On the Historical Origin of the “Roguelike” Term – Slashie's Journal

On November 11 2017, at the Roguelike Celebration in San Francisco, I had the pleasure of doing the opening talk, titled “What is a Roguelike?, 2017 Edition”, where I explored some of the interpretations on the Roguelike term.

Something interesting I found while doing research for the talk was what could be the first efforts to organize a group of games similar to rogue under an umbrella term. This happened on Usenet, the technological precursor to Internet Forums, throughout which most of the discussion of computer games happened.

My current progress on this research indicates that the efforts to define this hierarchical relationship, intended mainly to facilitate discussion, ended up serving the purpose of creating a community of niche developers and players who through the constant usage of the term ended up giving it an evolving, collective meaning that was relatively stable until the 2010’s decade, with the growth of the indie games scene and diffusion of the more action focused “Roguelike” labeled games.

Note that while there are indications that the term had been used already before this organization was proposed, based on the discussion it brought forth it seems clear that it was nowhere near established until it happened.

1. The first Request for Discussion

On July 2, 1993, Andrew Solovay sent the first version of a “Request for Discussion” to the news.groups Usenet newsgroup (2). He proposed to create a new hierarchy, rec.games.dungeon, to group together some games that shared similar characteristics.

I would like to propose formally that a new hierarchy be created, namely rec.games.dungeon. This hierarchy would contain groups dedicated to discussion of rogue-type games. These games share the following characteristics:

  • They are character-based; that is, the game generally presents a map of the current playing area, with different characters representing the player, opponents, objects, etc. Text is a secondary element at most. This is in contradistinction to other, “text-based” adventure games, such as Adventure, Zork (a.k.a. “Dungeon”), etc.

  • They are very portable. The games are usually written in a high-level langauge [sic], and as noted, they use character graphics; this makes them easy to port to various architectures (mainframes, workstations, and different types of personal computers). As a result, they don’t fit into any of the comp.sys.*.games groups, which are machine-specific.

This Request For Discussion was likely prompted by the repeated posts calling out for the creation of an Angband specific newsgroup, since most of the posts on rec.games.moria were now being related to Angband. Deriving from these posts, we have a proposal by Aliza R. Panitz, who insisted on “ reorganizing [the games] under rec.games.dungeon – recognizing the intrinsic similarities of all these games. ” (1)

This definition of “rogue-type” games (which would later be renamed to “roguelike”, extending over the same term) meant to group games sharing the following characteristics:

  • Character based display: The game generally presents a map of the current playing area, with different characters representing the player, opponents, objects, etc. Text is a secondary element at most. This is in contradistinction to other, “text-based” adventure games, such as Adventure, Zork (a.k.a. “Dungeon”), etc.
  • Highly Portable: Easy to port to various architectures (mainframes, workstations, and different types of personal computers).

As we can see, the reasoning for this grouping didn’t include random / procedural content generation, permadeath / persistent consequences, turn based interactions nor any other game design related characteristic from rogue or any of the existing similar games. The “character based display” they shared was mostly what prompted the original grouping.

His justification for this reorganization was the following:

There are several advantages to having such a hierarchy, notably:

  • Most people who play any of these games are interested in other games of the type. Thus, if the newsgroups were grouped together, people could easily find other groups that interested them.

  • It would provide an area for discussing dungeon games that do not currently have a USENET group (e.g. larn), and for discussing the relative merits of different games.

  • It would provide a uniform framework for adding new groups for
    dungeon-type games that become popular.

Even though this Request for Discussion didn’t give out a detailed definition of a Roguelike, we can see there is a notion that in some way, these games are connected and most importantly, there will be more similar games being developed in the future.

The games he intended to group together, based on the existing Usenet groups where:

  • Rogue
  • Hack and Nethack
  • Moria
  • Angband

He also suggested the creation of a new group (rec.games.dungeon.misc) for “ discussion of dungeon-type games that do not yet have mainstream groups (e.g. larn). “. This gives us an idea even at this point there was roguelike development activity beside these 5 games, albeit it was probably very small. Omega also comes up as part of this initial ecosystem in the ensuing discussion.

One of the key points of his request for discussion (besides doing an initial measurement on the interest of the community of proceeding ahead) was how this group of games would be named:

Should the hierarchy actually be named “rec.games.dungeon”? Some have expressed concerns about this name. They note that many text-based adventure games (e.g. “Zork”) are set in dungeons; indeed, one early version of Zork was named “dungeon”. Another concern: The unifying characteristic of these games is not that they’re set in dungeons, but rather the style of interface. There is no reason one of these games couldn’t be set, e.g., on the Death Star or in a gulag, and yet remain appropriate for the hierarchy. Again, perhaps “dungeon” is a misleading name.

