Business history can be highly misleading. Why? It’s written by the winners. Historians spend careers studying the making of the atomic bomb, but not the making of the atomic-powered plane (a spectacular Air Force failure). Research managers analyze DuPont’s development of nylon, not Coca-Cola’s creation of New Coke (a dud).

We need to pay more attention to the losers, because they can help us understand why innovators fail and when engineers should fold a project. I was reminded of this recently when Boeing, the world’s leading aircraft maker, announced the cancellation of a planned supersonic aircraft that promised to carry 300 passengers from Los Angeles to Tokyo in four hours and 20 minutes.

The experimental plane faced technical hurdles and required perhaps another 20 years of incubation. It wasn’t likely to meet future noise and emission standards either. But what really worried Boeing’s managers were surveys that found that neither passengers nor airlines would pay much of a premium for the plane. Boeing feared it would end up with a pathbreaking plane no one wanted.

Boeing folded at the right time. The company’s action revealed an appreciation for one of the great laws of innovation, the Law of the Rathole. It is a simple law: Know when to quit.

The rathole law isn’t popular.We live in a time when consultants routinely tell managers they aren’t “change-oriented” enough or that they’re afraid to make “quantum leaps” in processes and products.What the conventional wisdom misses is that many companies and lone inventors face the reverse problem: They are bold to the point of self-destruction.

Knowing when to walk away from a risky project, then, is as important as knowing when to stay the course. Here are five rules of thumb for deciding when to walk:

¦ Don’t Marry a Technology: Innovators fall in love with technologies at their peril. Often companies squander huge sums of money in order to satisfy the technical intuition of a senior manager. Sometimes older technology is better, if only because social, political and economic factors shape success and reinforce the position of existing techniques. That people the world over dislike the sonic boom effectively doomed the Concorde.

¦ Monitor Your Developers: At too many companies, research and development is kept in a black box. Managers know the box they need, but not the boxes inside the black box. To avoid a nasty surprise, managers must monitor a project from top to bottom. Bill Gates of Microsoft provides the model here. When he reviews the progress of a project, he meets with every layer of developer, which gives him a chance to sniff out differences of opinion and snags.

¦ Create Rival Ratholes: Two mistakes can be better than one. Consider the experience of Ashton-Tate, the leading maker of database software at the end of the 1980s. It needed a new version of its flagship program, but its development team scrapped one design after another. By the time the CEO assigned a new team to the task, it was too late. Ashton-Tate was dying, and the board fired the chief. Had he had two teams from the start, he might not have had a new program, but he would have realized the severity of his situation much sooner-and possibly saved the company by heroic measures.

¦ Spread the Risk: Some of the costliest projects, from new airplane engines to memory chips, are supported by alliances, not single companies. This cuts down the rewards of success-but also the size of the rathole.

¦ Manage Expectations: Big projects draw lots of attention from employees, investors and the tech-mad public. It is tempting to fuel the frenzy with grand claims. These make news. But leaders manage expectations. Often the hardest part of halting a muddled project is the loss of face that accompanies the decision. This loss is reduced sharply by having talked intelligently about the pursuit of innovation in the first place.

Follow these five rules, and the chances diminish that a pet project will vanish down the rathole. Look at Boeing. The airline maker certainly wasn’t married to supersonic technology. It spread its risks, essentially getting a silent partner, the U.S. government, to pick up a big piece of the tab. It understood customer needs and considered the social and political requirements needed to succeed with a supersonic plane. Finally, Boeing’s top managers had a realistic picture of the technical hurdles it faced.

Yet even Boeing fell short in some areas. It failed to manage expectations well, making its decision seem more like a black eye than it should. And it leaves the impression that the next generation of flight technology is a mountain that no one can climb, undercutting the company’s competitive image and feeding a potential hysteria over stagnation in commercial aviation. Finally, Boeing picked a time of financial stress-of layoffs and cutbacks in new jet orders-to drop the project. So even though the company obeyed the Law of the Rathole, it gives cynics the room to sneer that it merely responded to a short-term crisis by tossing its future overboard. And that’s the wrong message.