Mount & Blade II: Bannerlord — the sandbox strategy people modded forever

TaleWorlds Entertainment, a studio out of Ankara, entered Bannerlord into early access in March 2020 and didn’t call it finished until October 2022 — a two-and-a-half-year public build cycle for a game that was already a sequel to a decade-old cult favourite, Mount & Blade: Warband, and a prequel to it in setting. That’s an unusually long runway even by early-access standards, and what it produced is a game whose actual selling point was never really “finished” in the traditional sense at all: a combat model and a political sandbox robust enough that the modding scene treated the base game as raw material rather than a product to simply play as shipped.

Combat you have to aim, not just trigger

The mechanical decision that separates Bannerlord — and Warband before it — from almost every other melee combat system in strategy or action gaming is that hits aren’t resolved by a dice roll against a stat sheet. Every swing has a direction chosen by mouse movement, a windup and release timing the player controls, and a resolution that depends on weapon reach, momentum, and the relative angle between attacker and target. A couched lance charge at full gallop does genuinely more damage than the same weapon swung from a standstill, because the game is modelling momentum as a real physical quantity rather than a flat “mounted bonus” multiplier tacked onto a damage formula. Blocking works the same way — you aim a block in the direction of an incoming swing rather than pressing a generic “block” button — which means two players of genuinely different skill will have visibly different survival rates in identical gear, in a way a hit-chance-and-armour-value system can’t replicate.

That combat model sits inside a genuinely open political sandbox: no scripted main quest forces you down a single path, and a campaign might just as easily end with you as a mercenary captain for hire, an independent bandit-hunting freelancer, a vassal climbing a kingdom’s feudal ladder, or the ruler of a breakaway faction carved out of one of the map’s warring kingdoms through a mix of diplomacy, marriage alliances, and outright conquest. Clan management, kingdom politics, and siege engineering all sit on top of the same combat engine, so a siege assault resolves through the same momentum-and-positioning combat as a field battle, just staged against walls, ladders, and siege towers instead of open ground.

Why the modding scene adopted it as infrastructure

The historical setting itself deserves a word, since it’s easy to assume “generic medieval fantasy” from the marketing alone. Calradia’s factions are each built as a lightly fictionalised analogue of a real medieval culture rather than an invented one — armour silhouettes, unit names, and architectural styles are drawn closely enough from Norman, Rus, Arabian, steppe-nomadic, and Celtic historical reference that a player with any interest in medieval military history will recognise the source material even though no faction claims to be the real thing. That grounding shows up mechanically too: Khuzait horse-archer tactics genuinely reward hit-and-run harassment rather than direct engagement, in a way that reflects steppe cavalry doctrine rather than treating “cavalry” as a single undifferentiated unit category the way a lot of strategy games do.

TaleWorlds shipped Bannerlord with an unusually accessible modding toolset relative to its complexity, and the community response wasn’t just cosmetic reskins — full total-conversion mods restaging entire fictional settings on top of the engine, and multiplayer-focused persistent-kingdom mods that turned the base sandbox into an ongoing, player-run political simulation running for months at a stretch. That’s a similar trajectory to what happened with Cities: Skylines’s Workshop ecosystem in a completely different genre: ship a simulation robust enough that its underlying rules reward tinkering, and a community will extend the game’s actual lifespan well past what the base content alone could sustain.

The sandbox-career structure Bannerlord runs on didn’t originate with TaleWorlds, either. The clearest ancestor sits well within Jay’s own lived memory: Sid Meier’s Pirates!, MicroProse’s 1987 release, established the template of a player character free to choose piracy, mercenary work, trade, or a noble marriage, building a reputation and a fleet across an open Caribbean map with no single mandated path to victory. Bannerlord is that same open-career promise relocated to a fictional medieval continent and rebuilt around personal combat skill rather than ship-to-ship broadsides, but the underlying design conviction — let the player choose what kind of career this campaign becomes, rather than script it — is the same one Pirates! proved could work over three and a half decades earlier.

Sieges as the combat system’s hardest test

Bannerlord’s siege battles are where the momentum-and-positioning combat model gets stretched furthest, because a siege isn’t just a bigger field battle — it introduces vertical combat via ladders and siege towers, chokepoints at gatehouses where a handful of defenders can hold a much larger attacking force if positioned correctly, and siege engines (trebuchets, ballistae, battering rams) that have to be built or hauled into range and defended while they work. A defending garrison with a well-placed ballista covering a ladder approach can inflict disproportionate casualties on an attacking force before melee ever begins, which means a siege’s outcome is frequently decided in the approach phase rather than in the wall-top melee that follows — a direct consequence of the same momentum-and-angle combat model that makes field battles reward positioning over raw numbers.

