Three weeks into the NBA season, the standings don’t look quite as expected. With the majority of teams having played at least 10 games – fingers crossed – in three years in the first 82-game campaign, favorites like the Bucks and Lakers are in play-in position, while the conference’s top tiers are populated by more surprising ones. candidates .
The biggest early overachiever, relative to expectations, is Golden State, who went into the season with an over/under projection of 48.5 wins, according to Basketball-Reference’s archive of Vegas odds. That translates to 5.9 expected wins out of every 10 games, but the Warriors are a league-best 9-1, meaning they are already 3.1 wins, or 31 percent, above expectations.
Two teams that were expected to end up losing records have also started the season hot, making jumps of similar size. The Wizards are 7-3, as opposed to an over/under of 33.5 wins, or 4.1 wins in 10 games. And the Cavaliers – who went into the season with the fourth-lowest over/under total in the league (26.5 or 3.2 wins in 10 games) – have started 6-4 (now 7-4), in in fact double the expectation.
Generally, Vegas odds over an entire season are pretty good indicators of the quality of the team. Looking at all 82 game seasons this century (a sample of 565 teams), the 10 teams that started the season with the best overs/unders finished with an average record of 59-23; the 10 worst over/unders teams finished, handily, with the reverse average record of 23-59. Overall, Vegas projections alone could predict 56 percent of teams’ final records over this time period (the R-squared between the two variables).
But once a season starts, what happens in the case of teams like the Warriors or Wizards or Cavaliers? Should their warm starts affect our impression of their future chances, contrary to preseason expectations? Or, conversely, should we lose some faith in early underperformers like the Bucks, who went into the season with a 54.5 over/under win but started 4-6?
The answer, at least for now, isn’t much to change. Consider the biggest overachiever by 10 games in the 2018-19 season (the last with 82 games): Denver, which started 9-1 despite an over/underprojection of 47.5 wins. Over the remainder of the season, the Nuggets declined, winning at a pace of 51 wins – much closer to their original projection than their game over the first 10 games.
In many cases, the most surprising teams perform early worse than expected for the rest of the season, albeit by a small margin. The 2017-18 campaign featured two big early overachievers: The Pistons started 7-3 despite an over/under of just 38.5, then went on to win the rest of the season at a pace of 36.5 wins; the Knicks started 6-4 despite an over/under of 30.5, winning the rest of the way at a pace of 26 wins.
In the previous season, the Lakers were the biggest bouncer, starting 6-4 despite an over/under of 24.5. They won the rest of the season at a pace of 23 wins. The second biggest overperformer was the Hawks, who started 8-2 despite an over/under of 43.5, then went on to win the rest of the way at a pace of 40 wins.
And so on. Overall, we see that the 10 percent of the teams in this sample that performed best in their first 10 games, relative to expectations for the season, won 77 percent of their first 10 games, versus a Vegas expected win rate of only 49 percent. For the remainder of the season, they won 55 percent of their games — much closer to preseason estimates. The same relationship appears with early underachievers, in the opposite direction.
So if 10 matches aren’t enough to significantly change the perception of a team’s quality, how much data do we need to come to firmer conclusions? The answer is 25 games. After 10 games, a team’s record offers a 37 percent prediction for its performance over the remainder of the regular season; at the same time, a team’s Vegas projection for the season provides a more robust 54 percent predictability. But by Game 25, an early-season team’s actual record becomes more predictive than the future Vegas projection.
Another way to predict the future success of teams is to use points differential, which is useful faster than record as it offers a much wider range of outcomes than the simple binary win-lose. Point difference overtakes Game 17’s Vegas projections and surpasses it in terms of predictability for the remainder of the season by Game 18. This timeline generally agrees with previous research from the athletic‘s Seth Partnow, who used a different method of analysis and found that a team’s net rating becomes more stable after the first 15-20 matches.
It’s immediately apparent why deviations from early-season expectations often look flimsy in retrospect: a host of contextual factors can influence a team’s first record, from injuries (and this season, COVID-19-related absences) to randomness. luck to plan inequalities. For example, the Wizards are 5-0 in ‘matching games’ having been within five points in the past five minutes – a certainly unsustainable pace.
And the Warriors’ nine wins so far are against the league’s easiest schedule, according to Basketball-Reference, as they’ve already feasted on the Rockets, Pelicans and Thunder twice and haven’t met another top team yet. (No, the Lakers haven’t looked like a top team so far this season.) That disparity will be smoothed out somewhat in Game 25, as the Warriors play twice against Brooklyn, Philadelphia, and Phoenix, among other higher-ranking teams during that period. quality. — another reason to hold off on making statements about their bizarre performance so far beyond expectations.
Those early season wins count, of course — even if a 9-1 record doesn’t do much for predicting the Warriors’ record for the rest of the season, that’s nine wins on a bench and a cushion of 3.5 games on the play-in spots, which could prove vital as the western titans battle for seed later in the season. And there is reason for legitimate encouragement about the prospects of these three teams, from the more diversified attack in Golden State to the deeper rotation in Washington to the reciprocal rise of Evan Mobley in Cleveland.
But year after year, teams that have made surprisingly scorching starts fall back to their expected level as the season progresses. It is important not to overreact to the vagaries and caprices inherent in a small portion of a season. There’s not much reason to predict the playoffs or title races until Thanksgiving at the earliest — about the 18-game mark, when the point difference becomes a more meaningful indicator. By then, everyone will be more than excited to spin forward.