Endless Legend

Most 4X games are the same game wearing different hats. You settle, you expand, you research the same tech pyramid, you argue with an AI that either grovels or declares war for reasons it can’t articulate. The factions differ by a starting bonus and a unit skin. Everyone knows this. Almost nobody does anything about it.

Endless Legend, released September 18, 2014 by Paris-based AMPLITUDE Studios, did something about it. The setting is Auriga, a planet that has already ended several times and is queuing up for one more — a slow apocalypse of encroaching, permanent winter that functions as a soft turn timer on your entire civilization. That’s not flavor text. It’s a load-bearing mechanic, and it reframes the genre’s usual leisurely 300-turn sprawl into something closer to a race against thermodynamics.

Amplitude came into this with a reputation. They’d already built Endless Space and Dungeon of the Endless, sharing lore, a resource called Dust, and a peculiar house style: heavy art direction, heavier systems, and a community-voting platform (Games2Gether) that let backers pick faction designs. That last part shows. Endless Legend feels designed by people who play 4X games competitively rather than by people who studied them.

The market position is unusual. This isn’t a Civilization competitor — it doesn’t want the mainstream. It’s the game Civ players graduate to when they’ve internalized every Civ V opening and want their assumptions broken. Metacritic sits at 82. That’s a strong number for a game with this much friction, and one that has held up over more than a decade, which matters more. Amplitude has since moved on to Humankind and a sequel, but the original hasn’t been rendered obsolete by either. Very little has caught up to it.

• One City Per Region: Territories are pre-drawn on the map. Each supports exactly one city, which grows outward through specialized districts. Expansion becomes a land-grab for good regions, not a race to spam settlers into every empty tile.
• Eight Asymmetric Factions: Not stat tweaks — rule rewrites. The Roving Clans are physically incapable of declaring war. The Broken Lords are animate armor who consume Dust instead of food, converting currency directly into population. The Necrophages can’t do diplomacy at all. Each has a bespoke questline that doubles as a victory path.
• The Winter Clock: Seasons cycle, and winter gets progressively longer and more punishing. Production drops, movement suffers, and the endgame arrives whether you’re ready or not.
• Individual Unit Equipment: You craft weapons and armor from the strategic resources you actually control. A titanium shortage isn’t an abstraction; it’s why your spearmen are still carrying iron.
• Minor Faction Assimilation: Villages scattered across regions can be pacified or razed. Assimilate them and you gain their unique units plus an empire-wide bonus, turning the map’s neutral clutter into a real strategic layer.
• Influence as a Currency: A third resource alongside Dust and science, spent on diplomacy, empire-wide policy edicts, and pressuring rivals. It makes the non-military game feel like a game.

The Gameplay Loop

Here’s what Endless Legend understands that its peers don’t: scarcity makes decisions interesting, and abundance makes them arithmetic.

The one-city-per-region rule is the cleanest expression of this. In Civ, the correct answer to “should I found another city?” has been “yes” for thirty years. Here, regions are finite and unevenly valuable, so every claim is a commitment. You’re not optimizing a settler queue; you’re evaluating terrain, luxury deposits, defensibility, and whether the neighboring region’s minor faction is worth assimilating. When you do plant a city, it grows through districts placed on specific hexes — adjacency matters, terrain matters, and a city sprawling across a river delta plays differently from one wedged into mountains. Civilization VI shipped a version of this idea two years later to considerable acclaim. Amplitude got there first, and arguably got there better, because the district system is constrained by the region border rather than by unlimited map space.

Asymmetry as Design Philosophy

The factions are where this game earns its reputation, and it’s worth being precise about why. Plenty of strategy games advertise “unique civilizations” and deliver a percentage modifier. Endless Legend takes things away from you.

The Roving Clans can’t declare war. Full stop. That’s not a debuff — it’s a different game, one where you win by monopolizing the marketplace, buying mercenaries, and using economic leverage as a weapon. The Broken Lords don’t eat, so food production is worthless to them and Dust becomes population, military upkeep, and healing all at once; their entire economic curve is shaped differently. The Cultists get exactly one city, forever, and expand by converting minor faction villages into satellites instead. The Necrophages can’t negotiate, so their strategic horizon is permanently military.

Learning a new faction means unlearning your habits. That’s rare, expensive design, and it’s the reason people still play this in 2026. It also means the difficulty curve is jagged — a third faction can feel harder than the first, which is a legitimate onboarding friction problem the game never fully solves. The tutorial is adequate for mechanics and useless for strategy.

The Battle Layer Problem

Now the part that’s harder to praise.