Perhaps, since “rogue” is the granddaddy of these games, the hierarchy ought to be called “rec.games.roguelike”. Or perhaps some other name would be better.

2. Initial debate around the name

PLEASE, there has got to be something better that roguelike. — Robert A. Hayden

Since it was a Request for Discussion, a discussion ensued and it was pretty heated; while some people agreed, others failed to see the connection between these games.

Note that some of the discussion revolved around the practicality of the proposed Usenet group hierarchy in terms of discoverability, the actual design of Usenet hierarchies and whether the groups should better be categorized based on the computer architecture of the games. That part is less relevant for our purposes.

Here are some interesting excerpts: (3)

  • How about “Single-User Computer Role Playing Game” or .sucrpg. That is what Hack, Moria, and the rest share. — Michael Neylon
  • Well, I know of the rogue, but I think there’s a lot of people who don’t. After all, rogue is quite an old game, and not very common any more. — Juha Laiho
  • I do worry about confusion between “dungeon” and dungeon and dragons type people. “Roguelike” is the most accurate description anyone has yet come up with for these games. — Peter Glen Berger

This particular message provides a future point of expansion for the research:

  • […] it’s already pretty accepted that rogue-like games are a special category of their own. […] — Ron Asbestos Dippold

Although it’s possible that the rogue-like term had already been conceived as a category, it’s unlikely that it was already widespread, based on the messages we see on the discussion.

Following up on the discussion, Andrew iterated over the first RFD and posted a second one. At this point, besides strengthening his proposal for the “roguelike” term, Andrew further tries to narrow down the definition by adding another characteristic: being descended from rogue. (4)

There is general consenus [sic] that “rec.games.dungeon” would be a bad name for the hierarchy. The main problem is that “dungeon” describes far too wide a variety of games, including pencil-and-paper FRP games, games like ZORK (one version of which was actually called “dungeon”), graphic-oriented computer games, etc.

“roguelike” seems to me much superior. It describes all the games in the proposed hierarchy, and no other games. The one thing all these games have in common is that they’re descended from “rogue”, and have a basically rogue-style interface.

Of course, the issue here is determining the criteria for being a “descendant” of rogue, since there is no direct relationship between the source code of Rogue and any of the games intended to be grouped. This brought up a lot more of discussion as some people disagreed on Rogue’s legacy being a defining factor to group the games (3):

  • I’ve never seen rogue. Played rogue. Or even heard of rogue until this RFD and the name just rubs me the wrong way. […] ascii-dungeon is perfectly discriptive [sic] […] and shows the theme of the games without anyone having to know anything about rogue. — Robert A. Hayden
  • […] I don’t like the idea of having this group of games named after a single one of them. […] — Zodiac
  • […] Roguelike i could see applying to Larn and Nethack, but a program like Moria is somewhat a different now [sic] (I understand it’s origins in rogue).For example it is not a single screen game, as Rogue is. […] — Brian J. Dunford

Some others failed to see the need of a unified hierarchy

  • All the omegas, morias, hacks, etc…have diverged too far from each other to be unified into 1 hierarchy. — axl
  • […] considering how much the ‘roguelike’ (that has got to be the worst heirarchy [sic] name I’ve ever seen) games have diverged since the creation of rogue, they no longer need to be grouped together. […] Perhaps the difficulty in finding a unifying term for the games should serve as a hint to you that there aren’t enough similarities between them anymore to justify a subheirarchy. [sic] — John N. Trusell

Then we get glimpses of the character based display may be not being the real unifying characteristic

  • I realise that some of these games have acquired somewhat more sophisticated interfaces than ASCII graphics. At least in the ones I’ve seen, this is simply a matter of substituting coloured ASCII characters or small graphical images for the original plain ASCII characters, and the game still feels very much like an ASCII graphics one. — David Seal

And we also have some initial ideas at abstract ideas such as the “feeling” of the game (instead of a list of characteristics) being a factor for grouping

  • Some attempts to enumerate these features have failed (IE, “ascii”, “dungeon”, etc.) But for any of them you can look at the screen and say “Duzzn’t dat remind ya of rogue?” They don’t each have every common feature, but they form a reasonable grouping. — Tom O. Breton

Some of the alternatives that were discussed:

  • rogueish
  • rogue-style
  • alphabet-soup
  • ascii-dungeon
  • hack’n’slash
  • hacklike
  • hacked
  • morialike

3. Settling on a name

On the third Request for Discussion, Andrew finally settled for Roguelike for the lack of a better term. (5)

This is a third RFD. The discussion seems to have reached its conclusions; I don’t think many people will change their minds at this point, and people are mainly repeating the same points. Those people who agree on a name seem to favor “roguelike” as the least of all available evils.

