How do planners decide whether a family will have enough to live on at retirement?

In the old days, planners often just assumed an average annual return of 8 percent on a balanced investment portfolio.

These days many use a "Monte Carlo simulation" to examine hundreds of possible future financial scenarios.

Monte Carlo assumes the stock market and other investments will have fluctuating returns, as in real life. Like a rotating roulette wheel where the ball falls on a different number each time, Monte Carlo spins scenario after scenario, figuring the odds of unforeseen problems, such as another sudden stock-market downturn as seen after Sept. 11, 2001.

"When you do this a large number of times, a pattern starts to emerge," says Kamaryn Tanner, senior vice president of product development for financial-software giant SunGard.

World War II-era U.S.-based atomic scientists used Monte Carlo to simulate nuclear chain reactions as they strove to build the first atomic bombs.

Do it yourself: Individuals can purchase software to do Monte Carlo simulations ($495 for Vanguard Software Corp.'s DecisionPro program, for example), though planners say most are better off working with an expert.

Monte Carlo applied to the Volds' situation: Todd Vold hoped to retire at 55 but the simulation showed that even if he waits until 60, the family would have only a 76 percent chance of a secure retirement.

If he works through age 62, odds jump to 97 percent.