Combat resolves on a simultaneous-turn tactical layer: armies deploy onto a hex battlefield lifted from the world map, orders are issued, both sides execute at once over six rounds. Terrain grants bonuses, initiative determines strike order, and heroes anchor the line. On paper, it’s a smart middle ground between Civ’s abstract stack-bumping and Total War’s full RTS.

In practice it’s thin. Unit orders are coarse — mostly a target and a facing — and the simultaneous resolution means predicting the AI rather than outplaying it. Once the battles blur together, the tactical decision space collapses into “do I have more Dust in my equipment than they do.” The result is predictable: the system is built to be skipped, and auto-resolve is generous enough to make skipping it rational — a damning verdict on a layer that occupies a third of the game’s runtime. It’s not broken. It’s just not interesting enough to justify the clicks.

Late-Game Drag and the AI

Two more failures, both familiar to the genre and neither excused by that familiarity.

The AI is weak. It plays a competent early game and then loses coherence — it under-expands, mismanages armies, and misreads the winter mechanic entirely. Higher difficulties don’t make it smarter, they hand it resource bonuses, which produces the usual outcome: an early game that feels threatening and a mid-game where the outcome is already decided with ninety turns left.

And those turns are slow. By turn 150 the game asks for management of a dozen cities, each with a district queue, plus hero assignments, plus per-unit equipment loadouts that need refreshing every time a better alloy unlocks. The UI is not built to scale to this. There’s no meaningful automation, no governor delegation worth using, no bulk-equip. The systems that make the early game rich become an inventory-management chore late, and that’s where a lot of campaigns get abandoned.

The art direction is the strongest thing about this game after the factions, and it’s not close.

Auriga is rendered in a stylized, low-saturation palette — pale ochres, cold blues, bone whites — that reads as a world already past its prime. Cities look like illustrations. Unit design commits hard to each faction’s identity: the Broken Lords are hollow gilded armor, the Necrophages are chitinous and wrong, the Drakken are alien rather than “elves with scales.” The winter transition is the standout, with the whole map desaturating into snow as the season advances. It’s a visual mechanic, not a filter.

FlyByNo’s soundtrack is excellent — ambient, mournful, structurally patient. It’s one of the few 4X scores people listen to outside the game.

The UI is the weak link, and it’s a familiar Amplitude weakness: gorgeous, and legible only after it’s been learned. Iconography is dense and inconsistently sized, critical numbers hide inside tooltips-within-tooltips, and the city screen buries district adjacency information needed on every placement decision. Form beat function here, and it stayed beaten across six expansions.

Technically, a 2D-styled game of this scope is light on hardware — a 2014 Unity title that asks little of a modern machine. The bottleneck is design, not specs: late-game turn times climb badly on large maps with eight opponents, because the AI logic is single-threaded, and no amount of CPU fixes that.

Steam, mouse and keyboard, and that’s the honest scope of it. There’s no controller support worth discussing and no meaningful reason to want it — this is a game of hover-states, tooltips, and precise hex clicks.

What Steam actually delivers here is Workshop modding, and it’s substantial. The game’s data lives in editable XML, which put real modification within reach of people who aren’t programmers. Total conversions, faction packs, rebalance mods, and UI fixes all exist, and the modding framework is built to accommodate them. If the vanilla late game wears thin, the community has patched around it.

Multiplayer is the caveat. Simultaneous turns keep sessions moving faster than you’d expect, but the netcode is fragile, desyncs are a long-standing issue, and a full eight-player game is a scheduling commitment measured in evenings. Treat it as a two-to-four player experience with people who can be trusted to reconnect.

The DLC situation deserves a blunt note: the base game is complete, but the expansions — Shifters, Shadows, Tempest, Inferno, Symbiosis — each add factions and systems that meaningfully deepen the whole. Buying the base game alone in 2026 is a mistake when the complete bundle regularly drops to a fraction of list price. That price makes the value question trivial.

Endless Legend is a flawed game that matters more than most polished ones. Its combat layer is mediocre, its AI is a pushover, and its late game asks for administrative labor it should be automating. Those are real problems, and they’re the kind that end campaigns.

But the parts that work are the parts almost nobody else attempts. Eight factions that require eight different mental models. A map whose geography is a fixed constraint rather than empty space. A world that’s actively dying while you argue about trade routes. Amplitude took the biggest structural risks in the genre and landed most of them, and a decade later the industry has borrowed the districts and skipped the asymmetry — which tells you which idea was harder.

We rate it for the design, not the fighting, and recommend it to strategy players willing to forgive its friction. At current prices, the value proposition is hard to argue with.