because they refused to have their own arguments put in writing, but that's going off on a tangent here). And there was much rejoicing.

I'd be very interested in a more systematic break-down of what you would consider a proper Roguelike. So far, that's remained vaguely implied in your posts (IMHO).

As always,
Minotauros

[mushroom patch]

Slash's bullet points do include the game of life, but I don't think there's any reason to argue over that point. The bullet points fail to capture what a roguelike game is on one hand and fail to differentiate roguelike games from other types of games on the other. It has nothing to do with rhetoric.

But sure, let me try to give a definition.

A roguelike game is a computer game that implements roleplaying (in the sense of pencil and paper games like Dungeons and Dragons) game mechanics and content, following the tradition and idioms of early computer games on UNIX and VMS including rogue, moria, hack, and larn.

A "systematic breakdown" is somewhat beside the point, as it would really just be elaboration on the meaning of the words "tradition" and "idiom" as they relate to rogue, moria, hack, and larn.

A couple of things about this definition: First, it can be and often is pretentious to call one's game "roguelike." Roguelike games are not schlock.

Second, a game can be roguelike in large part by virtue of pedigree. For example, if a game is built from code bases that go back to the originals with reasonable continuity in spirit from one iteration to the next, it can be roguelike while differing from the originals in substantial ways. For example, tomenet is obviously a roguelike game, even though it's not "turn based" as understood by r.g.r.d alumni, it's not single player, and its permadeath mechanics are a bit iffy. It's more important to be able to point to how the game fits into previous work in the genre than to tick off a checklist of features.

Arguments about whether a game is roguelike shouldn't be "well, it doesn't have this feature, that feature, and it has this mechanic that's not allowed," as if it's a question of scoring something. The argument against a game should question its connection to existing work, especially the so-called "major roguelikes" (and that doesn't mean spelunky, btw), and what its contribution is (i.e. is there something new here or is it at best an homage?). If it does differ crucially in mechanics, are these differences so radical as to better place the game in another genre or do they address an issue in previous work and suggest a way forward?

You obviously put a lot of weight on the technological origin of Rogue et al. While that's interesting, I don't see how it constitutes the actual gameplay of the genre, much like how modern novels are considered epic literature, even though the earliest epics were part of an oral, rather than a written, tradition. This history of literature's origin is hugely significant, but still, the sorry sods who most assuredly claimed that Gilmgamesh or The Iliad stopped being "real" literature once they were written down, have been forgotten (probably because they refused to have their own arguments put in writing, but that's going off on a tangent here). And there was much rejoicing.

There's an obvious conceit here. Are graphical "roguelikes" (in the sense understood by many here) producing genuine improvement and innovation over more traditional roguelikes? You imply the answer is something like "Oh God, yes!" My answer would be, "Not really." I've played my share of recent games with no terminal interface, etc. I don't see the transition from cavemen sitting around a fire to robed scholars stroking their chins over a pile of scrolls. I see traditional roguelike efforts going on pretty independently of the discussion here and making real strides (see DCSS, Sil) and I see projects with lots of graphics that don't strike me as anything new. It doesn't help that so many of the latter have names like CommercialFranchiseRL.

Anyway, I wouldn't put graphics up there as a matter of definition, but I think a roguelike game needs a reasonable excuse to have them, for example, "some guy from my mailing list put them together and sent me a patch and they seem to work okay and aren't likely to cause trouble down the road, so I put them in as an option." I do think you need a damn good excuse not to have a terminal interface though.

This getting way too long, but two more things about my emphasis on the history and technical origins. These are things that seem to lost to current thinking. First, roguelikes were originally very social games. People played them on multiuser systems at universities, traded war stories, etc. (And war stories, in my opinion, remain one of the best things about the genre.) Then PC users started getting DOS versions via shareware catalogs and some of them seem to have thought that's how it all began. Well, it's not and multiuser systems remain popular and vital. Second, the centralized server-oriented aspect, while just a fact of life back then, represents an untapped opportunity now. There are significant technical advantages to delivering a roguelike game via ssh or telnet, even aside from the multiuser considerations (which are also particularly relevant in the current climate). Telnet is widely available and requires no installation of software. On the other hand, a game that runs primarily on servers can be made of whatever you want -- you want to use an integer programming engine, a numerical linear algebra package, a graph theory package, an image processing library, and ten other things your user is guaranteed not to have installed? You want to use lots of different programming languages, big databases of precomputed whatever? No problem. All your player needs is telnet.

[Bear]

I don't think anybody is going to come up with a definition of roguelike games that satisfies everybody.  That said, I'll throw something at the wall.

