Ray Stefani even as a young man was calculating sports odds … even against himself.

In 1967, the Los Angeles resident was debating whether to move temporarily to San Antonio to train for the horse-riding portion of the modern pentathlon.

He ultimately decided against it after calculating that improving his horse-riding score would at best put him at sixth place, short of what was needed to go to the Olympics.

"Life is a matter of decisions, and you really can't second-guess yourself," says Stefani, now an electrical engineering professor at Cal State Long Beach.

But guessing on sports is now Stefani's passion. Using statistical models, he has tried to predict the winners of football, soccer and basketball contests … and to develop a winning betting system based on his results. He has presented his research at major international statistics conferences.

Stefani analyzed five factors before making his NCAA Final Four picks this year. They include:

The Associated Press poll.

The seeding of teams.

The Division I RPI score, which is based both on a team's and its opponents' winning percentages.

The odds given at a popular bookmaker's Web site in England.

And "score rating" … the team's average points scored versus allowed, taking into account the same statistic of its opponent. This number is the most important factor in prediction, Stefani says.

His result: Duke beats Kentucky to take the championship. St. Joseph's and Connecticut round out the Final Four.

Stefani loves sports numbers. He uses statistical techniques to predict the margin of victory and to estimate the probability of a team winning, losing or drawing.

Part of that analysis is accounting for home field advantage, which Stefani has carefully calculated for each major sport both in the regular season and the playoffs.

Home is lucky

Stefani's data shows that in domestic soccer play, the home team wins about 22 percent more often than it loses. The home team wins 20 percent more than it loses in pro basketball, 15 percent more in the NFL, 10 percent more in hockey and only 8 percent more in baseball.

Why? Home field advantage consists of avoiding travel fatigue and crowd intimidation, and gaining tactical advantage that comes with familiarity of home court, Stefani says.

Hockey players can more easily ignore the crowd beyond the glass, he says. And baseball involves rest between innings and less frequent travel, creating less fatigue, he adds.

Fatigue is a bigger factor in basketball and soccer, sports which are more fast-moving and offer fewer opportunities for rest.

The conventional wisdom is that home field advantage matters more in the playoffs. True, Stefani, says, except for baseball, where it actually drops, and hockey, where it stays the same.

Although Stefani himself hasn't engaged in serious betting since it's illegal in California, his gambling technique in hypothetical tests gets a nice return, thanks to the "Kelly method." That method calculates how much of your bankroll you should devote to a particular bet, based on the probability of winning and the financial return.

Let's say Stefani calculates that he has a 50 percent chance of winning a particular wager. If the bookie is offering a return of $3 for each dollar bet, or two-to-one odds, he should devote a quarter of his bankroll to the wager, according to the formula.

His technique, if used last year to bet on soccer at a popular bookmaker's Web site in England, would have produced an 11 percent profit per dollar bet, Stefani says.

Not bad.

"If you could make 11 percent on the stock market every week, you would be a very rich man over time," he says.

Serious scholar

Stefani has done serious academic research into sports statistics, including an analysis of ranking systems. In fact, his studies of the international FIFA soccer ranking system led him to make several suggested changes that were eventually implemented, he said.

For example, the ratings system once used the last six years in calculating ratings.

Stefani in 1996 wrote to the international soccer governing body saying that its method made little sense, because the World Cup is played every four years, he says. Shortly thereafter, the organization had lengthened the time span to eight years.

"FIFA was at least was open to some suggestions," Stefani says.

Numbers may be vital to sports predictions, but no system, however complex, is perfect.

"Before the first game, we are all experts," Stefani says. "Two rounds later, some of us feel kind of dumb."