Game 7, 2016 NBA Finals. Stephen Curry, in transition, dribbles between his legs at the top of the key, takes one half-step left, and releases a 28-foot three-pointer with eleven seconds on the shot clock. The ball misses. LeBron James grabs the rebound, the Cavaliers rebound the loose energy of the building, and four minutes later they win the championship. The footnote is this: Curry took thirty-three contested jumpers off the dribble in those Finals, almost all of them from beyond twenty-three feet, in a series that, eight years earlier, would have been considered tactical malpractice. By 2016, it was the league’s most successful offense.
That night did not invent the pace-and-space NBA. It only made it impossible to ignore. Somewhere between Mike D’Antoni’s Phoenix Suns and Daryl Morey’s Houston Rockets, between Don Nelson’s Warriors and Steve Kerr’s, the league quietly stopped being the sport it had been for forty years. The thing that we measure on the floor, and the way we measure it, both changed. The metrics that survived from the 1990s — points per game, field goal percentage, even the box score itself — became unreliable narrators of what was actually happening. New ones rose in their place. The story of the modern NBA is, in a real way, the story of measurement catching up to a sport that had outrun its own statistics.
I have been writing about basketball analytics since 2014, mostly for a rotation of team blogs and my own newsletter, and the single conceptual shift that most fans still under-appreciate is the one this article is about. The pace-and-space era, the metrics it forced into existence, the metrics it broke, and how to read a 2026 box score without 1996 instincts, is the subject of this article.
The origin: where pace and space came from
Pace, in basketball, is a measurement of possessions per 48 minutes. A possession ends when a team takes a shot that ends in a basket, a defensive rebound, or a turnover. A high-pace team plays more possessions per game; a low-pace team plays fewer. The metric existed long before anyone treated it as central. Dean Oliver’s Basketball on Paper, published in 2004, formalized the math. Daryl Morey, John Hollinger, and Kevin Pelton spent the late 2000s carrying it into mainstream coverage.
The strategic insight under “pace and space” came from two adjacent observations. First, the corner three is the most efficient shot in basketball — the line is closer to the basket than from the top of the key, and the shot is worth fifty percent more than a two-pointer. Second, when you put more shooters on the floor, defenders have to space out, which opens lanes for drives and passes. Mike D’Antoni’s Suns of 2004-08 ran the prototype: seven seconds or less, four shooters around a creative center in Amar’e Stoudemire, and a willingness to take threes that the league treated as eccentric. They never won a title. The teams that came after them did.
By the mid-2010s, the analytical case had been made and adopted. The 2014-15 Warriors led the league in pace and three-point attempts and won the championship. The 2015-16 Warriors broke the wins record. The 2017-18 Rockets posted 50.0 three-point attempts per game, an inconceivable number a decade earlier, and pushed the Warriors to seven games in the conference finals. The league office, watching this, did not slow it down. They sped it up — through reset clock rules, freedom-of-movement enforcement, and a defensive philosophy shift that effectively banned hand-checking. The pace-and-space NBA is, in part, a designed product. The math was just the instigator.
How it works: pace and space in plain language
The arithmetic is simpler than the coverage often makes it sound. A possession is a chance to score. A team that plays a fast pace gets more chances. A team that takes more efficient shots scores more points per chance. Multiply the two together and you have offensive output.
The single most useful number that came out of this era is offensive rating: points scored per 100 possessions. A team’s offensive rating is independent of pace, which means it lets you compare a slow-paced Memphis Grizzlies team to a fast-paced Sacramento Kings team on equal terms. The 1995-96 Bulls, generally considered one of the great offensive teams in modern history, posted an offensive rating of 115.2. The 2024-25 Boston Celtics, in a year that felt almost mundane by their standards, posted 122.0. The league as a whole is now playing at an offensive rating that would have been historic thirty years ago. That is not because today’s players are better than 1996 NBA players, though some certainly are. It is because the math of shot selection has shifted.
Spacing, the other half of the phrase, is harder to measure but easier to see. A team with four or five shooters on the floor spreads the defense over a larger area. The space that opens up in the middle of the floor — the “paint”, “the dunker spot”, “the short corner” — becomes more available for drives, lobs, and cuts. A team with two non-shooters on the floor — say, a traditional center and a defensive specialist guard — gives the defense permission to collapse, which makes everything else harder. The math here is intuitive: if defenders have to respect more shooters, each individual defender has more ground to cover.
