TTLG Forums / / / / /

Reasons for the Fall: A Post-Mortem On Looking Glass Studios ~ James Sterrett

We’ve all been shocked and saddened by the sudden demise of Looking Glass. It seemed that Looking Glass was riding high, with System Shock 2 and the Thief series finally getting it the public acclaim and sales Looking Glass has always deserved. Suddenly, Looking Glass turned out to be dead. Why did this occur? The more I have learnt about this, the more complex the answer has become.

However, the quick answer is this:

Looking Glass died because a series of problems compounded each other into a financially lethal situation. No single factor is to blame. No single person or entity killed Looking Glass. No one problem was enough, on its own, to kill the company. Nonetheless, the problems were deadly when combined.
A hard fact of the world is that software development companies are rarely in a very good financial situation. The company enters a contract with a publisher to produce a specific game with a given set of features. The contract usually specifies that the publisher will provide the producer with money when the producer achieves agreed-upon milestones in the construction of the game. If the producer fails to achieve a milestone before the previous dollop of cash runs out, the producer has to find more money, or go out of business. Getting more money now almost inevitably means borrowing it in some manner, which leaves the producer with a debt to pay off in the future, reducing the net profit from royalties - payments the producer gets as a percentage of the publisher’s profit from sales of the game - the producer gets once the game is finished. Those royalties are only a small fraction of the actual revenue gained by sales of the game, and, furthermore, the payments the producer got during development are considered as advances against the royalties.

As a result, the producers are in a bind: they have to make a very complex product, on a tight schedule. If the schedule slips - and, because of the complexity of the product, the schedules nearly always do – the producer begins to get into trouble.

One of the ways that a producer can reduce their exposure to this risk is to have several projects running at once on an offset release schedule. If all goes very well, each product’s royalties then provide the operating expenses to get later products through their troubles. If one product fails to make money, the revenue stream from the others cushions the company from the blow. A producer needs to have a minimum of three or four active projects for this to work.

Inevitably, there are some catches in this plan. The biggest catch is that nothing guarantees the income cushion: no game is guaranteed to do well. If several games fail to do well at once, the scale of trouble is proportionately worse. It takes a large company to conduct multiple projects at once. Thus, when the company does have to search for extra funding, it needs a much larger infusion of cash than a smaller, single-project company. The risk involved is reduced, but the consequences are worse when things turn sour.

All in all, producing games is not for the risk-averse. Lots of things can go wrong. Many game projects never get to the point of being sold. Many of those that are sold do not make enough money to repay the cost of their production. Since the games cost several millions of dollars to produce, and more millions to publish, a lot of copies have to be sold just to break even. Consider that Thief II cost $2.5 million to develop, and probably, at the very least, another million to publish (and probably more). At a price of $40 per box, it needs to sell a minimum of 87,500 copies just to break even. And at 100,000 copies sold the profit margin is still only half a million dollars, which isn’t enough to fund the next game.

Keep in mind that this doesn’t all mean that the publishers are the evil villains of the game industry for failing to support obviously good companies and ditching the bad ones, and for failing to split the profits equitably with the producers. Publishers are, in fact, in the same binds as the producers, except on a much larger monetary scale. Bigger publishers are better able to offset losses with a few hits, which is why there are so few publishers around: most of them have died or been bought up.

The third link in the chain of getting a game into your hands, the game retailers, is also under immense financial pressures. Yes, retailers do some pretty amazing contortions that introduce big problems for the publishers. But when was the last time you saw an independent game retailer? All of their strange tricks (see The Retail Game) are designed to try to insure that the store won’t fold when it bets wrong on which products to stock. Game retailers have wound up in large chains because the greater mass of money allows a single store to survive mistakes better; the parent corporation may bail it out if it founders, while an independent retailer would simply fold.

The entire computer games industry is high-risk. Vast amounts of money are spent in the hopes that a few games will sell well enough to make up for the losses incurred on all the others. (If you want to know more about the perils of computer game development and publishing, there are several good articles about this on the web. I suggest beginning with these three: Staying in the Black; The Retail Game; and Why Being a Computer Game Developer Sucks .)

