At least one conference looks like a dartboard this spring, so I ran log5 calculations to help sort things out. This method finds the probability of a win in each game, given team quality and home court advantage. (I used expected or Pythagorean win percentage for team quality, and the traditional 0.6 HCA adjustment.) The added difficulty here (over baseball) is the length of series and the increased number of teams. (As such, it's a bit too much to find full postseason probabilities at this point ... though if someone handles it, let me know.)

Below, we'll run through the first round matchups and give the log5-based probabilities on which team will win and how many games the series will last.

EASTERN CONFERENCE

Boston v Atlanta

As you'd expect, Boston looks to cakewalk this one, with log5 offering a 98.4% probability of a Celtics win. log5 gives Atlanta a 46.5% of so much as winning a single game. log5 gives Boston a 90.4% probability of winning each home game; Atlanta has a 19.2% probability for a win in each of its home games. Boston's sweep probability is 53.5%. Atlanta's probability of sweeping Boston is ... 0.03%. Hmm.

log5-based prediction: Boston in 4.

Detroit v Philadelphia 

Philadelphia, unfortunately, doesn't fare much better in log5's (hungry) eyes. The Sixers do have a 69.2% probability of avoiding a sweep, though. Detroit's probability of a series win is 91.7%. Detroit in 5 (37.7%) nudges Detroit in 4 (30.8%).

log5-based prediction: Detroit in 5.

Orlando v Toronto 

Pythagoreas thinks much more highly of Toronto than does anyone south of the border. As such, log5 gives the Raptors a 25.4% probability of winning the series, and only gives Orlando an 11.5% probability of getting the sweep. In fact, log5 says Orlando in 7 is more likely than Orlando in 4. Those frisky Dinos!

log5-based prediction: Orlando in 5.

Cleveland v Washington 

HCA matters a lot in this one, as the teams have played at roughly the same level this season. Cleveland manages to hold a 61.9% probability of winning the series, but log5 thinks it will take seven games to get there. The system gives the 'Zards only a 14.7% chance of winning in four or five -- it seems D.C.'s best shot comes with a longer series.

log5-based prediction: Cleveland in 7.

WESTERN CONFERENCE

Los Angeles v Denver 

For the parity in the West, the Lakers are still much better (statistically) than Denver. log5 sets the Lakers series win probability at 74.3%. There's a 24.5% probability this one goes to seven ... but log5 thinks the Lakers will instead win in five (26.9%) or six (20.1%).

log5-based prediction: Los Angeles in 5.

New Orleans v Dallas

An early contender for 'series of the spring.' The most likely outcome is seven game series -- log5 gives little hope (11%) to a sweep either way. N.O.'s home court advantage comes in handy; Dallas's best shot to win would come in six, but even that falls behind the probabilities of the Hornets taking it in five, six, or seven. log5 calls the series for N.O. with a .647 probability of a win.

log5-based prediction: New Orleans in 7.

San Antonio v Phoenix

Pythagoreas actually thinks Phoenix is the superior team here, despite home court (and the superior record) falling in the Spurs' hands. But HCA makes up the difference, and log5 endorses S.A. in this one, with a series win probability of .576. It's interesting to note Phoenix is more likely to win a sweep than the Spurs (7% to 4.7%), and is also more likely to win it in six (16.5% to 15.2%). Still, having games 5 and 7 at home lets the Spurs move on, if log5 gets its way.

log5-based prediction: San Antonio in 7.

Houston v Utah 

Based on these teams' expected win percentages, they could play 100 times and each come up 50-50. It went seven last year, and log5 says it'll be going seven again (37.4%). Houston's home court advantage coaxs log5 into picking them by the slimmest of margins -- 50.7% to 49.3%. I'm not sure log5's mom would endorse this one.

log5-based prediction: Houston in 7.

There it is. Do with it what you will.