(MONEY Magazine) – Bill Miller looks and sounds more like a college professor than like the country's greatest mutual fund manager. When he was speaking before hundreds of investors at a recent conference in Las Vegas, his clothes were rumpled and his hair was sticking up. He sprinkled his talks with references to philosopher Ludwig Wittgenstein and poet T. S. Eliot. Nevertheless, this is indeed the man who just beat Standard & Poor's 500-stock index for the 13th year in a row.

Thirteen years--that's three Presidents, two recessions and a bout of Internet mania followed by a bitter bear market. That's even after subtracting hefty 1.72% management fees--and after his Legg Mason Value Trust has swelled to more than $13 billion in assets. It's a winning streak no other fund manager in history can match.

As Miller himself would say, however, the streak is one of those statistics that sounds incredibly impressive but in the end doesn't matter as much as the answer to his shareholders' basic question: How much have you, Mr. Miller, made for us in those 13 years?

That number is pretty remarkable too: a cumulative 680% through Dec. 15. That's more than double the return of the S&P 500 over the same period. Only three funds that invest in similar asset classes have performed better, and they're all run by teams of managers. Value Trust is all Miller, all the time.

So how does this guy do it? Is he just lucky? Michael Mauboussin, chief U.S. investment strategist at Credit Suisse First Boston and a Miller pal, calculated the odds that a fund manager would beat the market every year in any given 13-year stretch at one in 149,012. Sure, luck helps, but long streaks tend to happen to the most skillful, Mauboussin says.

Skillful, Miller is. He prizes free cash flow. He buys stocks no one else wants and then holds them for years. He decries stock options and pesters holdings about corporate governance. This is standard value-investing stuff. But Miller executes it with an unorthodox style befitting the former philosophy student that he is. While other value managers shunned tech stocks in the late 1990s, Miller loaded up. Value Trust's top two holdings remain Amazon.com and Nextel Communications. He makes huge bets--10 stocks represent nearly 50% of his portfolio--and has an iron stomach that allowed him to keep buying Amazon as it fell from $80 to $5.50 in 2001. His peers thought he was nuts, but this year, Amazon's 155% run is a big reason for Value Trust's market-beating performance.

Typical Miller. He absolutely delights in going against the grain. After MONEY called Value Trust "The Scariest Portfolio Ever!" in June 2002, Miller put the magazine's cartoon illustration in his 2003 annual report. At the Las Vegas conference, he gleefully announced to his shareholders that a high-powered venture capitalist had called most of Value Trust's holdings terrific candidates for short-selling. The more everyone's telling him he's wrong, he figures, the more the market is mispricing the stock--and the greater the opportunity to make money. And the more people are talking up a stock, the less it appeals to him. "If it's in the papers," goes one of his favorite sayings, "it's in the price."

His contrariness surfaces again in his observations about the stock market's direction in 2004. While most investors are skeptical about the U.S. economy's gains, Miller thinks rising worker productivity and quickening capital spending, among other factors, portend a banner year. "I'd be more surprised if the equity market was up 6%, 7%, 8% than if it were up 30%," he says.

Miller's results have been so good for so long that it's natural to think his market-beating days must be over. What are the chances he'll beat the S&P a 14th year? (One in 372,529, according to Mauboussin.) Something's got to go wrong. He'll get arrogant, overconfident. He'll buy one too many scary stocks.

Miller says he constantly guards against those mistakes, but he also knows that the streak does have to end eventually. "Just not this year," he says.