Romancing SaGa 2: Revenge of the Seven

Most JRPG remakes are exercises in embalming. Square Enix has spent the better part of a decade excavating its own back catalog with varying degrees of reverence, and the pattern is familiar: polish the sprites, patch the balance, leave the design philosophy untouched under glass. Romancing SaGa 2: Revenge of the Seven is a different animal, mostly because the thing being preserved was never safe to begin with.

The 1993 Super Famicom original — directed under series creator Akitoshi Kawazu’s stubbornly anti-commercial design ethos — was a Japan-only release built on a premise that no focus group would ever approve. You don’t play a hero. You play the Empire of Varennes, a political entity that outlives every character you control. Your emperor dies, or a generation simply ends, and the crown passes to an heir who inherits the accumulated combat techniques and spells of everyone who came before. The protagonist is disposable. The institution is not.

This remake, developed by Square Enix in partnership with xeen Inc. and released across Steam and Nintendo Switch on October 24, 2024 for $49.99, is a full 3D reconstruction rather than a texture pass. It landed at 82 on Metacritic.

That number tells only half the story. Graded against modern production values, this reads as competent-not-lavish. Graded against the SaGa series’ reputation for being magnificently user-hostile, it amounts to something close to unthinkable: a SaGa game that can be recommended to someone who has never played one.

• Generational Succession: When an emperor falls or an era closes, you crown a successor who inherits the empire’s accumulated skill and spell library. Progress is institutional, not personal — a permadeath system that costs you a character but never costs you time.
• Glimmers (Spontaneous Technique Discovery): Characters invent new attacks mid-combat, unprompted, signaled by a flash of insight. There’s no skill tree. Your arsenal grows through pressure, not menus.
• Use-Based Growth, No Experience Levels: Stats improve through what you actually do. Swing a sword, gain strength. There is no XP bar, and no “grind for twenty minutes to clear the wall” escape hatch in the traditional sense.
• Timeline Combat with Formations and United Attacks: Turn order is visible and manipulable. Formations grant positional bonuses that reshape a party’s identity, and United Attacks chain party members into combo strikes when the timeline aligns.
• Empire Management: Between expeditions you annex territory, fund research, and unlock new classes from a roster of 30+ recruitable job types, each with distinct weapons and abilities.
• Presentation Package: Full English and Japanese voice acting, Kenji Ito’s rearranged score with a toggle back to the original soundtrack, New Game+, and Casual / Normal / Hard difficulty tiers.

The Gameplay Loop

Here’s what makes this structurally strange, and why it works.

In a conventional RPG, the loop is: fight, gain XP, level up, fight harder things. Character growth is the reward, and character death is a punishment that reverses it. Romancing SaGa 2 severs that link entirely. Your emperor will die. Sometimes because you lost a fight. Sometimes because the story advances a generation and the timeline moves on without asking. Either way, the successor inherits the empire’s technique library and you continue, materially stronger than the person you just buried.

This is the design’s central trick, and it is genuinely radical even now. Death stops being a load-game event and becomes a resource-management decision. Do you push this emperor’s aging body into one more confrontation with a Seven Hero, or do you consolidate, expand a few territories, fund research, and hand a stacked deck to the next ruler? The game never forces the answer. You choose which of the Seven Heroes to hunt and in what order, and the world reshapes around your sequencing.

The Glimmer system is the other pillar, and it’s a sharp expression of “learning by doing.” Techniques arrive when you’re up against something difficult — a lightbulb over a character’s head mid-swing, and suddenly a move exists that didn’t three seconds ago. It converts hard fights from obstacles into training grounds. You don’t beat the wall by grinding weaker enemies; you beat it by repeatedly throwing yourself at the wall until your party invents a way through. That’s a fundamentally better feedback loop than an XP bar, and the genre should have stolen it thirty years ago.

Combat itself sits on a visible timeline that removes the guesswork from turn-based tactics. You can see the sequence, which means Formations and United Attacks stop being cosmetic and start being planning tools. Positioning a fragile mage in a defensive formation while queuing a three-person chain strike is real tactical work, not a stat check dressed up as strategy.