I think that roguelike games must give the player as much time as he or she needs to plan and input the next move.  Being turn-based single player games is the easiest but not the only way to accomplish this.  If having slow reflexes or a motor impairment is a handicap then you are not playing a roguelike game. 

I think that exploration of procedurally generated maps in the presence of opposing forces is an essential part of roguelike games.  It doesn't really matter whether the player is playing a medieval-fantasy character, or the crew of a starship, or a cockroach.

I think that an unambiguous 'game over' condition is essential to roguelike games.  How much of the game's objectives a player can accomplish (and how much of the game's content a player can interact with) before the 'game over' condition is met, must be limited by player skill rather than just by how much time the player spends playing. 

Those three conditions are the ones I think are absolutely essential. And they already leave out some games that people consider roguelike.  I'm okay with that; I don't think that those games are in fact roguelike. 

There are some other conditions that I consider 'nearly' essential to roguelikes, such as a power curve; the idea that the player makes long-term gains in power and resources during the game.  If a game has no power curve I will doubt that it's roguelike, and suspect instead that it's more sidescroller-like.  I think a symbolic (in the sense of being unambiguous) display is important; character-based displays are an easy way but not the only way to accomplish this.

[Darren Grey]

Single character focus is an important and oft overlooked aspect, and is why the likes of Civilisation don't quite match. Roguelikes tend to be adventure simulators, with focus on person's actions and experiencing as they would experience, often including field of view, unidentified items, single person resource management, etc. Permadeath and procedural content both enhance this experience.

[mounta1nman]

my roguelike def: 'a game that sacrifices visual interface for depth of gameplay'

yes, not historical nor technical. trying to restrict a def by those two dimensions is never successful, and as such not worthwhile. extract what really matters an keep moving.

[mushroom patch]

The problem with the bullet points people are putting forth now is that if someone came along with a new game with a compelling connection to existing traditional roguelikes that had, say, party based adventuring or player-generated maps crowdsourced via an online Dwarf Fortress-like simulation game, you would realize your definition is wrong.

The reason people are constantly arguing about these bullet point definitions is that they're wrong. On one hand, they always include something that has no connection with the genre as it has existed for decades and on the other they restrict the scope of new work.

my roguelike def: 'a game that sacrifices visual interface for depth of gameplay'

yes, not historical nor technical. trying to restrict a def by those two dimensions is never successful, and as such not worthwhile. extract what really matters an keep moving.

This captures an important part of the spirit of roguelikes, probably the most important part. Too often this is corrupted into "a game that sacrifices visual interface so that anyone can make one."

[Slash]

A roguelike game is a computer game that implements roleplaying (in the sense of pencil and paper games like Dungeons and Dragons) game mechanics and content, following the tradition and idioms of early computer games on UNIX and VMS including rogue, moria, hack, and larn.

Everybody is entitled to his own interpretation... in my opinion yours is interesting, super broad and completely useless for the two purposes I consider the most important: 1. To easily let other people know what these games are actually like and 2. To conserve the identity of the genre in time.

I've been thinking about the question of roguelikeness vs. roguelike; the definitions posted in the homepage and defined by the "Berlin Interpretation" were created in a time when our main interest was to allow the evolution of the genre, to make people experiment without fearing to go off bounds of a strict set of features; I think this should continue, but today we need a more strict definition that coexists with the current one, mainly for point # 2.

Second, a game can be roguelike in large part by virtue of pedigree.

I'll then proceed to check out Nethack source code, touch here and there and create the first fps roguelike.

This getting way too long, but two more things about my emphasis on the history and technical origins. These are things that seem to lost to current thinking. First, roguelikes were originally very social games. People played them on multiuser systems at universities, traded war stories, etc. (And war stories, in my opinion, remain one of the best things about the genre.)

Now that's very interesting! but in my mind that can added to the definition as a feature, regardless of the pedigree.

[mushroom patch]

A roguelike game is a computer game that implements roleplaying (in the sense of pencil and paper games like Dungeons and Dragons) game mechanics and content, following the tradition and idioms of early computer games on UNIX and VMS including rogue, moria, hack, and larn.

Everybody is entitled to his own interpretation... in my opinion yours is interesting, super broad and completely useless for the two purposes I consider the most important: 1. To easily let other people know what these games are actually like and 2. To conserve the identity of the genre in time.

Your definition (and maybe I am holding you more responsible for the "Berlin interpretation" than is warranted) is already undermining the identity of the genre. It has no sense of history or tradition and suggests that roguelikes are defined by some weighted average of feature scores. It's silly. On the other hand, if your goal is to constrain the genre through a list of required and forbidden features, you're not helping the genre. Genres develop over time as new work builds on, not copies or updates, old work. Identity of a genre is based on the connection and continuity between new work and prior work.

