Game Over

George Harris says the exercise was serious fun. "People got into it," he reports. "One of our competitors is from Turkey, so the guy who was running that company in the simulation -- one of our best guys -- became 'Ali-Ben' Donahue. Unfortunately, he made some bad decisions and crashed the company. The next day we put out a press release announcing that Ali-Ben had been shot for crimes against the state. He was running a nationalized company."

As the simulation unfolded, an interesting pattern emerged. No matter what the Harris team did with the Argentine mine, it couldn't gain ground on the competition. Meanwhile, as the other teams made counter moves, they recognized that Harris's strategic decisions didn't much affect their performance. "The guys running the other two companies said, 'We just don't worry about what you do except when it screws up the market,'" reports George Harris. "I don't know if that was a surprise to them, but it was a surprise to me."

The game changed how all 30 executives looked at the company's position. Harris didn't need access to more boron; the real problem was global overcapacity. Efforts to grab market share would weaken prices for everyone. The company reckoned with the results for a few weeks and took action. It passed on the Chile deal.

The Harris war game was valuable, McClellan says, because it recognized the basic rules of successful simulation. First, the exercise felt real. The computer model was a convincing representation of the market in which Harris operated, and each team's moves and counter moves represented its best efforts to approximate real life. "A good simulation generates counter intuitive but believable results," says McClellan. "People should see things and say -- "I've never seen that before. But now that I've seen it, I believe it."

Second, the exercise focused on a well-defined choice. The executives weren't trying to "get better at strategy" or "learn to work together." Those long-term benefits were outcomes of their concentrated efforts to wrestle with a narrowly defined question: Should we buy the mine and at what price? A clear focus makes the simulation manageable and its lessons -- no matter how surprising -- more concrete.

Third, the players saw a direct connection between their decisions and the company's performance in the market. One of the most powerful virtues of simulation is that it compresses the time between action and outcomes. In reality, it would have been years before Harris knew whether a decision to buy the mine was sound. In the simulation, the team initiated a range of strategic moves and saw the consequences in minutes. "Simulation kills bad ideas fast," says Mercer's Fay, "and it breathes life into good ideas that might be politically unpalatable."

That's why the best simulations often produce surprising insights -- even when "surprise" means verifying the status quo. ACS's Mark Chussil recently ran a war game not unlike the Harris exercise; a company was practicing in order to choose between several strategic options. As part of the game, the computer generated quarterly performance data that captured the impact of various strategies.

"One of the teams adopted the company's existing strategy and saw its profits decline every quarter for the first year," Chussil recalls. "So we asked them how they felt about that. Did they want to change their strategy? They said, 'No, we're confident our strategy will turn around.' Next quarter, profits were down again -- and then down again. Now there were five seconds between quarters. 'Dammit,' they said, 'we have to change our strategy. And their profits immediately started going up. You could hear the sighs of relief, and by the end of the five years they'd made lots of money."

"Then we said, 'Let's go back to when you changed strategy and run the simulation again. But let's assume you stuck to the original strategy.' It turns out they would have made twice as much money. It was an incredible learning experience -- for me, as well as for the people in the room."

How to Practice

In reality, of course, most business people won't use agent-based software to hone their project-management skills or spend a week at a war game. Still, these tools offer important lessons for anyone who wants to get better at business -- best practices on how to practice. Simulation experts point to three major principles.

First, never forget that the point of practicing is to learn how to get better. That means getting comfortable with failure. According to John McClellan, practice sessions work best when they encourage people to take chances in a blame-free environment. "If people experiment and fail," he says, "they get frustrated." Which is precisely the point. The best learning, he says, comes from "creative frustration in a safe environment."