I love roguelikes. Traditional roguelikes. Turn based, grid based, punch-you-in-the-gut roguelikes.

The complexity, depth, and emergent gameplay in these games rival any other genre out there. The heavyweights like Nethack, Crawl, and ADOM can be played for years without mastering them… or sometimes without a single victory. And in the last decade, roguelikes have proven beyond all doubt how valuable their ideas are. Despite being a genre invented in 1980, new games are constantly borrowing roguelike ideas and even the name itself.

And yet we haven’t seen a big adoption of traditional roguelikes. Roguelite action has done gangbusters of course. Games such as Binding of Isaac sell millions of copies. Nuclear Throne and Spelunky aren’t too far behind. But where’s our Binding of Isaac? Where’s the turn-based dungeon crawler blowing up the Steam charts? Clearly, permadeath is something many gamers are willing to stomach. Games like Civilization are proof positive that they enjoy turn-based stuff too.

I’d argue that there’s nothing inherent in the structure of traditional roguelikes that is holding them back from mainstream success. It’s simply that, for the most part, the genre is still stuck in the 1980s. I’m not the only person that thinks that.

This series is about 4 pet peeves I have when it comes to roguelikes. They’re things that not only annoy the hell out of me, but that I suspect might be at the heart of why the genre lags behind. I’m going to give examples of games that either solve or exacerbate the problems and share my tiny attempt at improving things with my game Golden Krone Hotel.

Before we start it’s important to point out, you know, this is just like my opinion, man. Some of the things I hate are literally the very features other people love. I think that’s OK, considering the huge variety available in the genre.

Burden of Knowledge

It’s no fun playing a game without knowing the rules.

Sure, there’s something to be said for the joy of discovery. But when you’re playing to win, not knowing the rules and not understanding the mechanics fucking sucks.

Ever been introduced to a board game by a friend who knows the game well? Ever had that friend suddenly “remember” an obscure rule halfway through (right after they’ve taken advantage of the rule)? There’s nothing more frustrating!

Burden of Knowledge1 is when game rules are so opaque that they’re nearly impossible to learn by just playing the game. Instead, you have to spend time outside of the game learning how to play the game… assuming you can find the information you need at all.


I tried to get into Dungeons and Dragons in college and it just didn’t click for me. At the same time, I was playing a ton of Magic the Gathering. I later realized these games are polar opposites when it comes to learning. With each, you have to learn a few rules to get started of course. Later on, however, the way you learn diverges dramatically.

When someone needed to learn a spell or pick a feat in Dungeons and Dragons, they often stopped the game for 30 minutes to skim the player handbook and decide what to do. The basic rules of the game are over 100 pages! You have so many choices at any given moment of the game. The problem is that understanding all those choices requires a lot of reading.

Contrast with Magic the Gathering. Even though the game has a lot of depth, you only have to deal with a handful of choices on each turn. When your opponent plays a new card you’ve never seen, something magical happens. You read the words on the card and figure it out in seconds.

In comparison to tabletop games, videogames actually have it pretty easy. They can simulate entire worlds and have the player learn on the fly just by mashing keys and seeing the result. In theory, we should never have to read rules while playing a computer game. In practice though…

Spoil Story

If we’re talking about polar opposites in the roguelike world, I can’t think of a better pair than Nethack and Dungeon Crawl Stone Soup. Nethack is notorious for it’s reliance on spoilers. In Nethack, there are dozens of ways to instantly kill yourself. Conversely, if you know the right secrets you can use ridiculously powerful abilities for free (like preventing monsters from attacking you). My first experience with Nethack was trying and failing for 10 minutes to get out of the very first room because I couldn’t figure out how to open a door. Let’s just say I’m not a fan of this approach. If you are, more power to you.

20+ ways to insta-die in Nethack

On the other hand, DCSS takes a stand against spoilers. Check out this tidbit in the “philosophy” section of the Crawl manual:

Things ought to work in an intuitive way. Crawl definitely is winnable without spoiler access. Concerning important but hidden details (i.e. facts subject to spoilers) our policy is this: the joy of discovering something spoily is nice, once. (And disappears before it can start if you feel you need to read spoilers – a legitimate feeling.) The joy of dealing with ever-changing, unexpected and challenging strategic and tactical situations that arise out of transparent rules, on the other hand, is nice again and again.


