Gaming

I think of roguelikes as a game genre often optimized for the long-term player.

Let's say that you want to make an image meme involving a cat. Microsoft Paint, if you're on Windows, is a good choice for that. It's very easy to pick up. So you spend maybe a ten minutes learning the tools and then you can make cat memes, add text to an image, and paint simple diagrams on the thing.

Photoshop is a bad tool for that, for most people. The vast majority of Photoshop just adds a lot of complexity that is totally unrelated to making a cat meme.

However, let's say that your job is making marketing brochures to sell houses. You want to use drop shadows and you want to airbrush out power lines from phots. You can do it in MS Paint, but the tools are very limited and the result probably won't be great. The amount of time you'll require to get a couple of good brochures is going to outweigh the time investment required to learn Photoshop's much more extensive set of tools.

Paint makes sense for users who are going to use the thing a limited number of times. Photoshop makes sense for someone who is going to spend a decade where a lot of their daily work is manipulating images.

I think that many roguelikes games are kind of like Photoshop is. They are heavily designed to be ideal for the long-term user.

Roguelikes often don't have very pretty graphics. But...I think that a lot of value of pretty graphics is in their novelty. You see them the first time, it's really impressive. The fiftieth or five-hundreth time, it starts to become old hat. If you play a roguelike for a decade, the graphics don't provide a lot of benefit.

Roguelikes often have single-key combinations that the player needs to memorize to perform operations. That's a hefty learning curve for a new player, but if you've been doing it for a year, it just doesn't matter. You got the "muscle memory" in the first month or so.

Roguelikes often don't have great "intro" systems to help teach you the game. Having a good tutorial system is a big deal if you're gonna play the game for ten hours and three are spent learning game mechanics. I remember some great essay about how Super Metroid was very carefully designed to teach the player game mechanics without him noticing. Having a really good system for doing that is really important if a game will be played for maybe ten hours. But if they're going to play it for a thousand hours...well, optimizing the time spent learning the mechanics becomes less critical, because that time is a much-smaller portion of the time spent playing the game.