Human After All: Why Roguelikes Need to be More Than Mere Algorithms

I’ll be the first one to admit that seeing the term “roguelike” in any game’s descriptor will to an extent put me off. Or at least instinctively make me feel a little cautious as to said title’s versatility or longevity alike. Scouring the thesaurus for an alternative won’t skew the fact either — you can describe your game as “evolving” or “mutating” or “transformative” all you want, but many of us are clued up on our lingo to know what’s up. I won’t deny anyone the satisfaction if such a thing as procedurally-generated rooms/maps/dungeons is one you can [honestly] find immense enjoyment from, but on a personal level — and even this may seem old-fashioned by today’s video game standards — the fact is that I will always prefer intentionally-structured level/world design over anything some back-end algorithm cooks up on boot-up. The former instills purpose, reason, a sign that developers have [hopefully] considered how everything is presented and inevitably delivered to the player. How does it look, how does it evolve, how much (if any) of the game’s additional mechanics can be tutorialized or otherwise shown off through the world-building?

With a roguelike, not only do you risk disregarding these fundamental details, but as many releases have sadly been a victim of, the limitations of one’s design can come to the forefront. The sub-genre’s greatest strength becomes its greatest weakness — algorithms care not for how well something flows or feels…it just picks and chooses and pronounces it as something inherently new. This might, in theory, be true and the premise of creating a seemingly-endless amount of technically different dungeons to trawl through or over-worlds to explore is a great feature to note on a press release. In much the same vein as the “survival” genre still finds itself in, the idea is interesting, but the execution almost always falls flat. Unfortunately, with how saturated the independent scene has gotten — long before you even factor in how many are flaunting the idea of roguelike being a “feature” — it’s becoming increasingly difficult for a game to stand out. From my perspective, simply doubling-down on the idea of procedurally-generating a player’s task won’t cut it anymore.


The reason I bring this up is that, despite my disdain for the mechanic — despite my slight withdrawal whenever I hear the sub-genre tag floating about — over the past few years, there have been those games that have managed to be more than just that. Better still, the idea of procedural-generation — integral it might be to such things as level creation beforehand or even dungeon-exploring in real-time — isn’t the cusp of the main gameplay. When “roguelike,” or developers themselves more likely, can find the means for this to sink into the backdrop, that’s when we see potential flourish. That potential has exuded with games like Seraph, Rezrog and most recently, the soon-to-be-released Dead Cells. The former of which, one of my personal favorites of 2016, succeeded because it focused on the gun-wielding combat, enemy variety, fluid movement and highly-octane pace with which you were moving about its created environments. The fact that the levels were predominantly blocky and maybe a little “samey” in parts didn’t matter — the genius lay far outside the contrived notion that “each play through is unique”or words to that effect. Let’s be frank: spewing out this same tired line just won’t cut it.

Where Rezrog did indeed have a novel pull, the randomness of its dungeons needn’t matter when the focus is squarely on strategically making one’s way to the target room, all the while keeping tabs on your character’s stats and much more. Dead Cells is a perfect example of integrating roguelike mechanics but not letting it dominate the space it’s sharing — resulting in a new micro-genre, the Roguevania, being coined by such a release. When you look back at these games — here’s some more: The Binding of Isaac, Enter the Gungeon, FTL, Downwell — a significant portion of their appeal isn’t so much the variety of possibility on show or even how close to organic/hand-crafted its levels may be, but in its gameplay that’s front-and-center. Visual flair and artistic aesthetic may help ease players in and alleviate some of the mathematic tinkering taking place, but games are of course meant to be enjoyed and an enticing (maybe addictive) mechanical loop will keep players coming back again and again. A game world’s environments will only do so much.

Seraph Review Screenshot
That’s not to say that this can’t itself bring some temporary satisfaction or emergent moments. In my recent review of The Swords of Ditto — another roguelike RPG, to add to the many — I admitted that one of the highlights during my time with the game was curiously stumbling into another generated dungeon whereupon the intrigue to delve deeper was rewarded by a fair amount of helpful loot. Likewise, I recently had a brief go at For The King prior to its release last month and while I did feel inundated with more mechanics and statistics than you can shake a stick at, I was compelled to figure out, at least, a sizeable portion of it all in order to make progress through the low-poly styled, hexagonal grid. Again, the game doesn’t simply end with its procedurally-random forte of generating a world map and letting players loose in it; there is so much more to dive into and best of all, that’s what is at the forefront of proceedings.

I’m not suggesting developers need to hide the fact their game is roguelike. If anything, that’s actually created a far more deflating reaction than outright declaring it as such from the get-go. Given how on-point its marketing had been, I was left disappointed when it was revealed that Strafe was yet another procedurally-generated venture. Rather than fully committing to the ’90s shooter mentality (it so obviously was painting itself as during its initial campaign) and delivering us an actual single-player campaign with intended levels and secrets to discover, part of the charm was tarnished somewhat by the notion that little felt entirely crafted from hand. That, once again, things were left in the cold and calculating hands of some back-end algorithm.


Maybe it’s my speaking to particular developers about this — they themselves admitting, some in a peculiarly caught red-handed expression, that going down the route of a roguelike is generally “easier to do” — that has spawned this niggling dismay with the sub-genre. Naturally, like anything, it all comes down to taste and preference. After all, there are some people out there who genuinely adored games like [vanilla] No Man’s Sky and even now are continuing to play Sea of Thieves. The latter, of course, hasn’t been marketed as a roguelike obviously, but its barebones content does share many similarities with the principle of a roguelike in that any real objective measure of investment seems to rest instead with the player strangely enough, rather than the developer themselves. Certainly there’s a market there to prop up these small-to-middle scale titles; why else would we be seeing more and more of them popping up on a seemingly weekly basis nowadays?

Such has been the popularity of the genre and the ease at which a back-end algorithm can maximize a game’s output, is it likewise surprising to find it make an appearance in even some AAA releases as of late? Bloodborne’s Chalice Dungeons, Persona 5’s Mementos, even Destiny 2 had a go at it…or at least tried to. The difference here is that those games treated it as an accompaniment and an addition, rather than the focus. Like the genre-defined examples mentioned above, the environments were but an extension of the gameplay loop rather than its priority. You could very well focus on looking for hidden items or combating unique monsters without ever noticing the similarities present within that temporary construct.


The roguelike is the double-edged sword to beat out all. While the ease of application and humble intrigue as to what players make of it might well be viable reasons for its inclusion, sooner or later there comes a point where the thinly-spread variety begins to show. But even if your game does feel at times like it has the depth of a puddle, providing players with a reason to stick with its mechanics, its visuals, its story or even its reason for being, will go a long way to keeping things feeling fresh. The last thing you want is for a game to unintentionally reveal its cold, calculated, machining self and so long as developers remember to put as much passion and hard-work into a game’s more organic and interactive elements, there’s no reason why the roguelike should turn stale. Besides, some of the genre’s best moments can be had when the procedural worlds merely set the stage and allow the discovery of the challenge, the real star of the show, to emerge slowly-but-surely — that’s something an algorithm could never hope to script or otherwise manufacture.