The theme of the young 2020-21 NBA season is unpredictability.
Through the first three weeks of action, an NBA campaign being played during the midst of the COVID-19 pandemic has been marked by upsets, inconsistency and lopsided scores. With more teams forced to play without key players due to the league's health and safety protocols, that doesn't figure to change anytime soon.
Just how difficult has the 2020-21 season been to predict so far? What helps explain the inconsistency from game to game? Let's take a look at the historic numbers.
Blowouts on the rise
We got an early indicator this NBA campaign might look very different in terms of results during the season's first weekend, when the Dallas Mavericks took a record-breaking, 50-point halftime lead on the road against the LA Clippers, one of the favorites to win the title.
Sure, there had been several lopsided games during the NBA's marquee Christmas Day slate of games, but all of those involved the favored team winning by more than expected. This was a case of a good team being completely blown off the floor by a lesser one -- at home. Per ESPN Stats & Information research, it was the third-largest margin of victory by an underdog team over the past 30 seasons.
With no fans in attendance for the majority of the league's teams as they resume play in home markets after finishing the 2019-20 season in a bubble, the lack of home-court advantage was predictable. So far this year, home teams have won slightly fewer than half of their games (48%) albeit with a neutral point differential.
There was no reason to expect limited home-court advantage to translate into more blowouts, however, and that's exactly what we've seen. Thus far, more than 10% of all NBA games have been decided by 20 points or more, which would be the highest mark in league history. Naturally, that might even out over time. But even if we limit to only teams' first eight games of the season, this is still the highest percentage of games decided by 20-plus points (12%) and 30-plus points (4%) over the same span since the ABA-NBA merger.
Atypically large performance swings
Inevitably, all those lopsided finishes mean it's tougher than ever to know what to expect from a team game to game. If you watched the Mavericks dominate the Clippers, they probably looked like championship contenders to you. Not so much if you watched their games immediately before (a 23-point loss on the same court to the Los Angeles Lakers on Christmas) or after (a 19-point home loss to the Charlotte Hornets).
Somewhat amazingly, neither the Mavericks nor the Clippers have been the league's most inconsistent team in terms of game-to-game variation of their point differential. That honor belongs to the Milwaukee Bucks, who have won games by 47, 39 and 29 points so far this season but also suffered a 20-point loss to the New York Knicks.
Again, let's limit this to the first eight games of the season to put 2020-21 on a level playing field with past years. In terms of the standard deviation of point differential game to game, the three most inconsistent teams in NBA history are all from the 2020-21 season, with Dallas as a fourth 2020-21 team in the top six.
Most inconsistent NBA teams (standard deviation of point differential)
1. Milwaukee Bucks (2020-21): 23.9
2. Golden State Warriors (2020-21): 23.8
3. LA Clippers (2020-21): 23.7
4. Portland Trail Blazers (1991-92): 23.6
5. Boston Celtics (1966-67): 23.5
6. Dallas Mavericks (2020-21): 23.4
Through first eight games
All that variation has produced a number of unexpected outcomes such as the Mavericks' win over the Clippers. According to ESPN Stats & Information research, teams that entered games as underdogs won an incredible 48% of games through the first 17 days of play this season -- making being the favorite little better than a coin flip. As you might imagine, that's much higher than usual. Over a full season, underdogs have won no more than 33% of games over the past three decades.
Like the number of blowouts and level of inconsistency, individual games being so unpredictable is also without modern precedent to start an NBA schedule. Over the same number of days of play, the highest win percentage for underdogs was previously 42.5% in 2003-04. Only in one other season, 2006-07, did underdogs begin the year by winning even 40% of the time.
Make or miss league to an extreme
One factor in particular seems to explain the inconsistency, unpredictability and blowouts we've seen this season: hot and cold 3-point shooting. Consider the differential between 3s made by the winning and losing teams, which has increased as 3s have become more frequent over time. Before this season, the biggest differential between winning and losing teams was 2.0 per game in 2018-19. This season, that has jumped all the way to 3.0 per game.
Naturally, that margin is even larger in blowouts. Remember that Mavericks-Clippers game? Dallas outscored the Clippers 42-12 beyond the arc as the Clippers shot 4-of-33 on 3s. Overall, the winners of games decided by 20-plus points have averaged an incredible 7.6 more 3-pointers made than their opponents have. Last season, that margin was 4.9 per game.
The modest increase in 3-point attempts (39.5% of all shot attempts, up from 38% in 2019-20) can't really explain the changes we've seen. Instead, it just appears that good shooting nights for one team have matched up with bad nights for the other far more than usual.
The abbreviated nature of training camps and preseason ahead of the 2020-21 regular season also looks like a factor in the inconsistency we've seen. Returning to the standard deviation of game-by-game point differential over the first eight games, averaged across all teams, 2020-21 is the highest in NBA history. No. 2? The 2011-12 season, which also begun abruptly after that year's lockout. (Oddly, teams were unusually consistent early in the 1998-99 season after a lockout.)
COVID-19-related absences on the rise
As my colleague Tim Bontemps detailed over the weekend, the NBA's ability to power through a pandemic without a protective closed campus is already being tested by growing numbers of players unavailable due to the league's health and safety protocols -- either due to a positive test or because they've been identified by contact tracing as a close contact to someone who tested positive. That figures to only exacerbate the inconsistency we've seen so far this season.
Injuries are already a factor that leads to unpredictability. Above and beyond the uncertain long-term health effects for players contracting the virus, the effect of COVID-19 on rosters in the short term is extreme because a positive test can sideline both the player who tests positive and a handful of close contacts, who are at increased risk of subsequently contracting the virus.
Unlike NCAA basketball teams, the NBA has attempted to follow the NFL's model of isolating close contacts in an attempt to prevent the spread of the virus among and between teams without shutting down their activity entirely. Whether that effort can succeed is dependent on the league's ability to perfectly identify close contacts. Otherwise, continuing to play risks COVID-19 spreading throughout rosters, as we've seen in college sports and other professional leagues.
The NBA's decision to go ahead with the Philadelphia 76ers game against the Denver Nuggets on Saturday despite the Sixers being reduced to seven active players in the wake of a positive COVID-19 test for Seth Curry has shown that -- as in the NFL -- ensuring teams can be competitive won't be part of the league's consideration. The show, it appears, will go on. (Sunday's Boston Celtics-Miami Heat matchup was subsequently postponed, the second such game this season, though because of the time required for contract tracing of the Heat rather than because the Celtics had just eight players available.)
According to tracking by my colleague Bobby Marks, 29 players around the league -- nearly 6% of all players -- are currently subject to the league's health and safety protocols. All told, players have already missed more than 100 games due to COVID-19-related absences, and those haven't been distributed evenly. So far, 11 teams have yet to have a player miss a game due to health and safety protocols, while the Chicago Bulls have had four players (starter Lauri Markkanen and reserves Ryan Arcidiacono, Chandler Hutchison and Tomas Satoransky) miss a total of 28 games.
Unfortunately, the pace of absences seems to be speeding up rather than slowing as the season continues. Until the COVID-19 vaccination process reaches a point where players can receive it, the risk of players contracting and spreading the virus will be a source of unpredictability in a season that already had more than enough.