Siege

Ali Atabek founded Mindcraft Software in 1989, with his wife Ugur Atabek and business partner Jim Thomas, to publish The Magic Candle, a role-playing game he had designed after being inspired by Ultima II. That game won Computer Gaming World’s RPG of the Year award for 1989 and established Mindcraft as a serious outfit in a genre dominated by larger studios. By 1992 the company was publishing six titles a year from its Torrance, California base and had grown to around thirty employees. Siege was the first of several attempts to extend the Magic Candle universe beyond role-playing, and the most successful of them.

The game was set in Western Gurtex, the same world that had served as the backdrop for the Magic Candle trilogy. The conflict was the familiar fantasy binary: the Forces of Light, comprising humans, dwarves, and elves, against the Forces of Darkness, fielding orcs, trolls, domugs, and tekhirs. Four fortresses were at stake: Highrock Fortress, Fort Neir, Castle Elissa, and Usk’hem Gart, each with six dedicated scenarios for a total of twenty-four. Players chose which side to command before selecting a scenario, and the game’s core proposition was that both perspectives were fully playable and mechanically distinct.

The attacker’s toolkit ran to battering rams, assault ladders, siege towers, mobile bridges, and catapults, supported by up to seventeen unit types ranging from berserkers and sergeants to engineers and magic-users. The defender had ballistae, burning oil to pour on troops attempting to scale the walls, and the same roster of magical support. Terrain affected which units could move where and how effectively, and spell-casting units on both sides could alter the course of an engagement through fireball attacks or healing. The game played out in real time, pausable at any moment to issue orders, with a difficulty range from one to eight that the documentation called “wimpy” through “hopeless.” A scenario editor allowed players to create new engagements or modify the included ones, which reviewers recognised as a significant addition to the package’s longevity.

Computer Gaming World reviewed Siege in 1992, praising the VGA graphics as “beautifully rendered” and approving of it as the first PC game to focus specifically on castle warfare. The criticism was directed at the AI, which the magazine found too predictable, and at the win conditions, which lacked detail. The conclusion was that the game was “an all-around unique and engaging simulation” that would have been outstanding with a more challenging opponent. Dragon magazine awarded it four stars out of five in 1993. The consistent point of agreement across reviews was that the concept was well-executed and the subject matter underserved; the consistent complaint was that the computer opponent could be outmanoeuvred too easily, and the only real difficulty came from numerical disadvantage rather than tactical ingenuity.

An expansion pack, Dogs of War, added multiplayer support, six new castles, and sixteen new unit types. Multiplayer addressed the central weakness of the base game directly: a human opponent could not be confused by feints in the way the AI could, and the reviewer for Pelit described it as the best thing about Dogs of War for exactly that reason. Mindcraft followed with Ambush at Sorinor in 1993, using an improved version of Siege’s engine for a different scenario type, and then Walls of Rome the same year, which applied the same engine to Roman siege warfare and was generally considered a stronger execution of the underlying system. The studio continued developing across RPG and strategy genres until 1997, producing fifteen games in total before winding down operations.

Siege occupies a specific position in the DOS strategy catalogue: a game that identified a niche nobody else was working in, built a technically capable system for it, and was let down by a single major weakness it never fully resolved. That the expansion and the sequel both sold on the strength of what the engine could do is evidence enough that the original concept was worth pursuing. The castle warfare subgenre it pointed toward would not find a definitive expression until much later, and several players who encountered Siege in 1992 have noted since that nothing produced in the following decade quite replaced what they were looking for there.