THE POWER RANK

Win probabilities for round 2 of the NBA playoffs, 2014

By Leave a Comment

nba_playoffs_2014_r2What an incredible first round of the NBA playoffs. Five games went to game 7, and a sixth easily could have if not for a buzzer beater by Portland over Houston.

The visual shows The Power Rank’s new odds for winning a title. Here are the odds for the 4 series of the second round.

Portland vs San Antonio.
San Antonio has a 72.9 percent chance of winning the series.

Los Angeles Clippers vs Oklahoma City.
Oklahoma City has a 56.2 percent chance of winning the series.

Washington vs Indiana.
Indiana has a 69.8 percent chance of winning the series.

Brooklyn vs Miami.
Miami has a 83.2 percent chance of winning the series.

Check out the interactive visual for the 2014 NBA playoffs

By Leave a Comment

nba_visual_2014You’re interested in which team will win the NBA playoffs.

While analytics can help, you don’t want to go blind squinting at a table of numbers. Visualization is a better way to view data.

The Power Rank has collaborated with Andrew Garcia of Chartball to make an interactive bracket with win probabilities for the NBA playoffs. Hover over a team or circle to see a team’s chance to advance through the rounds.

To play with the bracket, click here.

How to calculate win probabilities

It starts with the NBA team rankings from The Power Rank. These rankings consider margin of victory and adjust for strength of schedule.

Strength of schedule matters this year more than most, as the Western conference has dominated the East. Only two teams from the East, Miami and Indiana, crack the top ten in these final rankings after the regular season.

1. San Antonio, (62-20), 6.50
2. Oklahoma City, (59-23), 6.06
3. Los Angeles Clippers, (57-25), 5.73
4. Houston, (54-28), 4.50
5. Miami, (54-28), 4.08
6. Golden State, (51-31), 4.03
7. Portland, (54-28), 3.78
8. Indiana, (56-26), 3.34
9. Phoenix, (48-34), 3.15
10. Dallas, (49-33), 2.44
11. Toronto, (48-34), 2.39
12. Minnesota, (40-42), 2.15
13. Memphis, (50-32), 1.85
14. Chicago, (48-34), 1.76
15. Washington, (44-38), 0.64
16. Brooklyn, (44-38), -0.72
17. Atlanta, (38-44), -0.83
18. Charlotte, (43-39), -0.88
19. Denver, (36-46), -1.38
20. New York, (37-45), -1.53
21. Sacramento, (28-54), -2.10
22. New Orleans, (34-48), -2.25
23. Cleveland, (33-49), -3.12
24. Detroit, (29-53), -3.30
25. Boston, (25-57), -4.33
26. Orlando, (23-59), -4.92
27. Los Angeles Lakers, (27-55), -5.04
28. Utah, (25-57), -5.39
29. Milwaukee, (15-67), -7.87
30. Philadelphia, (19-63), -8.74

The calculation by the ranking algorithm also gives each team a rating (the number after each team’s record). The difference in the rating between two teams gives a predicted point spread. After adding a factor for home court advantage, this point spread is converted into a win probability.

Then a recursive algorithm calculates the win probabilities for a series. Multiplying these series win probabilities gives the likelihood that a team advances to later rounds.

Win probabilities for the NBA championship

nba2013_playoff_probThis visual ranks teams by their chance of winning the NBA championship.

As the top ranked team, San Antonio has the largest probability at 27.4%. However, they must first survive a wicked Western conference. The Spurs play the Mavericks, the 10th ranked team in the entire NBA, in the first round.

Miami finished the regular season ranked 5th. However, they have the 3rd largest win probability due to a weak Eastern conference. Miami has a higher win probability than the Los Angeles Clippers despite a rating almost 2 points worse.

Indiana gets a similar benefit from playing in a weak Eastern conference.

These probabilities appear at the center of the interactive bracket. To play with the numbers, click here.

3 point defense in the NBA: skill or luck?

By 3 Comments

Have you ever wondered about how much control a defense has over an opponent’s 3 point shooting?

On the one hand, there is obvious skill in closing out a shooter. This comes from proper positioning on the floor and alertness in identifying shooters.