When we look at things in this light, Looking Glass was actually a remarkably successful and long-lived independent developer. It survived all of these pressures for eleven years and produced at least 12 titles: Ultima Underworld I & II; System Shock I & II; Terra Nova; British Open Championship Golf; Flight Unlimited I, II, & III plus a Windows ’95 version of Flight Unlimited I; Thief I & II, Thief Gold, and, if the bankers permit it, Jane’s Attack Squadron. In addition, Looking Glass did conversions of various PC games to console systems, and at least one original console game: John Madden ’93 for the Sega Genesis, which sold over one million copies and was their greatest single hit.

Of all these, most were successes. Terra Nova was a disaster, not least because Looking Glass decided to publish on its own. Exactly why Terra Nova failed on the market is not clear. Possibly the gaming public saw it as a Mechwarrior clone and thus ignored it. If Terra Nova had been a success, this would have allowed Looking Glass to reap far greater profits, as they had done with the self-published Flight Unlimited 1. When Terra Nova was a commercial disaster (despite selling over 100,000 copies), Looking Glass was left holding all the publication debt. Looking Glass might have folded at this juncture, but was bailed out by Averstar, Inc.

Terra Nova would have killed most companies. Looking Glass, probably because of its previous string of successful titles, was able to find creditors and stayed afloat, but in debt. Over the next several years, the company appears to have led a troubled existence. Nonetheless, Looking Glass nearly died again in 1997. British Open Championship Golf experienced a difficult birth, and then bombed when released, as a third self-published game, in the spring of 1997. The company saved itself, in part by closing a studio in Austin set up by Warren Spector. While it reduced overhead, this closure was a serious loss. Not only did it cost Looking Glass staff who were important to ongoing projects, such as Thief, but the team involved comprised no small net talent: many of that team formed the group now completing Deus Ex. Moreover, the stress of not knowing if the office would be open from day to day caused many of the Looking Glass staff to leave, slowing projects while replacements were found and trained. In and around all these troubles, a project for a Star Trek: Voyager game foundered and eventually died. All told, Looking Glass incurred a significant quantity of long-term debt in this period, though it improved its financial situation at the end of the summer through a merger with Infometrics, a subsidiary of Averstar.

Yet Looking Glass survived. Flight Unlimited II appeared late in 1997 and eventually broke even. Nonetheless, the debt incurred in this period - from Terra Nova through Flight Unlimited II – comprises the precondition for Looking Glass’ eventual demise. If all went well, Looking Glass could control the debt, pay it off, and survive. If trouble occurred, it was not as well-placed financially to deal with it.

Thief appeared in late 1998 and was a major hit, eventually making millions of dollars for Looking Glass. Had Thief not been a big hit, Looking Glass would almost certainly have folded. Instead, Thief appeared to revitalize the company. Thief Gold sold well, and System Shock II sales were passable. Thief II is currently on the way to being an even bigger hit than Thief I. However, Flight Unlimited III (released in 1999) sold quite poorly and ate the gains from System Shock II due to an agreement with Electronic Arts to pay off advances for both projects with the royalties from both projects.

Furthermore, there is extensive circumstantial evidence that Thief’s success, great though it was, may have done little more than plug the holes that debt had left in Looking Glass’ budget. Thief ensured that, if all went well, Looking Glass could survive 1999 and 2000.

Unfortunately, a series of problems came together, and ensured that a lot of things went very badly wrong.

The fourth game in the Flight Unlimited series, Flight Combat, ran well over schedule. Worse, Flight Combat (Janes’ Attack Squadron) was knowingly undersold to Electronic Arts: Looking Glass apparently gambled that it could get the game out the door faster than the budget actually called for. The gamble failed: Flight Combat was instead slower than the planned budget, in part because of a major redesign insisted on by Electronic Arts. This redesign appears not to have been accompanied by a sufficient change in the budget and schedule for Flight Combat.

The multiple project system depends on each product coming out essentially on schedule. Since Flight Combat went over time, Looking Glass was not only losing money on the project, but also failing to get the income from it that was supposed to support its other projects.

In theory, multiple offset projects allows the developer to survive a failure. Thief was profit that plugged holes in short-term debt. System Shock II more or less offset the failure of Flight Unlimited III. Flight Combat was thus a financial problem that strained the multiple project system and put Looking Glass in danger again, and cash in hand began to run critically low in early autumn of 1999, despite the royalties coming in from Thief (in addition to, presumably, milestone payments for Thief: Gold.)

