The Miami Heat and San Antonio Spurs have reached the NBA Finals, setting up a rematch of last year's classic seven-game battle. With four days between the conclusion of the conference finals and Game 1, now comes the real fun: predicting the outcome.
Predicting the NBA Finals is somewhat different from predicting other playoff series. First, there's more playoff data available -- three full rounds against a variety of opponents, which gives a better idea of who's playing better right now. Second, because the teams have played each other just twice during the regular season, there's less head-to-head data to determine whether the matchup is good or bad.
So what has really mattered when picking the winner of past Finals? And what does that say about this year's series? Let's take a look.
Matters: Regular-season records
Ordinarily, statistical projections will start with point differential. But when it comes to the NBA Finals, during the past three decades, records have matched up better with results than differentials.
When the two Finals teams were separated by three games or fewer during the regular season, teams with home-court advantage have actually been more likely to get upset than win the series, going 4-5. But when they won at least six more games during the regular season than their opponent, they've gone a dominant 16-2, with 1995 (Houston over Orlando) and 2006 (Miami over Dallas) the lone exceptions. That's bad news for the Heat, who finished eight games behind the Spurs in the standings.