On the other hand, jump shots from distance introduce randomness into the game of basketball. Every fan knows that feeling when their shooter gets hot and carries the team on a run. Then at other times, the shooter can’t throw the ball in the ocean. Live by the 3, die by the 3.

Analytics on 10 years of NBA data

I’ve been wondering about this for years, so I dug into NBA team data over the last 10 years. Two facts emerged.

First, there is randomness in team’s 3 point defense. Let’s measure 3 point defense by the percentage that opponents shoot from behind the arc. Then 3 point defense one season explains 2.7% of the variance in 3 point defense the next season, a correlation coefficient of 0.17. This percentage regresses strongly to the mean.

However, that’s not the entire story. Teams that play better defense within the arc tend to have better 3 point defense. Measured by field goal percentage, a team’s 2 point defense explains 20% of the variance in 3 point defense, a correlation coefficient of 0.45. Ken Pomeroy has found similar results for college basketball.

While there isn’t a strong correlation between 2 and 3 point defense, you can’t ignore that strong defensive teams, like Indiana last season, tend to allow lower 3 point field goal percentages. Indiana led the NBA in both 2 and 3 point field goal defense last season.

Preseason predictions for 2013

Analytics suggests that teams with poor 2 point defense but strong 3 point defense last year will not get so lucky again.

What if we used this idea before the 2012-2013 (last season as of this writing) to predict a team’s 3 point defense? In 2011-2012, New Orleans and Houston finished 2nd and 4th respectively in 3 point defense. However, both teams finished in the bottom 10 for 2 point defense. The next season, New Orleans and Houston ranked 27th and 22nd in 3 point field goal defense respectively.

What does that mean for the 2013-2014 season (which starts tonight)? Here are teams ranked by 3 point defense last season. Their rank in 2 point defense and field goal percentage are in parentheses.

1. Indiana 32.7% (1. 44.4%).
2. Memphis 33.8% (3. 46.5%).
3. Portland 34.0% (29. 51.2%).
4. Boston 34.2% (8. 47.2%).
5. Chicago 34.6% (5. 46.7%).
6. Oklahoma City 34.6% (2. 45.2%).
7. Golden State 34.7% (11. 47.3%).
8. Milwaukee 34.8% (14. 48.3%).
9. Philadelphia 34.9% (19. 48.7%).
10. Washington 34.9% (7. 47.0%).
11. Miami 35.0% (10. 47.3%).
12. San Antonio 35.3% (4. 46.6%).
13. Sacramento 35.5% (30. 51.2%).
14. Los Angeles Lakers 35.7% (15. 48.3%).
15. New York 35.7% (23. 49.8%).
16. Orlando 35.9% (25. 49.8%).
17. Dallas 35.9% (13. 47.5%).
18. Detroit 36.0% (22. 49.2%).
19. Toronto 36.1% (18. 48.7%).
20. Denver 36.3% (12. 47.4%).
21. Brooklyn 36.6% (20. 49.0%).
22. Houston 36.8% (16. 48.4%).
23. Minnesota 36.9% (24. 49.8%).
24. Utah 37.0% (17. 48.5%).
25. Cleveland 37.2% (28. 51.1%).
26. Los Angeles Clippers 37.3% (6. 46.8%).
27. New Orleans 37.4% (27. 50.5%).
28. Atlanta 37.9% (9. 47.3%).
29. Charlotte 38.8% (26. 50.2%).
30. Phoenix 38.8% (21. 49.2%).

Over at bettingexpert, I wrote about the implications of 3 point defense for Atlanta, the Los Angeles Clippers and Portland.

Video of Los Angeles Clippers and their “poor” defense

The Los Angeles Clippers are team that should have better luck in 3 point defense. However, this NBA.com video disagrees. It talks about how poor they defended the 3 point line, putting the blame on Carron Butler and Willie Green. Moreover, they show 6 examples of opponents making 3 point baskets.

Personally, I thought the Clippers closed out 2 of those shooters quite well. They gave up a third wide open 3 pointer off an offensive rebound. While the 3 other shots were poor defense, you can find three open 3 pointers against any team. Since the Clippers usually play good defense (6th in 2 point defense), expect the 3 point defense of the Clippers to rebound this season.

Thanks for reading.