The critical component: shot value, not shot volume
The single most important conceptual breakthrough of the pace-and-space era is that shots are not equal. A corner three is worth, on average, 1.13 points; a long two from 18 feet is worth, on average, 0.78 points. A team that takes ten corner threes per game and zero long twos is generating, on the same volume of attempts, roughly 3.5 more points per game than a team that flips that ratio. Over a season, that is the difference between the eighth seed and a contender.
Shot-quality models like effective field goal percentage (eFG%) and true shooting percentage (TS%) capture this directly. eFG% adjusts for the fact that a three is worth 50% more than a two; TS% folds in free throws. League-average TS% in 2024-25 hovered near 0.580. Elite scorers — Nikola Jokic, Stephen Curry, Kevin Durant — operate north of 0.620 in good years. The 1995-96 Bulls were league average by TS%, and they were considered a historically efficient team. The math has moved.
Pace-era metrics vs the classical box score: a comparison
The traditional box score, designed in the 1850s for baseball and applied to basketball later, was never built to measure shot quality, possession context, or lineup effects. The pace-and-space era forced a parallel toolkit. A short comparison:
| Classical metric | What it misses | Pace-era replacement | What that captures |
|---|---|---|---|
| Points per game | Pace, role, opportunity | Points per 100 possessions | Per-possession efficiency, era-neutral |
| Field goal %% | Three-point value, free throws | True shooting %% (TS%%) | Total scoring efficiency per attempt |
| Assists | Setup quality, role context | Assist points created, AST%% | Playmaking impact, per-possession volume |
| Plus/minus | Bench unit context, sample size | On/off splits, lineup net rating | True team impact when player is on floor |
| Steals & blocks | Defensive scheme, communication, gravity | Defensive RAPTOR, EPM-D | Holistic defensive contribution |
The honest version of NBA writing, in 2026, reads both columns. The classical numbers still describe what happened. The pace-era numbers explain what it meant. A writer who quotes only points per game is, in the modern league, telling you what a player scored without telling you whether they were efficient about scoring it.
What the data needs: inputs
The pace-era metrics depend on a few data layers that did not exist publicly until the late 2000s. The most important is possession counting itself — a clean, consistent method for marking the end of one possession and the start of the next. The NBA’s official method, codified in Basketball-Reference’s glossary, treats turnovers, made baskets, defensive rebounds, and end-of-period buzzers as possession terminators. Different sources can disagree by 1-2 possessions per game, which compounds over a season.
The second necessary input is shot location data. Without knowing where a shot was taken — corner vs above-the-break, paint vs mid-range, restricted area vs floater — you cannot calculate shot quality. The NBA’s player tracking system, deployed leaguewide in 2013, made this data public. Before that, the league had assists logged but not the spatial detail that lets us evaluate a shot diet.
The third input, increasingly central in the late 2010s, is play-by-play with lineup data: which five players were on the floor for which possessions. This is what makes on/off splits, lineup net ratings, and matchup analysis possible. Basketball-Reference and NBA.com both publish this in clean format. The 2010s analytics revolution would have been impossible without it.
Building the analysis: a working framework
Pace-and-space coverage, at its best, follows a pattern. Start with the offensive context — what pace, what shot profile, what efficiency. Then move to the personnel — who is on the floor, in which lineups, doing what. Then move to the defensive response — what coverage is being run, how the offense is exploiting it. Then close with the takeaway — what does this mean for the next game, the next series, the next iteration.
A practical workflow during the season:
- Check pace and offensive rating for the team you’re writing about, season-to-date and last 10 games. Pace shifts can signal coaching adjustments or fatigue.
- Audit the shot diet. What percentage of attempts come from threes? From the paint? From the mid-range? The mid-range has not disappeared — Kevin Durant and DeMar DeRozan have built careers on it — but for most teams, an above-average mid-range diet is a yellow flag.
- Check the lineup net ratings. The closing five is the lineup that matters in the fourth quarter. A team whose closing lineup has a +12 net rating across 200 minutes is a different team than one whose closing lineup is +1 across 80 minutes.