And thus began the road to the formal Call for Votes, which was the official way to legitimize changes on Usenet groups hierarchy. There were further outcries against the idea (3)

  • […] the whole “roguelike” suggestion is obscure, overly limiting, poorly named, and generally abhorrent (the last IMO). […] I’d also like to see how the vote on this insane “roguelike” heirarchy comes out so I’ll know what steps need to be taken. — Chris Waters
  • The sub-hierarchy rogue-like is pointless. It doesn’t tell the average person what kind of games they are, unless they happen to know what rogue is. — Gregory R. Weiss
  • Someone could next come up with the idea that DnD is a subgenre that got it’s ideas from Tolkien, […] This insistence to search back to the ‘granddaddy’ of the genre and honor it with new newsgroup names is just plain silly… — axl
  • Try playing Nethack like it was Moria and see how far you get: They are two seperate [sic], even if related, games. The only real connection is their rogue ancestry, a connection that has no relevence [sic] from a player’s point of view — Sean O. Conell

As we can see in this point of time, having a character based display as a common feature was not consensually considered to be enough to provide grounds for grouping these games together. Some people considered this kind of display as something that any game could have, regardless of its origin or its gameplay, in a similar way that today we wouldn’t consider Call of Duty and World of Warcraft to belong to a same genre just because they are both rendered in 3D.

Of course, they were right on the long term, but at this point we also see more supporters of the idea appear, now further extending the definition towards gameplay related features.

  • These games all have something in common that the other D&D-type games do not — their ancestry. This includes style of play, techniques, and other things. — Richard Hartman
  • […] The games nethack, rogue, moria, larn, et. al. share certain common features, game goals, and strategies. Often, people who enjoy one game of this genre, and who follow the relevant newsgroup, learn about, and come to enjoy, other games of the genre, […] — Aliza R. Panitz

On the official Call for Votes, Andrew added another characteristic to his Roguelike interpretation: (6)

  • They are very addictive, and people who like one of these games often enjoy the others.

While pretty shallow, I think most Roguelike players will agree. I believe this was one of the defining factors of the community that was creating around these games.

The results of the Call for Votes were the following: (7)

RESULTS rec.games.roguelike reorganization results - 337 votes Yes No : 2/3? >100? : Pass? : Group ---- ---- : ---- ----- : ----- : ------------------------------------------- 229 100 : Yes Yes : Yes : rec.games.roguelike.announce 228 99 : Yes Yes : Yes : rec.games.roguelike.misc 212 115 : No No : No : rec.games.roguelike.moria 214 116 : No No : No : rec.games.roguelike.nethack 207 120 : No No : No : rec.games.roguelike.rogue

This initial process was thus half successful for the purposes of Usenet group reorganization, but proved to be the first step of an inevitable collision between these similar games.

Also, with the creation of rec.games.roguelike.announce and rec.games.roguelike.misc, a Request for Discussion on the creation of rec.games.angband was resumed (it had been put on hold, waiting for the decision on the rec.games.roguelike hierarchy). (8)

With the rec.games.roguelike reorganization vote now over, it is now time to restart the discussion on rec.games.angband. This is a continuation of the existing discussion, not a new RFD, so the vote will be held soon.

With a 184/50 votation, it was decided that the group would be created as rec.games.roguelike.angband. (9)

4. Further refining the name

As proposed in the Call for Votes, Aliza R. Panitz was appointed as moderator for the newly created rec.games.roguelike.misc. Part of her work included creating a FAQ for the group so that newcomers would know what the discussion was about.

In an early iteration of this FAQ, on July 1993, we find another early effort at defining what a Roguelike is, based on the original discussion for the rec.games.roguelike hierarchy (10):

Although the common features of rogue and its many descendants are ‘obvious’ to many people, they are difficult to describe in simple terms. All of the games mentioned below are single-user, fantasy role-playing computer games, generally set in a dungeon, run with a simple graphic interface. In all of the games, the player controls a single character, who roams around getting more powerful, in order to fulfill a difficult quest. Sword-and-sorcery rule the day.
Logistically, they’re all free games; executables, and generally sources, are available by FTP.

The FAQ also gives us an idea of some qualities that would disqualify a game from discussion on the group on the criteria on “not belong to the category”:

  • Commercial games (like Ultima and Wizardry)
  • Multi-User-Dungeon games
  • Text-mode games (like Adventure, Zork, and the whole family of Infocom text adventure games.)

5. Six months later

On April 1, 1994, David Grabiner revived the discussion about the creation of the rec.games.roguelike.{moria,nethack,rogue} hierarchy. (11)

It has been six months since the creation of the rec.games.roguelike.* hierarchy to dicuss Rogue and related games. The hierarchy is inconsistent as it now stands. Since the consensus was to put the Angband group in this new hierarchy, while the votes to move the existing groups for Moria, Hack, and Rogue failed, we now have rec.games.moria but rec.games.roguelike.angband.

