That's weird that your uni professors didn't mention forced choice pitfalls and double barreled questions. I used to be a research methodologist and statistician in a previous career, but this was stuff I learned in undergrad.
I wonder where you found that those were the genres that sell relevant numbers. There are a few genres I think you left out (i.e. pure sci-fi, cosmic horror, etc.) that have done well over the years.
I'd recommend
this for reading. I think they are a good example of a business that "succeeds". In general, if you hit the mark, you maybe have a few good years and then kinda fizzle, or pursue other revenue streams. DP9 went in the direction of minis games. Steve Jackson Games did the same with a variety of other (non-TTRPG products). As for the big kahuna of Wizards, well, D&D is a rather small chunk of their revenue, I'm sure. They could just straight up kill D&D tomorrow and they would be fine. Chaosium and Grey Ghost are other companies worth studying, though I'm pretty sure they are both run out of someone's basement at this point.
I'd also recommend reading up on
Graham Walmsley. He did something rather clever and original. He built an extremely rules-light game for Lovecraftian horror short play (generally one-shots). I mean, really, it was a lunchbreak RPG. He put it (and even some supplements) out there for free. Then, he created a commercial product bringing it all together with some additional material and kickstartered it. He funded for 25 days and drummed up almost 72k British of his 6k goal. This was far from his first (even published) project, though. But, he put something cool together to whet people's appetites (and also gauge the reception), then threw together a commercial project once he had a sense how it was going to hit.
I would be remiss if I didn't mention Vincent Baker, too. This is an example of another "successful" rpg designer. He had two hits with
Dogs in the Vineyard and
Apocolypse World (which, clearly you're familiar). I wouldn't even consider Dogs in the Vineyard a game, but more of a collaborative IF thing (like Fiasco just clunkier due to the pseudo-poker resolution mechanic. AW is ok. Personally, I think the bastard child Dungeon World was a much better designed game. But AW did make a chunk of change (twice as much as DW), so if that's what matters. Then again, there was also a cult of personality that surrounded Vincent Baker and pals that likely helped the numbers. Controlling for that, I would actually say that DW was more successful, as was Cthulhu Dark, for that matter.
I would also recommend
this. In terms of games where the designer sets a goal and designs a game to meet that goal, this is one of the best games I've ever seen. You probably never heard of it, but that's how it goes. And no, I'm not Simon Washbourne.
Lastly, I'd take a look at
this. Steve Jackson is an OG. Like, he probably goes as far back as Gary Gygax. Notice the section on "Breaking In". He has some good tips about making a name for yourself through blogs and other online media.
I absolutely love gaming and RPGs and have been in the hobby for decades. I've been designing for probably 15-20 years. At no point did I have the intention of becoming a "professional game designer" (I can proudly say that nobody in this hobby except my closest friends even knows who I really am), and I think it has become less and less viable over the years. Even when you break down what a self-publisher makes to an hourly rate, I'm pretty sure I get paid more to take a crap at work (and I'm not a millionaire or anything). The hobby itself is probably dying. Between a growing interest in digital mediums and the various sociopolitical issues that divide the sub-culture, people question whether the hobby can even survive, nevermind have any commercial value. It's kinda sad really. I'd much rather sit around a table with a bunch of pals drinking some beers and going on an adventure, than sit alone on my couch with a headset mindlessly blowing away aliens or whatever.
Anyway...
Based on the mechanics you posted, the first sounds kinda like d20 with crit successes and fails. Also, 1s and 20s hit on the rare side (there's actually a study out there that showed most d20s manufactured skew towards the middle, actually, due to construction and design). Not that a d20 game is a bad thing, but basing a mechanic around something that happens rarely isn't going to pack much of a punch. I think everyone has their own threshold for this, but (for me) anything less than a 10% likelihood is not worth the effort.
I'm not sure if I would call AW an algorithmic system. It's more like, each character template has a bunch of vague, broadly applicable abilities. Almost anything you can think of to do should fit one of them. The resolution is also far from concrete. It's basically, roll low and bad stuff happens, roll high and good stuff happens, roll in the middle and it gets complicated (and it's up to the group to decide what these vague things mean). All in all, I wouldn't say it's much meatier than Dogs and would probably still consider it to be collaborative IF. I would recommend taking a look at Dungeon World if you're thinking this direction. It's probably the most sound usage of the AW resolution mechanic.
Two dice have a stronger central tendency than one. However, I wouldn't call one die swingy. When you figure that most systems have a binary pass/fail resolution mechanic, you have a simple binary probability of success at various TNs. This is a double edged sword (and possibly why people who rebel against d20 for this reason don't often succeed). Having a stronger central tendency means a smaller range of viable TNs. Consider 1d20 vs. 2d10. 1d20 moves in flat 5% increments. Going from 1-2 means the same as going from 11-12. However, with the 2d10, going from 10-11 is the same as going from 2 to 6. I mean, if that's what you're going for, then great, but I'm not confident that many people who do this do it intentionally.