- Look at the on/off splits for the team’s two or three highest-usage players. A star whose on/off is +9 is carrying real water. A star whose on/off is -1 is, at minimum, deserving of a longer conversation.
- Cross-reference with the eye test. The numbers tell you which lineups work. The film tells you why.
Where this gets weird: common mistakes
The pace-and-space metrics have known failure modes. Writers who use them well usually name the failure modes out loud.
Pace inflation distorts era comparison. A 110-point game in 1995 was an offensive explosion. A 110-point game in 2026 is a lazy Tuesday in November. Comparing players’ scoring across eras without adjusting for pace is the single most common analytical sin in modern coverage. Per-100 stats are not optional; they are the entry fee.
Three-point variance is real and underweighted. A team that goes 18-of-32 from three in one game and 8-of-32 in the next has had wildly different offensive results from very similar process. Single-game three-point shooting is mostly noise. The “hot shooting” narrative attached to a single performance is, in most cases, a story about variance. The team that took 32 quality threes is the team that, over a season, will win the shooting argument.
Lineup data lies in small samples. A five-man unit with a +20 net rating in 80 minutes is impressive on the leaderboard and almost meaningless as evidence. Lineup net ratings stabilize slowly. Two hundred minutes is a hint. Two thousand is an argument. The most-quoted lineup data in NBA coverage is, in fact, the least reliable.
Defensive metrics still trail offensive ones. The pace-era revolution solved offensive measurement. Defensive measurement is improving — defensive RAPTOR, EPM-D, the Cleaning the Glass defensive splits — but the public toolkit for evaluating individual defenders remains the weakest part of the modern stat stack. Steals and blocks are noisy proxies. The film still matters more here than in any other facet of the modern analytical conversation.
When pace-era metrics shine: use cases
The strongest applications:
Era-adjusted player comparisons. Was Michael Jordan in 1996 a more efficient scorer than LeBron James in 2024? The question is genuinely answerable using TS% and offensive rating, and the answer is more nuanced than most fans would expect. The math here lets you compare across eras without the usual hot-take noise.
Team strength assessment, in season. A 16-6 team with an offensive rating of 118 and a defensive rating of 112 is, by every analytical standard, a better team than a 17-5 team with ratings of 113 and 113. The records can lie. Per-100 numbers usually do not, in samples larger than 20 games.
Roster construction and offseason evaluation. When a team trades for a player, the most useful question is what their on/off and lineup data look like in the new context. A scorer who posted a +6 on/off in a system that featured them as the primary creator will rarely produce identical numbers as a secondary option on a new team. The data lets you ask, in advance, whether the trade makes sense.
Identifying overperforming and underperforming teams. The same kind of regression analysis that EPA enables in the NFL works in the NBA via net rating. A team whose record is meaningfully better than their per-100 differential is, on average, going to regress. A team whose record is meaningfully worse than their per-100 numbers is, on average, going to improve. The math, applied carefully, has predicted playoff disappointments and surprises with reasonable accuracy for two decades.
The limits: what pace-and-space metrics cannot tell you
The honest version of this kind of writing names the limits.
Per-100 metrics cannot tell you whether the game is enjoyable to watch. That is not a flaw of the metric; it is a category mistake to ask. The numbers measure efficiency. The game is also drama, conflict, and the unrepeatable moments that no model can predict.
Pace-and-space metrics cannot, on their own, predict postseason performance. A regular-season offensive rating of 120 against a varied opponent schedule does not survive seven games against a coordinated playoff defense that scouts the offense’s third action, double-teams the primary creator, and forces uncomfortable touches to the third or fourth option. Some teams scale up in the playoffs. Some scale down. The regression to playoff-team-quality is real, and it is one of the harder unsolved problems in public analytics.
The metrics cannot evaluate non-statistical contributions reliably. Communication on defense. Gravity from a non-shooter who happens to be a great screener. The slowing effect of a veteran guard on a young team’s turnover rate. These show up in lineup data, sometimes, but not always cleanly. The writers who do this well combine the numbers with the film, and they say out loud when the film disagrees with the data.
A working example: the 2022-23 Sacramento Kings
The 2022-23 Sacramento Kings are the most useful single-team illustration of how pace-and-space metrics rewrote a franchise’s identity. Mike Brown’s first year as head coach produced a top-three offense in the league, a +5.2 net rating, and a third seed in the West that the franchise had not approached in fifteen years. The offensive profile was a textbook pace-and-space build: high tempo, elite three-point shooting around Domantas Sabonis’s playmaking, lineups that stretched the floor with shooters at every position. The team’s offensive rating of 118.6 would have been historic in 2010. In 2023, it was merely excellent. The Kings led the league in points per game and pace, two metrics that twenty years earlier would have been treated as suspicious for a contending team. The math had moved.
What followed in 2023-24 was the equally instructive case study of regression. The same roster, the same coach, posted a much lower net rating in the following season — partly luck, partly defensive scheme adjustment by opposing teams, partly the inevitable variance of three-point shooting. The pace-and-space framework predicted both the rise and the regression. A team built on three-point variance lives by the same variance when it turns against them. The 2024-25 season returned the Kings closer to the middle of the Western Conference, where the underlying numbers had said they belonged from the beginning. The 2022-23 surge was real and partly luck; the 2023-24 retreat was real and partly bad luck. Both pieces of analytical writing existed at the time. The narrative-cycle pieces did not survive contact with the second season.
One additional limit worth naming. The pace-and-space framework is itself a product of a specific era, and the league is already producing the next adjustment to it. Defenses have, in the late 2020s, started exploiting the over-spacing tendency by switching aggressively, packing the paint against weaker shooters, and forcing the offense’s third or fourth option into uncomfortable touches. The data is starting to show a slight retreat from peak three-point volume as defenses learn the math. The metrics will catch up. The frameworks that define modern coverage are not permanent; they are the snapshot of a sport mid-evolution.
Frequently asked questions
What is a good offensive rating for a modern NBA team?
League average offensive rating in 2024-25 was approximately 115. An elite offense posts a season rating of 120 or higher; the 2023-24 Boston Celtics led the league at 122. A bottom-tier offense lives in the 108-112 range. The Super Bowl-style championship pattern holds in basketball too: title winners are almost always top-five in offense, defense, or both.
How do I know if a team is actually fast-paced or just inconsistent?
Pace stabilizes much faster than efficiency. A team’s pace over 15-20 games is usually a reliable indicator of coaching intent. Wild game-to-game pace fluctuations almost always reflect game state — a team playing from behind in the fourth quarter speeds up dramatically. Filter for similar game states if you want a clean read.
Is the mid-range dead?
No, but it has become specialist territory. League-wide mid-range attempts as a share of total shots have fallen by half since 2010. The mid-range survives in the hands of elite scorers — Durant, DeRozan, Booker — who shoot it at percentages high enough to make the math work. For a team to base its offense on mid-range, every shooter needs to be elite. Most teams do not have that personnel, which is why the mid-range diet is, on average, a yellow flag.
How can I learn this myself?
Three resources, in order of accessibility: Basketball-Reference for free historical and current data; Cleaning the Glass for paid, beautifully presented advanced metrics with garbage-time filtering; the NBA’s own Stats site for player tracking data and lineup splits. Read Zach Lowe, Tom Haberstroh, and Ben Falk for the writing that translates the data into prose.
Sources and further reading
- Basketball-Reference’s stat glossary — the canonical definitions for every public NBA metric, and an indispensable reference for analytics writing.
- Cleaning the Glass — Ben Falk’s site, founded on the principle that garbage-time minutes pollute the public data. The single best paid resource for serious NBA analytics.
- Zach Lowe’s writing at The Ringer — the model for how to translate NBA analytics into prose without losing either the math or the texture of the game.
- Dean Oliver’s Basketball on Paper — the 2004 book that formalized possession-based analysis. Still the foundational text.
- NBA.com/stats — the league’s official stats portal, including player tracking data that powers most of the public analytical conversation since 2013.
The Curry miss that opened this article — Game 7, 2016, eleven seconds on the shot clock — was a textbook pace-and-space shot in a possession that the math, in retrospect, did not love. Curry’s shot diet that series was historically aggressive. The Cavaliers won anyway, on a play that came down to LeBron James guessing right at the rim and Kyrie Irving hitting a contested three of his own that was, on the analytics ledger, an even worse shot. Basketball is high-variance. The metrics describe the underlying terrain. The game decides what happens on a given night. For the broader frame on reading NBA numbers without the usual noise, our field guide to NBA advanced stats is the natural next read.
