Okay, a few notes:
1. I think that, personally, the difference between "roguelite" and "roguelike" are mainly in terms of how the penalties for death are handled. And how progression is handled.
2. With Bionic we went with "lite" because it has a lot of progression outside of the individual levels... and the levels themselves are not really full runs anyhow. Overall the entire campaign is kind of your run, but kind of not. It's a mix because there's not an exact analogue there.
3. With Bionic we also went with lite because it has a back and forth strategy element to it. Losing has consequences, but in the same sense that a strategy game does: you lose ground. Eventually that can lead to an overall loss, but it's not a given.
Regarding Rogue Legacy:
4. I definitely view that as more of an RPG in a lot of respects given how much is based around you definitely dying and then doing another run. And how much is revolving around stats in general.
Regarding Isaac:
5. I view that as a roguelike because you're unlocking things only based on the run you are in. And the other things you unlock that are for cross-run are basically just making future runs more varied and interesting, or harder, or what have you. But each run is self-contained.
Regarding Starward Rogue:
6. There's not overarching progression like BD or RL, per se. That said, you do have an overarching objective like you do in RL or BD: in this case, repairing your ship. And like BD, you can't do this without doing multiple runs.
7. That said, the ship repair work is more of a meta-goal between runs, and doesn't really affect individual runs much. You also have to win runs for that to be improved.
8. Overall that provides goals for progressively harder runs, and progressively more complexity in the runs, etc.
9. The RPG elements are absent. You can't grind stats. This is a skill-based game, and if you can avoid taking hits you could win every run without getting any upgrades or whatever at any point (that would be incredibly tedious, though, if you never upgraded your weapons as part of a given run -- but not upgrading your health any would certainly be possible yet tense).
Regarding genre labels:
10. Overall we have to choose terms that match as closely as possible so that people have some idea of what the heck they are looking at at all. If I tell you that a game is "blue" or "it's a game," then you have no idea what I mean.
11. If I get TOO specific to the point that you are misled, that's a problem obviously. But I don't think most games do that. Saying something like "this is a clone of Rogue, but minus these elements and plus those" would be really descriptive, but obviously a lie if that wasn't an accurate descriptor. Also, the odds of a game being able to be described as "Triple Town but with a snow theme" is very... questionable. That rarely happens with legit games.
12. Beyond that, getting as close as possible with subgenre labels that mostly-fit-but-not-completely is the best that anyone can do. Genre labels are easy: people agree that if you are holding a gun and it's a first person view, that's a FPS game. Although actually even the strategy and simulation labels can be contentious: what degree of simulation do you have to be before it's simulation? How much focus has to be on strategy before it's strategy? Anything with levels and EXP is an RPG? Etc.
13. Subgenre labels will usually be only partly-correct unless a game does nothing that is too novel. How realistic do the controls have to be for a game to be a legitimate flight simulator? If I make a very arcade-y flight sim that would be accessible and fun for non-purists of that genre, but that would be a huge disappointment for someone looking for the true pilot experience, did I describe it well or poorly? Probably the best descriptor there is "an arcade-y flight sim."
In other words, to sum up:
14. I think that devs almost never give a game a given description in order to deceive you. That would be incredibly stupid, because unhappy customers is the last thing anyone wants -- at least if they plan on making more than one game and then disappearing into the night.
15. Describing games with the right amount of precision that people can find things that interest them means going a bit too broad sometimes. If there are things that are kinda-sorta-roguelikes, don't you want to know about those? I know that I do. It's easy for me to read more about a given one and then hit Not Interested if it's not my bag. And a ton of them aren't. But they were definitely in the realm of what would scratch my roguelike itch, so the tag did its job. I wasn't getting gas stations when I search for restaurants or something (thanks Yelp).
16. Overall that's kind of what it comes down to. If I get gas stations when I look for restaurants, that's a problem. If I get McDonalds and Ruths Chris in the same list... well, I think that's valid. Both are restaurants, despite one being fast food (and both being an immediate "no" if I'm looking for typical medium-priced food options). Finding Rogue Legacy in your list of roguelikes is I think like finding a Chilis and going "hey I just wanted non-chain restaurants!" or something. I won't do them the discredit of comparing them to Micky D's. But basically I get that you think roguelike means something a lot less mass-oriented, by the very nature of part of its definition. So that's going to be something you just kind of have to sort through. The alternative is that you might miss out on some really cool games every so often.
17. The other alternative, of course, is to come up with constant new genre labels. If we say that something is a "castle intrusion upgrade" game, what the heck does that mean, though? Is it some kind of stealth game? What exactly am I upgrading there? Do you mean a literal castle, or just some sort of stronghold? And imagine being presented with another 40-such labels that you suddenly have to remember what each means. That quickly becomes less meaningful to you than subgenre labels that only sort of fit, or fit but kind of look wonky while being worn by the game in question.
18. Or we could just do away with labels entirely and go witih two-page descriptions of every game. But people realistically would read a line or two of that and form an initial impression based on that whether to read further a lot. So those first lines would get more and more symbolic and jargon-y to the point that we are back to labels with then the added description that is longer. ...Which is the system that we have now!
19. In short I think that the system we have now (as an industry) is pretty good. New subgenres would be nice to add a bit more often than we do, perhaps. And sometimes a game gets a bit off-categorized, but usually that opinion is not universally held on a given game. Sometimes specific devs could do a better job of categorizing a given game. But ideally in those instances the description text and/or bullet points helps to clear up any confusion that might result.
And... whew. Back to work for me.