It’s not that there are no surprises to discover in DCSS. There definitely are. The Abyss anyone??? Rather the surprises are discoverable simply by playing the game. If you play very carefully, you have a reasonable chance to deal with any surprises on the fly. Remember the plethora of instadeaths in Nethack? There’s only two kinds in Crawl: starvation and falling into water/lava. In both cases, the game gives you ample warning that things are about to go belly-up.

Burden of Knowledge goes way beyond cheap deaths and big spoilers. There are tons of things to learn. What does this monster do? What does that spell do? Where do these stairs lead? What options do I have available on this turn? The amount of material you have to figure out in a big roguelike must add up to a few college courses (which might explain a lot for some of us). And that doesn’t even cover all the basic stuff (like what red bars signify in games) we take for granted.

So how are we going teach players all that stuff? Here are some common techniques. I’ve ordered them by their proximity to the ideal method, actually playing the real game.

The 6 Circles of Roguelike Learning Hell


Tutorials are great for quickly learning the basics. On the plus side, you are playing the game… sort of. It’s a stilted, challenge-free, and dead end version of the game, but at least you’re playing that version instead of reading a wall of text.


Hints are like mini-tutorials. A minimal amount of text presented in-game at exactly the right time you need the information.  You do have to read, but you do so right in the middle of playing. Hints should be made unobtrusive as possible, preferably with an option to turn them off completely.

Modal hints force the player to read them, but are sometimes needed to convey crucial information.

Tooltips & Descriptions

The third circle isn’t that bad: on-demand information delivered in game that (hopefully) obscures very little of the screen. Tooltips help you quickly learn the UI. Good descriptions greatly reduce Burden of Knowledge regarding mechanics. It’s particularly useful when monster descriptions tell you what kind of spells, resistances, and attacks to expect.

Message log

Messages logs are a weird holdover from text adventures. Most modern games have too much going on too fast to even support a log, but the pace in roguelikes sort of allows them to work. A message log is a safety net for understanding. If the game fails to explain something visually, it can always go in the log and you don’t even have to stop playing to read it.

The utility of the message log is what makes it so dangerous. It’s very tempting to just dump stuff in the log instead of doing the very hard work of making it clear what’s happening on the play field. Even worse is that the log explains events after they happen. For small effects, that’s alright. For instadeaths, it doesn’t tell you anything until the game is over and you’ve potentially lost 20 hours of your life.

Help menus & manuals

To a first approximation, nobody reads help menus. As a designer, you absolutely cannot expect people to do it. Hopefully, the pattern is clear enough: the further removed from the game you are, the less enjoyable the learning and less likely someone is to seek out the information. In some sense the help menu is in the game, but you’re definitely not playing while you read it.

Wikis & guides

Abandon hope all ye who enter here. Because if you alt-tabbed over to a wiki, you have indeed given up on all hope of the game teaching you how to play.

Don’t get me wrong. On many occasions, I’ve been grateful for detailed wikis. But that’s really only after the game has hooked me for good. If I try out a new game and get stumped, I’d rather dump it than go read an essay about the difference between Flails and Polearms. Now consider your average steam user who has no idea what a roguelike is and stumbles across one.

Learning through level design

The above techniques are not all bad. If all roguelikes had good tutorials, hints, and tooltips, the genre would be dramatically more welcoming for beginners. I’ve added these things to Golden Krone Hotel and the game has clearly benefited. I have tooltips on every single button and UI element, detailed descriptions, a small hint system, and I developed a pretty neat little tutorial.

Still, I can’t help but feel we could do a lot better.

Let me hand it off to this Sequelitis video, which explains things so much better than I can:

Also check out How I Got My Mom to Play Through Plants vs. Zombies. This video by the designer of PvZ is a masterclass in teaching new concepts to players. The learning curve in PvZ is so damn good that, when development was completed, they completely replaced the Help menu with a joke:

Anyway, the theme of these two videos is that games shouldn’t have separate tutorials. People don’t like to read and they really don’t like to be forced to learn. In my experience, many players would rather struggle for hours not understanding stuff than spend 10 minutes in a boring tutorial.