Work began in early 1999 on a modern-day stealth-based game called "Deep Cover". This was a joint project with Irrational, similar to the System Shock 2 project. Looking Glass was to provide a version of the Dark Engine while Irrational would provide the bulk of the game design. Unfortunately, finding a publisher for this project took 9 months.

As a result, Looking Glass’ cash shortage began to catch up with it. Payments to Irrational were affected, placing Irrational in danger of extinction. As a result, Irrational informed Looking Glass that it was looking for other work, and eventually found it in November 1999 in the form of a contract for a Playstation 2 game. Shortly after Irrational signed onto its new project, Looking Glass came to an agreement with Microsoft to publish Deep Cover in conjunction with Irrational - but Irrational felt it had to pull out of the Deep Cover project to honor its own recently-signed contract elsewhere.

The rights and materials to Deep Cover reverted to Looking Glass, and Looking Glass was allowed to continue to hire the services of the Deep Cover design team from Irrational. Microsoft, however, was unhappy that Irrational would not be involved in the manner originally intended. Despite negotiations, Microsoft pulled out of the deal in February 2000.

This apparently subtracted at least one million dollars from the cash that Looking Glass had expected to use to cover its operating expenses. That’s a lot of money, but keep in mind that one million dollars is almost certainly less than Looking Glass’ payroll expenses for one year: Looking Glass employed some 60-80 people, and even if all of them were paid a spartan salary of around twenty thousand per year, its salary costs per year would be around one and a half million dollars. Even that figure ignores all its other operating expenses of, probably, at least another million dollars a year. In other words, the failure of this contract could only be a mortal threat to Looking Glass if its finances were already in bad shape. Furthermore, had Irrational remained in the contract, a substantial fraction of the money from Microsoft would have gone to keeping Irrational afloat, instead of to Looking Glass.

Thus, Flight Combat’s cost overruns, earlier debts, the failure of Flight Unlimited III, and the failure of the Deep Cover project with Irrational and Microsoft, all combined to leave Looking Glass in a perilous condition.

Looking Glass sought new sources of funding. In particular, it tried to get itself bought by a company with enough resources to keep Looking Glass solvent.

A deal was in progress with Sony when Sony hit economic troubles. Sony reorganized, and the executive who had been in charge of the deal with Looking Glass was replaced. The replacement canned the deal, terminating this possible rescue.

Eidos then stepped into the breach and firmly intended to buy Looking Glass, presumably intending to protect their projected income stream from Thief II Gold, Thief III and other potential future projects such as Deep Cover. The deal was agreed upon, and, while the paperwork was processed, Eidos began to keep Looking Glass afloat. It seemed that Looking Glass had been saved.

However, the economic instability sweeping the entertainment industry slammed into Eidos, which found itself with insufficient funds to complete the purchase. Eidos asked its bankers for more money. The bank was itself in a period of difficulty, and thus the bankers said no.

Left with only enough cash in hand to pay off its employees, Looking Glass was forced to shut down.

Looking Glass had a long run; it survived eleven years in a business where all too many companies have a lifespan a mayfly would not envy. It consistently produced superb games, most of which were commercial successes, in an industry where many games are second rate and many fail to make a profit. It repeatedly survived disasters that would have killed most companies. All in all, Looking Glass was a spectacular success as a games producer.

But the games industry is a high-risk environment of speculative capital. Looking Glass beat the odds for eleven years. Eventually, the odds caught up with it.

Several conclusions can be made from this:

First, as mentioned before: No one factor killed Looking Glass. A number of attempts have been made to blame various single parties. While all of these accusations have a small grain of truth, all of them fail to appreciate the wider picture. To debunk some of the more prevalent myths:

Eidos did not kill Looking Glass. True, the final word came from Eidos. However, Eidos was deeply involved in trying to keep Looking Glass afloat. Blaming Eidos is like claiming that a drowning victim was killed by the person who attempted the resuscitation.

Eidos’ bankers did not kill Looking Glass. True, the final blow was dealt by Eidos’ bankers. But without an enormous chain of events beforehand at Looking Glass, Eidos’ bankers would never have been involved. If Eidos itself had not hit financial trouble, the bankers would not have been involved. Had the bank been in good shape, there is a good chance it would have approved the rescue.