Clan and party management sits underneath all of this as the layer that determines whether a player even has the troops to attempt a siege in the first place. Companions recruited across taverns and quests can be assigned to lead their own parties, garrison towns, or govern fiefs, and the perk system attached to each of a character’s skills — one-handed weapons, tactics, leadership, engineering, and more — lets a long campaign specialise a ruler’s personal strengths in a way that meaningfully changes what kind of war they’re equipped to fight, whether that’s a siege specialist commanding engineering bonuses to build works faster, or a cavalry-focused warlord whose personal build rewards mounted charges over infantry lines.

The parts still catching up

Bannerlord’s full 2022 release closed most of early access’s rougher edges — diplomacy between AI-controlled kingdoms, which had been thin for most of the early-access period, is deeper now, and the persistent bugs around companion inheritance and clan succession that plagued early builds are largely resolved. What hasn’t fully caught up is the strategic-layer AI’s tendency to snowball: a single kingdom that gets an early advantage in territory can end up steamrolling the map’s diplomatic balance faster than the game’s political systems can meaningfully check it, which is a genre-wide problem for open political sandboxes rather than one unique to this engine, but it’s more visible here because the combat layer is detailed enough to make every individual siege feel consequential even when the strategic outcome already feels foregone.

The smithing loop as a second sandbox inside the first

One system worth calling out on its own is Bannerlord’s weapon-smithing minigame, which lets a character with enough smithing skill design custom weapons from individually unlocked component parts — different blade shapes, guards, grips, and pommels, each affecting weight, reach, damage type, and handling independently — then forge and sell them for profit or equip them personally. It’s close to a small crafting game bolted onto the larger political sandbox, and it became one of the more surprising ways players found to fund an early campaign without relying on trade caravans or battle loot, since a well-designed two-handed sword built from unlocked high-tier parts can sell for a substantial sum once a character’s smithing skill and unlocked part list are deep enough. It’s also one of the clearer examples of the game rewarding a genuinely different playstyle investment — a character who never fights and instead sinks skill points into smithing and trade can still build significant wealth and influence, which is consistent with the sandbox’s broader refusal to mandate combat as the only path to power.

Spoilers below

There’s no fixed campaign to spoil in the traditional sense, since Bannerlord’s story is generated by whichever path a player chooses, but the game’s scripted background lore is worth knowing before diving in: the setting depicts the Calradic Empire in a state of terminal civil war between three imperial claimant factions in the years following the death of the previous emperor, with five additional independent kingdoms — Vlandia, Sturgia, the Aserai, the Khuzaits, and the Battanians — carved from or bordering the fracturing empire, each modelled on a real historical culture (Norman, Rus, Arabian-Bedouin, Mongol-steppe, and Celtic respectively) rather than invented wholesale. The endgame state most players eventually reach, if they push a personal kingdom far enough, is founding their own breakaway realm and being crowned its ruler by vassals they’ve personally recruited — a state the base game’s political simulation supports but does not narratively resolve with any scripted ending, since the entire point of the sandbox is that there isn’t one waiting to be spoiled.

The long early-access period is also worth judging on its own terms rather than dismissing as simply overdue. TaleWorlds used the two and a half years between the 2020 early-access launch and the 2022 full release to rework core systems in view of an active player base — the clan and kingdom diplomacy layer in particular went through multiple substantial revisions based on early feedback that the initial political AI made peace and war decisions too erratically to support a believable campaign. That’s a genuinely different development approach from a studio building a sandbox strategy game behind closed doors and revealing it fully formed, and whatever friction the long public build cycle caused, the diplomatic AI at full release was measurably more coherent than the version early adopters first encountered in 2020, which suggests the extended early access period bought real design improvement rather than just extended marketing runway.

If the momentum-driven combat is the part that hooked you, there isn’t a cleaner ancestor than Sid Meier’s Pirates! for the open-career structure underneath it; and for a different strategy game that also trusts restraint and a bounded map over unchecked scale, Total War: Shogun 2 is worth the comparison for how differently two studios solved the same “make every battle matter” problem.