Therefore, I propose that we clean up this hierarchy by putting everything under rec.games.roguelike; this will make it easier for readers of one group to find the others.

The final results overwhelmingly supported the move (12)

rec.games.roguelike.* results - 295 valid votes Yes No : 2/3? >100? : Pass? : Group ---- ---- : ---- ----- : ----- : ------------------------------------------- 229 45 : Yes Yes : Yes : rec.games.roguelike.moria 230 60 : Yes Yes : Yes : rec.games.roguelike.nethack 219 59 : Yes Yes : Yes : rec.games.roguelike.rogue

6. Conclusions

Based on the examined material, we can draw some conclusions on these early definitions:

  • They make no mention at all of the game design features that we currently consider being inherent to the games of the genre, such as randomly generated content (procedural), permanent consequences (permadeath) or being real time or turn based. As a result it’s hard to consider these historial interpretations as a criteria on the discussion of whether Roguelikes must to be turn based or lack meta progression or non-permanent death.
  • They make a great emphasis on features which nowadays may seem distant and secondary except on the most traditional roguelike circles:
    • Character based display: While still a staple of roguelikes, nowadays we wouldn’t consider it a requirement.
    • Being “Free”: Disowning commercial and closed source games.
    • Being portable: A very specific quality of these first games.
  • They have been contested since the beginning, but nobody ever found a better term.

Sources

  1. Panitz, Aliza (June 18, 1993) “Rec.Games.Dungeon.[Moria|Angband|Hack|Rogue|Omega|Misc]” Newsgroup: rec.games.moria. Usenet Message-ID: 1vsuhk$mj7@bronze.lcs.mit.edu. Retrieved November 12, 2017 using Google Groups
  2. Solovay, Andrew (July 2, 1993). “RFD: rec.games.dungeon.* hierarchy“. Newsgroup: news.groups. Usenet Message-ID: 211gvmINNsnp@rodan.UU.NET. Retrieved November 12, 2017 using Google Groups. Also available on November 13 2017 in the ISC.org archive here
  3. Followup posts on Solovay, Andrew (July 2, 1993). “RFD: rec.games.dungeon.* hierarchy”. Newsgroup: news.groups. Usenet Message-ID: 211gvmINNsnp@rodan.UU.NET. Retrieved November 12, 2017 using Google Groups here
  4. Second RFD
  5. Solovay, Andrew (July 20, 1993). “3rd RFD: rec.games.roguelike.* hierarchy“. Newsgroup: news.groups. Usenet Message-ID: 22ie07INNbaa@rodan.UU.NET. Retrieved November 12, 2017 using Google Groups. Also available on November 13 2017 in the ISC.org archive here
  6. Solovay, Andrew (July 28, 1993) “CFV: rec.games.roguelike reorganization” Newsgroup: news.groups. Usenet Message-ID: rec.games.roguelike-CFV1@uunet.uu.net. Available on November 13 2017 in the ISC.org archive here
  7. Solovay, Andrew (August 18, 1993) “RESULT: rec.games.roguelike reorganization” Newsgroup: news.groups. Usenet Message-ID: 24tdibINNb6t@rodan.UU.NET. Available on November 13 2017 in the ISC.org archive here
  8. Grabiner, David (August 18, 1993). “2nd RFD: rec.games.angband”. Newsgroup: news.groups. Usenet Message-ID: 24u5agINNpei@rodan.UU.NET. Available in the ISC.org archive here
  9. Dipold, Ron (September 27, 1993). “RESULT: rec.games.roguelike.angband passes 184:50”. Newsgroup: news.groups. Usenet Message-ID: rec.games.roguelike.angband-RESULT@uunet.uu.net. Available on November 13 2017 in the ISC.org archive here
  10. Panitz, Aliza (October 7, 1993) “Roguelike Games Mini-FAQ” Newsgroup: rec.games.roguelike.misc. Usenet Message-ID: 291sr3$f5e@bronze.lcs.mit.edu. Retrieved November 12, 2017 using Google Groups
  11. Grabiner, David (April 1, 1994). “RFD: rec.games.roguelike.{moria,nethack,rogue}”. Newsgroup: news.groups. Usenet Message-ID: rec.games.roguelike-newgr-RFD1@uunet.uu.net. Available in the ISC.org archive here
  12. Manyard, Jay (May 31, 1994). “RESULT: rec.games.roguelike.{moria,nethack,rogue} all groups pass”. Newsgroup: news.groups. Usenet Message-ID: rec.games.roguelike-newgr-RESULT@uunet.uu.net. Available in the ISC.org archive here

Oh, I intend to mine this for inspiration when I get to my own roguelike directory. :slight_smile: