What Can We Learn from NBA Preseason?
Preseason NBA games are valuable for different reasons.
Is offering a glimpse of what's to come in the regular season one of them?
Exhibition contests are, at their heart, not supposed to be crystal balls. The best players don't routinely log big minutes. Entire quarters, halves and sometimes games are devoted to shoring up rotations and roster spots instead.
TOP NEWS
Braun: I'm Nuggets' Leader 🤔
Former NBA player fires back at Draymond
Finch Questions Wemby Blocks🤔
That means marginal players are typically the ones who most impact the outcomes of games. Ergo, the NBA's preseason shouldn't matter beyond the extent that it shapes a team's regular-season cosmetic makeup.
But what if it does? What if collective exhibition performances—wins, losses, points scored and allowed—portends the future?
What if preseason basketball matters more than we thought?
Measuring Relationships
Let's start with how we'll get to where we're going.
Direct comparisons between preseason and regular-season performances might seem misguided and ambitious. The disparity in sample size—usually six to eight preseason games weighted against 82 regular-season contests—is enormous and cannot be ignored.
But the results cannot be dismissed because of volume alone either.
In an attempt to fix the NBA's back-to-back scheduling problem, ESPN.com's Tom Haberstroh (subscription required) reminded us that bigger sample sizes can be superfluous, not better:
"Forget the health aspect for a second. In its current form, the NBA season is excessively predictable. How predictable? Let's look at the data. Brilliant research by FiveThirtyEight's Neil Paine found that we learn as much about the true abilities of an NBA team after 22 games as we do about an MLB team after it plays its full 162-game slate.
Let that marinate for a second. Twenty-two NBA games. That's all it takes.
"
If 22 games is enough to render season-defining conclusions, there's risk in glossing over samples that, in theory, can account for roughly one-third of the bigger picture. Yes, the rotations are different and the level of play inferior, but that's also why we're here: to see if any of this matters. That includes determining whether the sample size is too small or overly manipulated by different agendas.
Focus will lie within point differentials to do this. They're a better measurement of a team's standings than wins and losses, since they better account for conference imbalance and overall strength of schedule.
Regular-season point differentials typically average out to zero, thus allowing us to classify teams as above-average and below-average. But since not all teams play the same number of exhibitions each year, preseason means will vary.
Fear not, though. This doesn't change the spirit of designation, just the number teams are measured against.
Preseason vs. Regular Season
Last year's preseason results provided some valuable insight.
Fifteen teams registered above-average point differentials during the 2013-14 preseason. Eleven of them (73.3 percent) would go on to notch positive—above-average—point differentials during the regular season. Of the 15 squads with below-average preseason point differentials, 10 of them (66.7 percent) recorded below-average marks in the regular season as well.
That leaves last year's preseason displays as solid indicators of whether or not teams would eventually end the regular season with better-than-normal point differentials. Eleven of those 15 above-average units would make a playoff appearance. Another one of them—the Phoenix Suns—still managed to rattle off 48 wins in the brutal Western Conference.
Can the same be said of a larger sampling?
Those willing to brave the cluster of data from the last 10 years can look at how the 300 different teams charted here. But if you're cluster-phobic—or value your eyes—rest easy.
Of the 300 teams evaluated, 157 mustered average or above-average point differentials during the preseason. And of those factions, 64.3 percent (101) did the same come regular season.
Similar results are found when looking at the 143 teams with below-average preseason differentials. Sixty-seven percent of them (96) closed out the regular season in the red.
Put another way, point-differential classifications for 65.7 percent of teams (197) since 2004 have been mirrored in the regular season. That's almost two-thirds of the league, which, admittedly, isn't overwhelming.
The best we can say for now is an above-average preseason most likely precedes an above-average regular season.
Let's Hear It For the Champs
When in doubt, turn to the champions.
Title-toting teams are going to have good point differentials. To survive the regular season and put themselves in position to contend for a championship, they have no choice.
As for whether that success begins in the preseason, most of the time the answer is no.
Below you'll see how each championship team over the last 10 years fared through the preseason and regular season. Remember, preseason scores represent how each club fared against the league's exhibition average:
Four of the last 10 champions came out of the preseason with below-average point differentials, including the most recent winner, the San Antonio Spurs. And therein lies one of the major issues.
Point differential swings between preseason and the regular season can be insane. The Spurs increased theirs by 12.2, from minus-4.5 to 7.7, last year. Seven of the last 10 champions experienced positive jumps of at least four points between the preseason and regular season. One of the three exceptions, the 2011-12 Miami Heat, saw theirs go 8.5 points in the wrong direction—from 14.5 to 6.0—because the lockout-truncated campaign allotted only two preseason games.
Consider that a team's average point-differential swing over the last decade—whether it moves down or up from preseason to regular season—is nearly 5.2 points. To put that in further perspective, only 42 of the 300 teams we looked at (14 percent) finished with regular-season margins of equal or greater value.
Winning seven of eight games and posting the seventh-highest point differential (6.5) last preseason ended up meaning little for the New Orleans Pelicans. They went on to win 34 games and tally the Western Conference's fourth-worst point margin (minus-2.7).
When there's that much annual movement, it's difficult if not impossible to make any significantly specific judgments.
Proceed with Caution
Every team is going to value the preseason differently, using it to test different things and answer various questions.
Take the Spurs. Coach Gregg Popovich made his feelings on the importance of preseason basketball perfectly clear last November.
"This is the preseason, right?" he exclaimed when TNT's Craig Sager attempted to interview him. "We gotta do this in the preseason? Are you kidding me? In the preseason we gotta do this? The preseason?"
Asked about the biggest challenge during this time of year, Popovich responded in kind:
"Getting to the restaurant on time."
Somewhere, buried deep inside Popovich's exasperated temper tantrum, is something worth taking away: Preseason basketball is complicated.
Teams that have their rotations and starting lineups set, like the Spurs, won't necessarily take it as seriously. They'll play their best players even less than usual in favor of fringe roster candidates who perhaps might not be as locked-in when chances of making the team are slim.
Other squads that are fine-tuning and still figuring out prominent personnel situations will, in turn, be playing for something different. That something needs to be acknowledged accordingly—as does preseason basketball itself.
Kevin Pelton, writing for Basketball Prospectus in 2012, put it best:
"It's easy to get seduced by preseason play. Since it's the first we've seen any of these players in months, there's a bit of "what you see is all there is" at play in elevating results into something more meaningful than it really is. Still, be careful not to overcorrect for this tendency. There is some meaning to the preseason. The trick is sorting the useful results from the misleading ones, and that's part science and part art.
"
Basically, proceed with caution, even though—as mentioned before—65.7 percent of teams analyzed were dealt similar regular-season hands.
Despite what we know and what we have the ability to find out, the new-look Cleveland Cavaliers won't be crowned Eastern Conference champions during the preseason. Kobe Bryant's play in October won't foreshadow every aspect of his game in April.
What you see now isn't an all-inclusive look at what's coming, especially when it comes to established title contenders.
But it's enough to gain a strong idea of how the regular season will inevitably unfold.
Stats courtesy of ESPN.com unless otherwise cited.