Instead a game should present the player with a series of challenges with increasing difficulty such that an “aha moment” is inevitable long before the real difficulty comes.

I actually tried this with early versions of Golden Krone Hotel. You would encounter a high level foe in a narrow hallway at the start of every game. The only way to advance was by exploiting the main mechanic, killing vampires with sunlight. It did work as a teaching tool and it made people feel very clever. But it also forced experienced players to waste time whenever they started a new run and it sacrificed part of the procedural generation on the first floor.  I removed this inline tutorial years ago and never thought much about it.

After recently watching the above videos, however, I was desperate to once again incorporate a “blended tutorial” into my game. It’s not an easy goal to accomplish within the context of a roguelike. You don’t want to bore the player by forcing them to repeat the same content again and again. I was stumped until I thought back to the most polished roguelike I’ve ever played and it hit me…


Yet again we have two polar opposite roguelikes, Caves of Qud and Sproggiwood. These wildly different games are in fact made by the same developer, Freehold Games!

Caves of Qud

I love this game so much, but the user interface is a nightmare for beginners. Chalk it up to a short attention span or a lack of patience, but I’ve had a very hard time getting into the game. The help menus are good and a “Ten Things You Should Do When You Start To Play” section is nice, but playing the game itself can be quite bewildering.

When you start up Caves of Qud, this is the sidebar staring you in the face. There’s over a dozen incomprehensible stats. As far as I know, there are no tooltips to explain them. If some of these numbers get big, you’re in real trouble (my favorite being T for Temperature). There’s also a bunch of cryptic symbols on the left. Even words like Verdant confuse me. Qud make Jeremiah feel dumb.

Here’s the Abilities menu. First, recognize that a new player doesn’t even know this is a thing they can get to. Once there, you’ve got to manually bind all of your abilities. Meanwhile, you don’t get a description of exactly how the abilities work. I guess they were explained during character creation, but I often roll a random character and don’t see that.

To be fair, it does seem that Qud’s UI is slowly improving and there’s a lot more coming on the roadmap.


Sproggiwood on the other hand has a beautiful interface.

  • Every possible action you can perform (including wait!) is on screen somewhere
  • A magnifying glass icon lets you examine any of your abilities
  • Despite seeming rather sparse, a good deal of information is represented: the minimap, the duration of the character’s invisibility, XP, and the level/cost of each ability.
  • There are keypress hints on every button (more on this later)

Sproggiwood also has a delightful tutorial. This is how it should be done. Mandatory, but short. Not even labeled a tutorial. Valuable on it’s own and not a slog (the tutorial is very funny).

So Sproggiwood helped me figure out what to do: jam the tutorial right on the start of my game. When the tutorial is over, it jumps straight into the regular game and when the player dies, the next run skips right over the tutorial.

Which is better?

Obviously, it’s easier to make a simpler UI if the game is simpler. It’s easier to reduce Burden of Knowledge if there’s simply less to learn. Sproggiwood fits this description, but so do many of my favorite roguelikes: 868-HACK, Hoplite, and Auro. And to some extent, so does my game.

But my point isn’t to only make simpler games. Caves of Qud, even with all its flaws, seems like the more successful game because of its complex systems. Just check the Steam reviews: 85% positive for Sproggiwood vs 99% positive of Caves of Qud. My point is let’s strive for the best of both worlds. Games as deep as Qud, but as easy to learn as Sprog. That’s not an easy task, not at all. But the payoff could be huge.

If you want to see how well I accomplished my goal of reducing Burden of Knowledge, check out the latest update to Golden Krone Hotel. And join us next time when we talk about Identification systems.


[1] As far as I know, the term Burden of Knowledge was popularized by designers at Riot Games. All in all, that’s pretty hilarious considering League of Legends is a game with about 700 abilities, of which a great portion are indicated by similar glowing/flashing effects.