John Romero did not kill Looking Glass. True enough, Eidos’ finances might have been much healthier if Daikatana had been on schedule and on budget. But Romero’s flaws would have had no bearing on Looking Glass’ survival, if Looking Glass had not already been in deep trouble for its own reasons - including games that went over time and over budget.

Irrational did not kill Looking Glass. True, pulling out of the Deep Cover contract hurt Looking Glass. But this single failure could not have been mortal without the severe problems that already existed in Looking Glass’ budget.

Flight Combat did not kill Looking Glass. True, the cost and schedule overruns were very harmful, and the cost gamble turned sour. Yet without the other problems that came before and during this project, it would not have been lethal.

Poor sales of Thief did not kill Looking Glass. This hasn’t even a grain of truth. Thief sold well and, according to Tim Stellmach (Gamespy interview), Looking Glass made millions of dollars from it. If Thief had failed to be a hit, Looking Glass would have died. Instead, Thief kept the company going.

Poor sales of System Shock II and Flight Unlimited III did not kill Looking Glass. Obviously, better sales would have helped. Flight Unlimited III was not a success. However, System Shock II did not sell poorly and, according to Tim Stellmach (Gamespy interview), should break even or better in the long term. In any event, had other things gone right – Flight Combat, Deep Cover, the Sony deal, or the Eidos deal – then the multiple project system would have covered for the failures.

Neither Terra Nova, nor the Austin studio, nor Star Trek: Voyager, nor British Open Championship Golf, nor even Elvis killed Looking Glass. Looking Glass survived the worst moments of all of these. True, the damage these caused helped to drag it down in the end. But the damage itself was insufficient to actually cause Looking Glass to keel over.

So what can we blame? If you must blame something, blame the system as a whole. The games industry is extremely volatile. Millions of dollars must be spent to design, test, and manufacture a product, and as often as not the products fail to repay that investment. The whims of the buying public are quite difficult to predict. Companies whose products fail to pay off rarely get a second chance, because nobody is in the business to lose money.

Looking Glass, in retrospect, did a stunningly good job in this harsh environment. If we discount the Windows ’95 version of Flight Unlimited 1, Thief: Gold, and its console conversion, it produced 11 titles in about 11 years (if we start counting from the late 1989 inception of Underworld 1). Of these, three were commercial failures (Terra Nova, British Open Championship Golf, and Flight Unlimited III), and another one might fail to break even (System Shock II has not yet broken even, but is expected to do so). Two thirds of its titles were commercial successes. Some of its titles are legendary as creative achievements; all have received critical acclaim.

That’s no mean feat. Surviving for a decade, through shocks that would have killed lesser companies, without compromising its design principles, and producing many of the greatest games ever to grace a computer, is a legacy to be proud of. While we may be saddened by its death, we should also be grateful that Looking Glass had the tenacity and agility to survive so long.

Finally, the Looking Glass legacy does not prove that deep, complex gameplay is a recipe for market failure. If that were the case, Looking Glass would have died long ago. Ultima Underworld would be cult hit known to few, instead of consistently being praised as a seminal computer game. Thief and Thief 2 would not have been hits. If anything, Looking Glass’ long survival proves that deep, complex, immersive gameplay tends to be commercially successful. Only one of Looking Glass’ clear-cut money losers, Terra Nova, resides in its "core competence" of first-person action-adventure titles. The other two major commercial failures – Flight Unlimited III and British Open Championship Golf – went head to head with established powerhouse titles (Microsoft Flight Simulator and Links, respectively) on those title’s home turf. That’s a daunting and risky course of action, and in both cases it failed to pan out. Arguably, Terra Nova suffered from the same fate, being seen as a "big robot game" and thus a competitor to Mechwarrior. By contrast, Ultima Underworld, System Shock, Flight Unlimited 1, and Thief all broke new ground and were rewarded for it.

In other words, where Looking Glass innovated most, and provided a game experience that was most different, it did best commercially. This, and Looking Glass’ longevity, strongly suggests that in the long run, good innovative games are good profitable business.

James Sterrett ~ 31/5/2000

Go Back

TTLG Forums / / / / /