Adversarial collaboration helps scientists understand controversial issues. Businesses should consider the technique for the great make-or-break decisions. The process can avoid the pitfalls that so often lead to strategic errors.

Decision-making within business is generally pretty good. But when it’s not pretty good, it can be disastrous. Everyday decisions are small decisions, so they don’t put the corporation at risk. Errors can be seen quickly. Managers learn from their mistakes, getting decisions right the next time. What economists call economic progress reflects many small decisions in which managers better served customers at lower cost. That means lots of small decisions were right.

Big decisions are not so successful. Major acquisitions turn out to be colossal wastes. Major software conversions fail badly. New products flop. Preparations for the unknown fail to protect the business when major surprises occur. And these failures occur even at successful companies with a track record of making good decisions about their day-to-day operations.

Two challenges prevent major decisions from going well. The paralysis of analysis leads some companies to never make the big decision one way or the other, but the fear of paralysis leads many companies to accept shoot-from-the-hip judgments.

Confirmation bias is the second challenge, leading people to find evidence to support their prior beliefs. Even worse, when staff members know what the CEO’s prior belief is, they find plenty of evidence to support that belief. Contradictory evidence is considered irrelevant or unreliable. Elaborate briefing binders end up buttressing a hunch rather than weighing conflicting evidence.

An alternative approach is called “red teaming,” from the military training exercises in which one team, the blue team, competes against another side, called the red team. If a corporation is considering a large acquisition, it might ask a few staff members to write up the best arguments against the acquisition to ensure that all the issues are on the table. My fellow Forbes contributor Bryce Hoffman has a good write-up of red teaming based, believe it or not, on a zombie movie. He also has a book on red teaming. A similar role was played at the Vatican by the Devil’s Advocate, an office created in 1587.

If we think of red teaming as an internal debate, though, we may well get caught up in all the usual flaws of debates: misuse of statistics, exaggerated arguments, as well as the occasional cheap shot.

The result should be information that the decision-maker can use without fear of being told what the staff thinks the boss wants to hear.

A process called adversarial collaboration, developed by Daniel Kahneman, avoids those drawbacks of internal debate. Scott Alexander has organized several of these on his blog. He writes, “an adversarial collaboration is where two people with opposite views on a controversial issue work together to present a unified summary of the evidence and its implications. …it’s a good way to make sure you hear the strongest arguments and counterarguments for both sides – like hearing a debate between experts, except all the debate and rhetoric and disagreement have already been done by the time you start reading, so you’re just left with the end result.”

Alexander’s latest adversarial collaborations can be reviewed for a better understanding of what the results look like. They range from whether calorie restriction slows aging to the potential for space colonization.

The two people (or two teams) present their best arguments to each other. Each person points out the other’s evidentiary flaws, such as cherry-picking data or over-reliance on limited information. They progress, hopefully, to laying out the facts they both agree on. The report will have sections about which both people agree and sections about their continued disagreement. The result should be information that the decision-maker can use without fear of being told what the staff thinks the boss wants to hear.

Adversarial collaboration can have defined deadlines, given by senior management, to avoid paralysis of analysis. Some teams have an arbitrator, a neutral party who settles internal disputes. With well-chosen participants, though, arbitrators have little to do. They may be useful in keeping the process on track.

Adversarial collaboration has been used in science a number of times, but there is little documentation of business applications. Yet big-bet decisions would benefit from the clear exposition of what is known and what is in dispute.

For corporate use, it would be imperative that the end product of adversarial collaboration be highly valued and that the participants’ service on the project be recognized regardless of the final decision on whether to ahead with the project under consideration or not.