Representationalism in Graphical Roguelikes

Every hardcore roguelike player can relate to that scene in The Matrix where Neo is looking over the shoulder of the guys who just stare at the incessantly-changing green ASCII characters.  To everyone else, our screens look inscrutable, but to us, they're dwarven fortresses under assault by forgotten beasts, faithful monkeys fighting off goblin necromancers, and trap-lined dungeons riddled with deadly traps.  Not everyone sees what we see.

It comes as no surprise to anyone, then, that roguelikes began adopting graphics in an attempt to make them more accessible.  Now that games like Dungeons of Dredmor are making the roguelike accessible, is there no more room for our beloved ASCII roguelike?

Actually, I think there still is room.  The enduring popularity of Dwarf Fortress, Dungeon Crawl Stone Soup, and others are a testament to that.  The ASCII roguelike does something that very, very few video games do nowadays: require you to use your imagination.  There's an upper limit to what most graphics can convey - especially on an indie budget - so, counterintuitively, abstracting your display away from representationalism can actually actually increase immersion.  In the absence of visual detail, your imagination must fill in.

I briefly flirted with the idea of making Lone Spelunker a "graphical" roguelike.  I even started on some "pixel art" mockups of what it might look like in a graphical version.  Against my better judgement, here's what it was looking like:

The early 90's called - they want their tileset back.

While it would almost certainly be more accessible to players new to roguelikes, I just didn't like it.  It was too literal, robbing you of a lot of the mind's-eye imagination of what the tiles represented.  For instance, in the screenshot above, the spelunker looks male and caucasian, whereas when it's just a "@" in the ASCII version, no gender or ethnicity is implied.  The walls look blocky, somewhat out of necessity due to the mechanics, whereas the walls in the ASCII version, while just as blocky, imply that they are representational of the *mass*, not trying to represent the actual outline of the structures they represent.  And don't get me started with what was happening when ropes and pitons came into the picture.

Worst of all, when I tried a functional mockup with the tileset above, I discovered that it just did not feel as exploratory as the ASCII version.  It felt...like an arcade game.  Like I was supposed to just run-n-jump my way through it.  Like I was only supposed to see what was drawn there.  Like someone laid out a level for me to move through in a level editor, and that it was a human-created place instead of a place carved naturally by geological activity over eons.  It just wasn't the feel I wanted for Lone Spelunker.

(Granted, some of the above comes from my embarrassing pixel art, but I suspect the above would continue to be true even if I were a much better pixel artist than I am.)

Perhaps someday, after Lone Spelunker has found its foothold, I can corral some excellent artists and produce a graphical version to expand its reach.  I am bothered, a bit, at the ASCII display's effect of shutting people out of the game.  For those people, a graphical roguelike would be better than a roguelike they wouldn't play at all.

But considering that A Dark Room soared to the top of the iOS App Store - a place where casual gamers and eye candy rule the roost - perhaps ASCII graphics don't suffer the level of stigma that I believe they do.  Maybe, with the right presentation, even gamers new to ASCII graphics will accept it.  It will be interesting to see what happens when the game is released.