Easily letting people know what something is is not what definitions are for. Definitions are supposed to be definite, not easy. If I want to let you know what topology is, I don't give you the definition. I give you examples and down to earth explanation. The answer to the question of what a roguelike is does not have to be a definition.

And my definition is not overly broad. It excludes a lot of games you would call roguelike.

I've been thinking about the question of roguelikeness vs. roguelike; the definitions posted in the homepage and defined by the "Berlin Interpretation" were created in a time when our main interest was to allow the evolution of the genre, to make people experiment without fearing to go off bounds of a strict set of features; I think this should continue, but today we need a more strict definition that coexists with the current one, mainly for point # 2.

I think this shows a peculiar attitude toward the genre. It's not yours nor can some central authority define it. If a lot of people find your definition lacking, it's a reflection of a lack of continuity with preexisting notions and work and, I'm sure, a reaction to the idea that you or a group of your associates is in the position to hand down a definition (one intended to guide future development, no less!) of something that you did not create (and had little to do with either at the time of its inception or in its rise to prominence, as far as I can tell).

Second, a game can be roguelike in large part by virtue of pedigree.

I'll then proceed to check out Nethack source code,

Great, Nethack could really use new developers!

This getting way too long, but two more things about my emphasis on the history and technical origins. These are things that seem to lost to current thinking. First, roguelikes were originally very social games. People played them on multiuser systems at universities, traded war stories, etc. (And war stories, in my opinion, remain one of the best things about the genre.)

Now that's very interesting! but in my mind that can added to the definition as a feature, regardless of the pedigree.

I'm glad you think so. This part of my comment was less about definitions than explanation of previous commentary on the provenance of roguelike games and its importance in understanding the genre.

[Krice]

I'm working on the definition, but writing a scientific article is harder than I thought.

[Omnivore]

I'm working on the definition, but writing a scientific article is harder than I thought.

You think that's hard, try modeling the Berlin interpretation as a set of simultaneous equations using matrix calculus with Rogue as the identity matrix!

[AgingMinotaur]

I think each will have to rely in part on intuition wrt whether or not a certain game is a RL, and opinions will differ on corner cases. Some aspects that matter to me (why I like RLs to the exclution of almost every non-RL computer game I've played since childhood), are the notion of replayability, and also that RLs play in a lot of ways similar to boardgames. The replayability aspect touches upon typical features like permadeath, procedural/emergent content, and the id subgame (although I find this poorly implemented in most RLs) – consider how Rogue's creators have stated that they wanted to make a game that they themselves would enjoy playing many times over – but might also be achieved in other ways.

So I think it's possible (and may be an interesting experiment, maybe as 7drls) to make a RL that "scores" zero points according to the Berlin Interpretation, or conversely, a non-RL that adheres strictly to the constraints of that definition.

Just to clear up some misconceptions, shroompatch: Slash had no hand in creating the Berlin Interpretation; it was conceived at IRDC in 2008 and put up on this website after that. Slash explains some more about its historical context above. Also, please note that I take this as a personal insult:

Are graphical "roguelikes" […] producing genuine improvement and innovation […]? You imply the answer is something like "Oh God, yes!"

I actually prefer a terminal interface nine times out of ten, but I also consider that a question of taste rather than truth, and graphical representation a question of makeup. And I respect devs who for some reason wrap their system in a GUI (I consider games like Shiren, Powder, and TOME to be RLs without a doubt). My comparison to the evolution of written texts wasn't to make a point that graphical RLs are better or more evolved, just that tradition implies ongoing change rather than constant stasis. After all, oral and written literatures still exist side by side, overlapping at some places, and with their particular strengths and weaknesses. I think it's impossible to talk about "the RL scene" any longer, as one could as little as fifteen years ago. Today, the community have branched out in different (and in places overlapping) communities: you'll come over incompatible conceptions of the genre at rgrd, anband.oook, amongst the 7drl crowd, here, in mainstream "indie" communities (TIGSource and the like), etc. Some will never have heard about Nethack, others nothing but.

As always,
Minotauros

[mushroom patch]

I think each will have to rely in part on intuition wrt whether or not a certain game is a RL, and opinions will differ on corner cases. Some aspects that matter to me (why I like RLs to the exclution of almost every non-RL computer game I've played since childhood), are the notion of replayability, and also that RLs play in a lot of ways similar to boardgames. The replayability aspect touches upon typical features like permadeath, procedural/emergent content, and the id subgame (although I find this poorly implemented in most RLs) – consider how Rogue's creators have stated that they wanted to make a game that they themselves would enjoy playing many times over – but might also be achieved in other ways.

