NBA Teams That Are Better or Worse Than Their Records Suggest
A team's win-loss record is not always the best indicator of how good or bad said team really is. Simple wins and losses don't actually tell all that much of the story, paying no mind to important factors like margin of victory (MOV), offensive efficiency, defensive efficiency, clutch performance, strength of schedule, and so on.
Pythagorean expectation is a concept developed by baseball statistician Bill James as a way to get a more complete picture of a team's performance than by simply going off win-loss record.
Pythagorean wins and losses estimate how many games a team "should" have won to date based on the number of points it scores and allows. The idea is that factoring margin of victory into a team's record is a way to lessen the impact of some of the dumb luck that occurs in close games that could go either way.
Whereas some standard wins and losses are determined by a single moment -- a fluke play, questionable call or miracle last-second shot -- measuring expected wins and losses effectively eliminates those high-variance occurrences from the equation. The result is a more representative ranking system that separates the strong teams from the weak, without leaving much up to chance or lucky bounces.
Here is a list of all 30 NBA teams, ranked by the differential in percentage between their current win-loss record and their Pythagorean win-loss record. Comparing the two percentages can show us which teams are better than their actual record suggests and which are worse. Teams near the top of the list that have more Pythagorean wins (PW) than real wins, and teams near the bottom that have more real wins than Pythagorean wins.
Rank | Team | Wins | Losses | MOV | PW | PL | Difference |
---|---|---|---|---|---|---|---|
1st | Brooklyn Nets | 9 | 47 | -8.75 | 14 | 42 | +5 |
2nd | Minnesota Timberwolves | 22 | 35 | -0.89 | 27 | 30 | +5 |
3rd | Charlotte Hornets | 24 | 32 | 0.25 | 28 | 28 | +4 |
4th | Milwaukee Bucks | 25 | 30 | 0.27 | 28 | 27 | +3 |
5th | Toronto Raptors | 33 | 24 | 4.14 | 36 | 21 | +3 |
6th | Dallas Mavericks | 22 | 34 | -2.29 | 24 | 32 | +2 |
7th | Miami Heat | 25 | 32 | -1.00 | 27 | 30 | +2 |
8th | Utah Jazz | 35 | 22 | 4.54 | 37 | 20 | +2 |
9th | Denver Nuggets | 25 | 31 | -1.18 | 26 | 30 | +1 |
10th | Detroit Pistons | 27 | 30 | -0.44 | 28 | 29 | +1 |
11th | Phoenix Suns | 18 | 39 | -5.46 | 19 | 38 | +1 |
12th | Golden State Warriors | 47 | 9 | 12.82 | 47 | 9 | 0 |
13th | Houston Rockets | 40 | 18 | 6.45 | 40 | 18 | 0 |
14th | Los Angeles Clippers | 35 | 21 | 3.80 | 35 | 21 | 0 |
15th | New Orleans Pelicans | 23 | 34 | -3.14 | 23 | 34 | 0 |
16th | New York Knicks | 23 | 34 | -3.12 | 23 | 34 | 0 |
17th | Portland Trail Blazers | 23 | 33 | -2.80 | 23 | 33 | 0 |
18th | Los Angeles Lakers | 19 | 39 | -6.28 | 18 | 40 | -1 |
19th | Chicago Bulls | 28 | 29 | -0.67 | 27 | 30 | -1 |
20th | Sacramento Kings | 24 | 33 | -2.72 | 23 | 34 | -1 |
21st | San Antonio Spurs | 43 | 13 | 8.36 | 42 | 14 | -1 |
22nd | Memphis Grizzlies | 34 | 24 | 1.29 | 32 | 26 | -2 |
23rd | Indiana Pacers | 29 | 28 | -0.91 | 27 | 30 | -2 |
24th | Cleveland Cavaliers | 39 | 16 | 5.38 | 37 | 18 | -2 |
25th | Washington Wizards | 34 | 21 | 2.80 | 32 | 23 | -2 |
26th | Boston Celtics | 37 | 20 | 2.75 | 34 | 23 | -3 |
27th | Philadelphia 76ers | 21 | 35 | -5.61 | 18 | 38 | -3 |
28th | Orlando Magic | 21 | 37 | -6.59 | 17 | 41 | -4 |
29th | Oklahoma City Thunder | 32 | 25 | -0.11 | 28 | 29 | -4 |
30th | Atlanta Hawks | 32 | 24 | -0.46 | 27 | 29 | -5 |
Observations
The Brooklyn Nets are five wins better than their 9-47 record suggests (tied for the biggest discrepancy in the league), but even their Pythagorean win-loss record of 14-42 is the worst in the NBA. Any way you slice it, they stink.
The Atlanta Hawks are perhaps the luckiest team in the league, having a 32-24 record that's five games better than their Pythagorean mark of 27-29. They are currently eight games over .500 and the 5 seed in the Eastern Conference, but if the playoffs were determined by Pythagorean wins, they'd be two games under .500 and in 8th (behind both the 9th-place Milwaukee Bucks and 11th-place Charlotte Hornets and only 0.5 games up on the 10th-place Miami Heat).
The race for the 8 seed in the Western Conference is an interesting one, with the Denver Nuggets, Sacramento Kings, Portland Trail Blazers, New Orleans Pelicans, Dallas Mavericks, and Minnesota Timberwolves all separated by only 3.5 games for the final playoff spot. If that final postseason berth were decided by Pythagorean wins, it would go to the Timberwolves with a Pythagorean record of 27-30, despite the fact that they are currently behind the rest of the pack in 13th in the real win-loss standings with only a 3.0% chance of making the postseason, according to our algorithms.
The Toronto Raptors have the biggest jump from regular wins to Pythagorean wins of any current playoff team with a difference of three. The Boston Celtics and Washington Wizards have been on fire of late while the Raptors have been ice cold, but Pythagorean wins and margin of victory suggest that the Raptors have still been the best team of the three and the second-best team in the Eastern Conference on the season as a whole. Let's see if the addition of Serge Ibaka helps them in reclaiming that position in the actual standings.
Going the other way, the Oklahoma City Thunder have the biggest dropoff from real wins to Pythagorean wins of any current playoff team, with their record of 32-25 coming out as 28-29, a difference of four games. They are currently hanging onto the 7 seed in the Western Conference at seven games above .500, but Pythagorean wins would put them one game below (although still in seventh).
The Golden State Warriors (47-9), Houston Rockets (40-18), Los Angeles Clippers (35-21), Portland Trail Blazers (23-33), New Orleans Pelicans (23-34), and New York Knicks (23-34) all have identical records by both standard and Pythagorean wins and losses. In other words, their current win-loss records are realistic representations of how good (or bad) each of those teams is.