You view roguelikes and roguelites as positively addicting because of your personality and the fact that those games test for the traits you care about, not because they're objectively better than other games that focus on another trait.
You probably don't view MMORPGs where the entire point of the game is just grinding mindlessly for months or years as positively addicting because your personality isn't the type that would like that, but there are tons of people in the world who aren't like you and like the average HN reader who would agree with you.
I've written more extensively about this before so I feel like keeping arguing here might be pointless, but in case you're curious for a more full version of the argument: https://github.com/a327ex/blog/issues/66
In any case, I think making something that is emergently addictive or addictive because it is fun is different from engineering something to be addictive by liberally lifting from all the years of research that went into slot machine design.
I don't know whether you frequent casinos but if you sit in the slot area for a while then compare that to big budget mobile p2w games (including Clash Royale which is actually also fun) is quite alarming. There are so many design similarities that lootbox games are basically slot machines without needing to pay out money...
If this is done by being fun and enjoyable, and the developer doesn’t cross the line into preying on addictive behaviour, no problem.
If the developer implements mechanics that are designed to “hook” the player to the detriment of the enjoyment and their health, then we have a problem.
- Static/dynamic typing: borders -- I prefer static typing to dynamic typing; this is a change from preference for dynamic typing in the past; today, I find "mixing and matching" of types to almost always cause runtime errors, not new insights
- Engines/frameworks: no borders -- I dislike frameworks in general, gamedev or otherwise, I much prefer to compose libraries myself
- Roguelike/Roguelite: ??? -- I don't care about a strict definition, but having a somewhat clear definition is useful for managing expectations
- Progression/No progression: ??? -- I like both
- Grinding/No grinding: no borders -- Grinding is boring, though I'll do it if you hang a nice enough carrot in front of me
- Gacha Games: no borders -- As expressed above, I consider them immoral, and that has less to do with mechanics and more with why they're employed. That said, I don't think Gacha mechanics have any sort of special borders compared to many other fun, well-defined mechanics.
- Professions: ??? -- My professional software dev is mostly a borderful experience, but I also do gamedev on the side...
- Easy Modes: ??? -- I don't see how it fits in your argument at all, or how is it even a topic at all (except for competitive multiplayer games, about which I don't care all that much). Most games have "easy mode" in form of a wide variety of cheats to choose from.
It seems I'm leaning a bit towards "no borders", but then I think this borders/no borders split isn't really "carving nature at its joints". It does not factor personality cleanly.
Also, as additional objection, I think you could use this argument to justify getting people addicted to gambling as a matter of "personality difference". I feel this explains too much.
I'm familiar with all of these except for "time-gating." What is it?
--
Mobile games use time gating for "lives" or "energy" or something so you can only play so much then you have to wait. Then they try and get you to spend real money to circumvent the time gate.
MMOs use time gating to slow player progression, with mechanics where you can only earn certain rewards once per day or once per week or something. It slows players becoming too powerful too quickly, or from completing their collections too quickly or from grinding the new rare drop in the first week.
This type of stuff is even creeping into other types of games, unfortunately. Especially with every game nowadays being an online game, even the single player ones.
For instance, as a kid, I've spent some ridiculous amount of hours playing Unreal Tournament (and to the lesser degree, Quake 3 Arena) - some of it with friends in an Internet cafe, but a lot in single-player mode too. I'd consider these two games to be pretty addictive for their time. But they were addictive because they were fun. Not because they tried to lure me in with skinner boxes, keep me in with time-gating, and monetize me with microtransactions.
Or, more recently, I've spent countless of hours playing 2048. At some point it became my go-to activity for every moment I wasn't actively concentrating on something. It may sound like an "intellectual" game, but in reality it's pretty braindead once you get the gist of it. You could probably say that the time I spent with it was unhealthy, but again, at no point it tried to be anything but fun. There were no features in there designed with exploitation in mind.
Or, roguelikes and roguelites. Between Cataclysm: Dark Days Ahead and RimWorld, I'm probably approaching some 1000 hours of total play time. Both are fun and addictive, in the positive sense. Neither has any of the addictive money-making features in them.
I could go on. But the point is, I maintain that ill intent is a better way to categorize games than "addictiveness". And I hate to invoke the "you'll know it when you see it" cliche, but the difference between the games designed to be fun and the games designed to milk you is glaringly obvious. Entertainment has value, and good entertainment tends to be addictive (it's pretty much a tautology). But it takes more than good entertainment to create a problem.