Bottom Line: A monstrously large, genuinely strange visual novel with a competent tactics game bolted to its side — and one of the few games in years where "100 endings" is a structural thesis rather than a marketing lie. It demands more hours than most people have, and rewards nearly every one of them.
The Gameplay Loop
The loop is a metronome: story chapter, downtime, battle, repeat, across a 100-day countdown that the game keeps visibly ticking. It's clean, legible, and after twenty hours, it's also the source of most of the game's problems.
Combat is the weakest limb, and it's worth being precise about why — because "the tactics are bad" is the lazy read, and it isn't accurate. The tactics are fine. Maps are readable. Voltage economy creates a real tension between spending now for a clean sweep and banking for a boss phase. Specialist Skills are meaningfully differentiated; several characters are less units than puzzle keys. When a chain detonates across a packed formation and clears twelve enemies in one turn, the feedback is excellent.
The problem is that the encounter design never develops. It's built for power fantasy throughput, not tactical pressure — you're mowing, not maneuvering. That works beautifully for the first act, where it reads as escalation and the numbers going up feel earned. By the fourth or fifth pass through similar map layouts on a replay route, it reads as toll collection. The critical consensus lands on "serviceable rather than exceptional," and that's the honest verdict. This is a combat system that supports the narrative without ever competing with it.
Which is, arguably, the point. But it makes the route grind the game's real design risk.
The Structure Is the Argument
Here's where the game earns its reputation. The branching isn't a feature layered on top of a story — it is the story. The academy's mystery is deliberately unsolvable from any single route, and the game is explicit about this. You are meant to die, restart, and carry knowledge across attempts. Uchikoshi has been building toward this shape his entire career; here he has the budget and the collaborator to push it further than Zero Escape ever could.
Be clear-eyed about the number, though. 100 endings does not mean 100 full campaigns. The distribution is heavily tiered: a substantial tail consists of short, punchy failure states — some of them jokes, some of them genuinely upsetting — reachable in minutes from a branch point. The main branches, the ones that carry real narrative weight, number closer to a dozen. Marketing will let you believe otherwise. Anyone selling this as a hundred discrete forty-hour routes is misrepresenting it.
That's not a con. The tiering is smart. The short endings function as probes — cheap experiments that teach you the rules of the world, so the long routes can assume literacy. But it changes the value calculation, and you should know before you buy.
Onboarding Friction and the Opening Problem
The first act is slow. Not slow-burn slow — slow. It spends its opening hours establishing fifteen characters through a battery of anime archetypes played almost entirely straight, and there is a real risk you bounce before the game starts dismantling them. Critics flagged this near-universally, and the complaint is legitimate. The payoff is real and the setup is doing necessary work, but "trust me, it's good at hour fifteen" is a hard sell at full price.
The trope reliance persists past the opening, too. If the loud-genki-girl, the stoic-swordsman, the perverted-comic-relief register as nails on a chalkboard for you, this game will not convert you. It deploys those types knowingly — subversion is the whole engine — but it deploys them constantly, and knowing self-awareness doesn't make a bad joke land.
Pacing Across a Hundred Hours
The single sharpest structural critique: the game does not adequately compress on repeat. Skip functions exist, and they work. But the ratio of new content to content you're skipping past to reach new content degrades in the back half, and the mandatory combat between narrative beats means you can't simply fast-forward the text and move on. You have to fight your way back to the fork. For a game whose entire premise is replay, that's a real friction cost — the one place where the design's ambition outran its ergonomics.