Yeah, I agree about most of these aspects. I still think it's a mistake to try to think about what a roguelike is in a vacuum. In essence, what people are saying today is "We've looked at a bunch of games we think are roguelikes, including the original examples as particularly representative, we've extracted a set of common features (and non-features) that make them distinct from most other games, and those traits are what we're going to use to define what a roguelike is." What I'm saying is you don't need to reify a particular set of defining traits, you only need defining examples and work that follows them. This is both more specific and less restrictive. In reality, what people are putting forth here with their bullet point definitions are not so much definitions as specifications -- which is ridiculous. This is most true of the so-called Berlin interpretation.

Just to clear up some misconceptions, shroompatch: Slash had no hand in creating the Berlin Interpretation; it was conceived at IRDC in 2008 and put up on this website after that. Slash explains some more about its historical context above.

I don't think I have a misconception about this. It's the definition he puts forth on a prominent outlet for discussion of roguelikes. I hadn't heard of it before I read it here. He's done a lot to popularize it. It doesn't matter whether he created it.

Also, please note that I take this as a personal insult:

Are graphical "roguelikes" […] producing genuine improvement and innovation […]? You imply the answer is something like "Oh God, yes!"

I actually prefer a terminal interface nine times out of ten, but I also consider that a question of taste rather than truth, and graphical representation a question of makeup. And I respect devs who for some reason wrap their system in a GUI (I consider games like Shiren, Powder, and TOME to be RLs without a doubt). My comparison to the evolution of written texts wasn't to make a point that graphical RLs are better or more evolved, just that tradition implies ongoing change rather than constant stasis. After all, oral and written literatures still exist side by side, overlapping at some places, and with their particular strengths and weaknesses.

Well, okay. I'm not trying to suggest you're not "hardcore" or whatever. For your reference, when you draw a parallel between an instance of fundamental technological progress (e.g. written language) and any other instance of change, it will be interpreted as "I think this change is awesome" every time. I don't see how you get that I favor stasis from what I've said here. I've said that the current situation wrt to the "definition" of roguelikes (or roguelikeness or whatever) is untenable, that a better way to understand the notion is through the tradition and history of work within the roguelike genre especially as it relates to the original examples, that the genre naturally extends by continuity of new work with old, and that feature/non-feature lists cannot capture this dynamic.

Put it this way: Suppose you're talking about tomenet/mangband and related angband variants. Krice comes in and says "That's not a roguelike, it doesn't have permadeath narrowly defined and I can't think fast enough to play angband in real time." If you go by my definition, you just say, look the game is derived from angband which was derived from moria, it doesn't differ that fundamentally from either game, and where there are fundamental differences, they address a shortcoming of their predecessors in an interesting, playable way (even if it's a little rough and could be improved). On the other hand, with the Berlin interpretation, you find yourself in the peculiar position of having to say, "You know, strictly speaking, you're right, Krice."

[AgingMinotaur]

Well, okay. I'm not trying to suggest you're not "hardcore" or whatever.

Because, you know, I am, I am, I am! (or rather, I don't exactly see it as a mark of excellence to think that a pretty or costly presentation raises the inherent worth of a cultural artifact)

Put it this way: Suppose you're talking about tomenet/mangband and related angband variants. Krice comes in and says "That's not a roguelike, it doesn't have permadeath narrowly defined and I can't think fast enough to play angband in real time." […] On the other hand, with the Berlin interpretation, you find yourself in the peculiar position of having to say, "You know, strictly speaking, you're right, Krice."

It's probably fair to note, though, that even upon its conception, the ("so-called" ) Berlin Interpretation wasn't intended to be definitive, but rather to list some typical features. Though I do agree it can form an unhappy condition for defining RLs. OTOH, I think the danger of relying too much on technosocial genealogy is that one risks excluding classics like Powder, the whole spectrum of Japanese Roguelikes, or even ADOM (for being closed source, whether of not you personally consider that a cardinal sin).

Also, regarding "war stories", YASDs and the like, yes, this points to the unique strengths of the RL genre – and relies a lot on both gameplay qualities and Internet culture.

As always,
Minotauros

[Slash]

Identity of a genre is based on the connection and continuity between new work and prior work.

How is this connection and "continuity" granted from an external point of view if not by assessing the features of the game?

I'll shorten your definition to make it even more definite!"A roguelike game is a computer game derived from rogue. Go look it up and reconstruct the whole story of the genre if you want to know what I'm meaning."

[JeffLait]

It's probably fair to note, though, that even upon its conception, the ("so-called" ) Berlin Interpretation wasn't intended to be definitive, but rather to list some typical features.

