We’ve seen in some of my other recent articles how companies like Sierra and Origin, taking stock of escalating complexity in gameplay and audiovisuals and their inevitable companion of escalating budgets, began to systematize the process of game development around this time. And we’ve at least glimpsed as well how such systematization could be a double-edged sword, leading to creatively unsatisfied team members and final products with something of a cookie-cutter feel.

MicroProse, suffice to say, didn’t go that route. Stealey took a hands-off approach to all projects apart from his beloved flight simulators, allowing his people to freelance their way through them. For all the drawbacks of rigid hierarchies and strict methodologies, the Darklands project could have used an injection of exactly those things. It was plagued by poor communication and outright confusion from beginning to end, as Arnold Hendrick and his colleagues improvised like mad in the process of making a game that was like nothing any of them had ever tried to make before.

Hendrick today forthrightly acknowledges that his own performance as project leader was “terrible.” Too often, the right hand didn’t know what the left was doing. An example cited by Hendrik involves Jim Synoski, the team’s first and most important programmer. For some months at the beginning of the project, he believed he was making essentially a real-time fighting game; while that was in fact some of what Darklands was about, it was far from the sum total of the experience. Once made aware at last that his combat code would need to interact with many other modules, he managed to hack the whole mess together, but it certainly wasn’t pretty. It seems there wasn’t so much as a design document for the team to work from — just a bunch of ideas in Hendrick’s head, imperfectly conveyed to everyone else.

It’s small wonder, then, that Darklands went so awesomely over time and over budget; the fact that MicroProse never cancelled it likely owes as much to the sunk-cost fallacy as anything else. Hendrick claims that the game cost as much as $3 million to make in the end — a flabbergasting number that, if correct, would easily give it the crown of most expensive computer game ever made at the time of its release. Indeed, even a $2 million price tag, the figure typically cited by Stealey, would also qualify it for that honor. (By way of perspective, consider that Origin Systems’s epic CRPG Ultima VII shipped the same year as Darklands with an estimated price tag of $1 million.)

[...] Combined with the only slightly less disastrous failure of the new point-and-click graphic-adventure line, Darklands was directly responsible for the end of MicroProse as an independent entity. In December of 1993, with the company’s stock now at well under half of its IPO price and the creditors clamoring, a venture-capital firm arranged a deal whereby MicroProse was acquired by Spectrum Holobyte, known virtually exclusively for a truly odd pairing of products: the home-computer version of the casual game Tetris and the ultra-hardcore flight simulator Falcon. The topsy-turvy world of corporate finance being what it was, this happened despite the fact that MicroProse’s total annual sales were still several times that of Spectrum Holobyte.

Stealey, finding life unpleasant in a merged company where he was no longer top dog, quit six months later. His evaluation of the reasons for MicroProse’s collapse was incisive enough in its fashion:

[...] But then, Darklands has been polarizing its players from the very beginning. Shortly after the game’s release, Scorpia, Computer Gaming World magazine’s famously opinionated adventure-game columnist, wrote a notably harsh review of it, concluding that it “might have been one of the great ones” but instead “turns out to be a game more to be avoided than anything else.” Johnny L. Wilson, the magazine’s editor-in-chief, was so bothered by her verdict that he took the unusual step of publishing a sidebar response of his own. It became something of a template for future Darklands apologies by acknowledging the game’s obvious flaws yet insisting that its sheer uniqueness nevertheless made it worthwhile. (“The game is as repetitive as Scorpia and some of the game’s online critics have noted. One comes across some of the same encounters over and over. Yet only occasionally did I find this disconcerting.”) He noted as well that he personally hadn’t seen many of the bugs and random crashes which Scorpia had described in her review. Perhaps, he mused, his computer was just an “immaculate contraption” — or perhaps Scorpia’s was the opposite. In response to the sidebar, Wilson was castigated by his magazine’s readership, who apparently agreed with Scorpia much more than with him and considered him to have undermined his own acknowledged reviewer.

The reader response wasn’t the only interesting postscript to this episode. Wilson:

This anecdote reveals what may be a deciding factor — in addition to a tolerance for complexity for its own sake — as to whether one can enjoy Darklands or not. Wilson had been willing to simply inhabit its world, while the more goal-oriented Scorpia approached it as she would any other CRPG — i.e., as a game that she wanted to win. As a rather plot-focused, goal-oriented player myself, I naturally sympathize more with her point of view.
​