Onboarding Friction — Fixed, But Only Halfway

The remake’s most valuable addition is boring and essential: a world map with legible objectives. The original expected you to derive the entire structure of a continent-spanning campaign from ambient dialogue. This version tells you where to go. Combined with the Casual difficulty tier, it’s the reason this is now widely — and fairly — called the most accessible entry point into the series.

But the guidance stops at geography. The systems underneath remain sparsely explained. The relationship between weapon usage and stat growth, the specific conditions that trigger Glimmers, how empire research actually converts into class unlocks — these are all still things you infer, or look up. The game modernized its map and left its manual in 1993. For a certain player that opacity is the appeal, the pleasure of a system that respects you enough to stay quiet. For everyone else, it’s a wiki tab open on a second monitor.

The Difficulty Problem

This is the sharpest legitimate criticism. The difficulty curve is lumpy. The mid-game is too forgiving — the quality-of-life additions and the compounding technique library combine to flatten a long stretch where the original applied constant pressure. Then the endgame spikes, hard, and players who coasted through the middle hit a wall they have no muscle memory for. It tracks with how the systems compound.

There are also grind-heavy stretches, which is a peculiar failure in a game whose entire design thesis is that grinding shouldn’t work. When the fastest path forward becomes repeating encounters to fish for a Glimmer, the elegance leaks out of the machine.

The art direction is functional rather than lavish, and precision matters here. This is not an ugly game. The chunky, slightly toy-like character models and clean environmental geometry read as a deliberate translation of 16-bit sprite work into three dimensions — closer to a well-made diorama than a prestige production. It’s coherent. It’s also clearly working within a budget that Square Enix reserves for its mid-tier revivals rather than its flagships.

Where the presentation does best is audio. Kenji Ito’s rearranged score is strong, and the decision to include a toggle back to the original SNES soundtrack is the kind of small, confident choice that signals the developers understood who this was for. Full English and Japanese voice tracks are a substantial addition to a game that originally had none.

The UI is the quiet MVP. Formation selection, technique management, and the empire-development screens handle an enormous amount of state without becoming a spreadsheet. Given that this game tracks inherited skills across generations and 30+ classes, that’s real interface engineering.

Steam is the fidelity-first option, unsurprisingly. A 2D-rooted game of this scope is light on hardware, and the menu-heavy design suits mouse-and-keyboard. If you’re the type who’ll be cross-referencing technique lists in a browser, having them one Alt-Tab away is a meaningful convenience for a game this under-documented.

Nintendo Switch is the better fit — and, by most accounts of the port, the weaker build. The trade-off is the usual one: a softer resolution and more conservative effects. What you get in return matters more here than in most games. This is a title built around long, self-directed sessions punctuated by natural stopping points — an era ends, a successor is crowned, you put it down. That rhythm belongs to a handheld. Combat is entirely turn-based, so nothing about the design depends on input latency or frame consistency in the way an action RPG would.

Pick Steam for fidelity. Pick Switch for the rhythm.

The 82 Metacritic average is defensible and slightly unfair. Graded on production values, this is a mid-budget remake with modest visuals and a difficulty curve that needed another balance pass. Graded on design, it’s carrying an idea that no other RPG has successfully copied in three decades: that progression can belong to an institution instead of a person, and that a character’s death can be a transaction rather than a tragedy.

This is not a game trying to be for everyone, and the remake’s real achievement is that it added an on-ramp — a world map, a Casual mode, a legible UI — without filing down the teeth underneath. Square Enix and xeen could easily have made this friendlier by making it duller. They didn’t.

The mid-game slack is real and the endgame spike will lose people. But when the systems align — a Glimmer fires mid-fight, your formation holds, a United Attack chains through a Seven Hero you’ve been circling for two generations — it produces a specific satisfaction that no leveling curve can manufacture. You didn’t grind for that. Your empire learned it.

We recommend it to the audience it was built for. Then hand the crown to someone else.