On the one hand, the game has a solid story, a quirky cast of characters, amazing animation, and a unique real-time-turn-based combat system that I'm sure someone other than me would probably be really into (I mostly just used the characters that did the most damage with the least amount of fuss).

On the other hand, the platforming sucks.

And I get it, at some point if I complain about games having bad platforming enough it starts to seem like maybe the problem is I suck at platforming. But I really don't think that's the problem here. I have liked platforming games in the past, after all. I really liked Celeste, and while part of why I liked it was because of the Assist Mode feature I only used it a handful of times. And I play my favorite platforming game, Big Tower Tiny Square, once or twice a year.

What those games have in common is simplicity. Celeste has basic side-to-side movement, a jump, a dash, and wall-climbing. Big Tower Tiny Square is even more barebones, just movement and a jump (and I guess wall-jump if you want to count that as a separate thing). This simplicity in the controls lets you concentrate more fully on navigating the increasingly complex environments.

Indivisible, meanwhile, has decided to pair its platforming with metroidvania-style skill acquisition. What this means is you'll spend the first half of the game collecting abilities that make platforming not easier but more complicated, as each of these skills is bound to only one of the main buttons, with control-stick direction acting as the modifier. At a purely physical level, this increases the odds that in any given moment you'll accidentally use the wrong ability (with potentially disastrous results). Maybe more important, though, is the cognitive load. In some of the more difficult platforming sections, where you're having to execute a wide variety of moves in quick succession, my brain would just short-circuit as it tried to remember what all the moves were, let alone execute them correctly.

None of this is helped by an incredibly floaty and imprecise physics engine that makes it nearly impossible to replicate moves from one attempt to the next, meaning you can't get useful feedback about what you did wrong because you could do exactly the same things next time and get a completely different result. They also make the inexplicable choice of completely killing your momentum when you land on a platform, which makes the simple act of jumping up a series of platforms feel unexpectedly clunky and challenging.

So I don't think this is a situation where I just need to git gud or shut up. I think there is a fundamental flaw in the design philosophy here that, when combined with less-than-stellar execution, led to a nearly unplayable experience, particularly in the game's final area (that is to say, all three times you tackle the games final stage). And it sucks because I really like everything the game is doing on an aesthetic and narrative level—but not enough to recommend putting up with its mechanical failures. (Still couldn't bring myself to give it 2 stars though.)

p.s. Two really smart choices they make with the story (with very big spoilers): As Ajna is going around trying to save the world she leaves a trail of destruction in her wake, and her friends keep trying to get her to change her approach but to no avail. Finally, when you get to what is supposed be the final boss fight, Ajna loses control and is about to basically destroy the world when of her friends sacrifices himself to stop her. Smart choice #1 is that Ajna then revisits the three main areas of the game where she'd previously intervened so recklessly and actually works with some of the people there to help fix things. Smart choice #2 is that her friend (who also wasn't exactly her friend, given that he murdered her dad, but still) who sacrificed himself, who to be clear is a playable character and was actually a mainstay of my party right up until that point, stays dead. It's a really well-handled character death: earned by the story, fits the character arc, isn't an arbitrarily shocking twist, doesn't rely on arbitrarily removing the player's agency, has actual important consequences for the story and the main character, and is not undermined by a miraculous resurrection an hour later.

p.p.s. There were several locations in the game that you can only get to right at the end, and you have to go through some really long and difficult platforming sections to get to, and then you get there and there's just nothing there. Like literally nothing. Just an empty room. Most of them were big rooms that felt set up for a boss fight, but there's no boss, though one of them was on top of a mountain so at least there was a view. I don't know if these were part of some sidequest(s) that I just completely missed or if they were intended to contain something that just never got implemented, but they were really annoying.

p.p.p.s. The map is also a little janky, occasionally implying that there's a passageway where in reality there's just a wall. It also doesn't distinguish between the down smash green barriers and the upsmash green barriers, so it will show a barrier as being passable when it in fact is not.

p.p.p.p.s. Also the fact that the combat takes place in the real environment has some unpredictable consequences, such as melee characters just falling off a ledge and leaving combat. I don't know if that's necessarily a good or a bad thing but it's notable.