N0 LyF3 ScRuB
05-05-2014, 02:08 PM
With the first round of the NBA playoffs in the books, the championship picture is hardly any clearer than it was two weeks ago. Favorites in both conferences survived scares that took them the distance, while the Miami Heat (http://espn.go.com/nba/team/_/name/mia/miami-heat) got needed rest after becoming the only team to sweep its opening matchup.At the conclusion of the conference semifinals, we'll have a much better idea of who's headed to the NBA Finals. For now, here's a look at what the numbers say about which teams are most likely to hoist the Larry O'Brien trophy.
Methodology
My playoff projections are based on a regression that maps point differential and head-to-head results to the outcome of series from 2003-13. As I wrote about Giant Killers (http://insider.espn.go.com/nba/story/_/id/10802035/identifying-potential-davids-goliaths-playoffs-nba), regular-season matchups are a good predictor of playoff series, and indeed the only team without home-court advantage in the first round that won the season series (the Washington Wizards (http://espn.go.com/nba/team/_/name/wsh/washington-wizards)) pulled off a first-round upset. The Portland Trail Blazers (http://espn.go.com/nba/team/_/name/por/portland-trail-blazers) were the only team to advance after losing the season series. (Three other winners split head-to-head with their opponents.)
For point differential, I used a blend of 80 percent season-long differential adjusted for opposition, plus 20 percent the same measure from the first round of the playoffs. That and head-to-head results projected the likelihood of a series win in the next two rounds, while differential alone was used for the NBA Finals because two-game interconference series don't have proven predictive value. With estimates of each team's likelihood of making it to a given round, I projected out all possible matchups to determine the likelihood of winning the title.
Rankingshttp://a.espncdn.com/combiner/i?img=/i/teamlogos/nba/500/sa.png?w=110&h=110&transparent=true
1. San Antonio Spurs (http://espn.go.com/nba/team/_/name/sa/san-antonio-spurs) (26.4 percent)
ESPN insider
Methodology
My playoff projections are based on a regression that maps point differential and head-to-head results to the outcome of series from 2003-13. As I wrote about Giant Killers (http://insider.espn.go.com/nba/story/_/id/10802035/identifying-potential-davids-goliaths-playoffs-nba), regular-season matchups are a good predictor of playoff series, and indeed the only team without home-court advantage in the first round that won the season series (the Washington Wizards (http://espn.go.com/nba/team/_/name/wsh/washington-wizards)) pulled off a first-round upset. The Portland Trail Blazers (http://espn.go.com/nba/team/_/name/por/portland-trail-blazers) were the only team to advance after losing the season series. (Three other winners split head-to-head with their opponents.)
For point differential, I used a blend of 80 percent season-long differential adjusted for opposition, plus 20 percent the same measure from the first round of the playoffs. That and head-to-head results projected the likelihood of a series win in the next two rounds, while differential alone was used for the NBA Finals because two-game interconference series don't have proven predictive value. With estimates of each team's likelihood of making it to a given round, I projected out all possible matchups to determine the likelihood of winning the title.
Rankingshttp://a.espncdn.com/combiner/i?img=/i/teamlogos/nba/500/sa.png?w=110&h=110&transparent=true
1. San Antonio Spurs (http://espn.go.com/nba/team/_/name/sa/san-antonio-spurs) (26.4 percent)
ESPN insider