The definition of roguelike has long been contentious, long before the lightning rod that is the Berlin Interpretation was put out there.

It does not help that when each individual built their own definition, they did it with a different corpus of games.  For example, when I started on POWDER I was a bit concerned people would declare it Not A Roguelike, so added my own definition to the page.  My definition included this interesting item:

Complex interactions of properties. While the commands for a roguelike are simple, the potential interactions are not. My favourite example is equipping a silver ring as a weapon in order to damage a creature vulnerable to silver, but not one's other weapons.

Rogue, which one would naively expect to be the touch-stone for roguelikeness, fails this test.  It is something I got out of all of my playing of ADOM and Nethack and reflects the little experience I have with the huge *band family.

The term "Berlin Interpretation", rather than "Berlin Definition" is rather intentional.  The model is off the Copenhagen Intepretation: an attempt by so-called experts in the field to come to an agreement about something inherently non-agreeable.  Of course, the result is something no-one is entirely happy with.  But what I like about the Berlin Interpretation is it puts a stake in the ground for starting point, something we did not have before.  I know the inclusion of "ASCII display" may seem ridiculous to the modern crowd, but;

• It was actually a significant issue at the time; and quite progressive for it to be moved to the low list.
• I'm not sure how much videogame genres should encompass.  Personally I'm for them focussing solely on gameplay, but I think practically in many people's minds it will also include art-style and setting.

It is also important to note that genre construction, as an aspect of taxonomy, is by definition going to be imprecise.  We are not dealing with abstract mathematical constructs here.  There is a precise definition for Topology not because the things that are topologies have a precise definition, but it was decided to redefine things to match a precise definition.  Our goal with a roguelike definition is a functional lens for us to better understand games.  A precise definition for roguelike would be pretty useless.  There is no gain to evenly dividing games into Roguelike and Non-Roguelike camps!

So why bother at all?

There is gain to identifying what is it that causes us to think of some games as roguelike, and some games as not.  We can then choose to keep or discard these items in other games.  Or identify them in existing games and better understand what mechanisms caused us to enjoy/hate those games.  I consider it very convenient to have "Sci-Fi" and "Fantasy" and "Romance" as genres, despite the fact that obviously any particular novel may or may not respect those divisions.

What makes something a roguelike isn't any particular feature.  Just like being "plot-driven" doesn't make something a fantasy novel.  But, you can't just make a checklist either!

Compare it with taxonomy of animals.  Often these do devolve to checklists: species A differs from species B due to a slightly longer clavicle.  But these lists very much are designed merely to distinguish existing species - we could invent an animal that catastrophically fails any checklist based taxonomy system.  Instead, there exists the holotype for each species, which represents the complete animal to match against.

Though I do agree it can form an unhappy condition for defining RLs. OTOH, I think the danger of relying too much on technosocial genealogy is that one risks excluding classics like Powder, the whole spectrum of Japanese Roguelikes, or even ADOM (for being closed source, whether of not you personally consider that a cardinal sin).

Techno-social?  I guess I should read the entire thread...  But isn't POWDER very much a derivative of Nethack, and ADOM likewise traces Omega and Nethack?  Unless you only count inheritance via code, in which Nethack is no longer a Roguelike, as, AFAIK, Hack is a clean break from rogue?

My main argument against genealogical approaches, however, is that there is a lot of convergent evolution in videogames.  And, the cases of convergent evolution are probably the most interesting for genre definition, as they represent an attractor in the game-play space.  When two games have the same gameplay because one was inspired by the other, it doesn't tell us as much if two independent people were led to the same gameplay.  OTOH, I'm not sure how "clean" the Japanese Roguelikes are from western influence, Nethack had a Japanese port since '93.  And, of course, we know the Diablo designers were aware of *band.

Also, regarding "war stories", YASDs and the like, yes, this points to the unique strengths of the RL genre – and relies a lot on both gameplay qualities and Internet culture.

War stories apply to all video games worthy to be called video games.  It also seems a pointless thing to add to a definition.  How am I supposed to write a game that has war stories?  Answering this starts to ask the question of what sort of warstories, and runs us into something that is starting to look like the Berlin Interpretation: a list of techniques/gameplay elements found in most roguelikes that seem to be most necessary for generating a roguelike experience.

So, in summary, I'm all for debate about what makes a roguelike, but mostly because I want to learn of new ways of looking at games to understand how they work and how they differ from each other.  In the end, I don't particularly care where anyone draws the line between roguelike and non-roguelike, and find it entertaining how much people are now using it as a catch-all for "Procedural".