Bennedict Mathurin came off the Indiana bench in Game 3 of the NBA Finals and dropped 27 points on Oklahoma City in 22 minutes. Indiana won the game 116-107. The bench scoring differential in that single game was 49-18 in Indiana’s favor. The Pacers had built their entire playoff run on this — a deep, two-deep-at-every-position rotation that produced double-digit bench scoring advantages in roughly 70% of their postseason games — and against the team that had won 68 regular-season games, the formula had stretched a series almost all the way to seven. Then Tyrese Haliburton tore his Achilles in Game 7 and the formula stopped working, and the Thunder won 103-91, and Oklahoma City was the 2025 NBA champion.
The Pacers’ run had been almost entirely a bench story. Nine players logged meaningful minutes across the postseason. Three of them — Mathurin, Obi Toppin, and T.J. McConnell — produced playoff impact numbers that, if measured per-100-possessions, would have landed in the top tier of the league. The starters had been good. The bench had been the differentiator, and the differentiator had been good enough to push the league’s best team to the brink in a seven-game series.
What follows is what the Pacers bench actually did across the playoffs, why the depth-over-stars roster construction nearly worked against OKC, and where the model the Pacers used breaks down even when the bench plays at its ceiling.
What the Pacers bench produced, in numbers
Across nineteen playoff games, the Indiana bench averaged 41 points per game. The league postseason average for bench scoring was 28. Indiana’s gap of +13 points per game from the bench was the largest in any single playoff run since the 2003 Detroit Pistons, and the Pistons that year had a different structural reason — Larry Brown’s deep rotation was the prevailing style of the early 2000s. The Pacers’ version of depth in 2025 was anachronistic. Nobody in the modern NBA had been running nine players this deep in a playoff series, and the public analytics community had largely written off this kind of construction as a regular-season optimization that did not travel to the playoffs.
The bench’s net rating across the run was +6.2 per 100 possessions. The starters’ net rating was +2.1. That is unusual — in most playoff runs, the starting unit carries the heavier load and the bench provides connecting minutes. The Pacers had inverted the relationship. The starters were closing games and the bench was winning quarters. Across the first three rounds, Indiana’s largest single-quarter scoring outbursts had come from lineups featuring three or four bench players against opposing starters.
Our garbage-time tax piece covers how playoff lineup data can be distorted by minutes context. The Pacers bench numbers were not garbage-time inflation. The +6.2 net rating came in high-leverage minutes against opposing closing units in the first three rounds. The pattern was real until OKC.
What changed in the Finals
| Series | Pacers bench pts/G | Bench pts diff | Series result |
|---|---|---|---|
| vs Bucks (R1) | 44 | +15 | Won 4-1 |
| vs Cavs (R2) | 42 | +13 | Won 4-2 |
| vs Celtics (ECF) | 43 | +11 | Won 4-2 |
| vs Thunder (Finals) | 33 | +2 | Lost 4-3 |
The Finals number is the part the league’s tactical community has been quietly studying. Indiana’s bench averaged 33 points per game against OKC, down ten from their average across the three prior series. The bench scoring differential collapsed from double digits to +2. The structural advantage that had defined the Pacers’ run did not survive the OKC depth and the OKC defensive intensity. Mark Daigneault’s rotation had been the league’s deepest all season. He matched Indiana’s depth, neutralized the bench scoring gap, and let SGA do the work the Pacers’ best player could not finish.
Haliburton’s Game 7 injury is the headline. The pattern the Pacers were running into was visible before Game 7. The bench scoring advantage had been compressing across every game of the series. The 49-18 Game 3 outlier was actually a return to the prior pattern, not the dominant pattern of the series. By Game 6, the bench differential was zero. By Game 7, with Haliburton out, Indiana had no answer left.
Where this gets weird
The clean “Pacers nearly upset OKC because of depth” reading misses three things.
The first is that the Pacers were not actually playing deep most nights in the Finals. The minutes for their nine-deep rotation compressed against OKC. Indiana’s three least-used bench players saw their minutes drop by half across the series. The “nine deep” framing was true through the conference finals. By Game 4 of the Finals, the rotation was effectively eight-deep, and by Game 6, seven. The compression was a coaching response to OKC’s matchup advantages, not a roster failure. The depth that the Pacers had used to grind opponents down in earlier rounds was not the depth that beat OKC. The eight-man rotation was.
The second is that Indiana’s depth was partly an artifact of the regular season. The Pacers had cycled minutes during the regular season to keep players fresh; OKC had done the same. Both teams arrived at the Finals with cleaner injury profiles than was historically normal for a deep run. Indiana’s regular-season formula traveled to the playoffs because OKC’s regular-season formula was almost identical. In a series against a top-seeded team that had played its starters heavier minutes — Boston, for instance — the Pacers’ depth advantage might have been larger. Against OKC specifically, the advantage compressed because the opponent had been built the same way.
The third is that the public conversation has converged on “Indiana’s depth proves the depth model works” and that conclusion is wrong in the direction the conversation is taking it. Indiana’s depth got them to a Game 7 against a team that was healthier, deeper, and better. The Finals proved that depth without a top-fifteen player ceiling caps your playoff path at the second-best team in the league. The depth model is necessary; it is not sufficient. The Pacers proved the necessity. The Thunder proved the insufficiency.
What other teams should take from this
- Two-deep at every position works in the regular season and the first three rounds. The Pacers proved this at scale. Front offices weighing depth vs star acquisition can lean depth in the regular-season-to-conference-final window without sacrificing playoff path.
- Depth alone does not beat a healthy top-five team. The Thunder closed out the series in seven because they had SGA and depth. The Pacers had depth. The math does not balance.
- Rotation compression in the Finals is structural, not coaching failure. Indiana’s nine-deep rotation became eight-deep and then seven-deep across the Finals because that is what playoff intensity does to bench impact. Front offices should plan for it.
- Injury luck at the top of the rotation is the unrecognized variable. Haliburton’s Game 7 injury swung the championship probability by an enormous margin. The depth model is more fragile than the public conversation has been suggesting.
The callback
That Game 3 in Indianapolis, when Mathurin came off the bench and dropped 27 in 22 minutes and the Pacers bench outscored the Thunder bench 49-18, looked like the night the depth model had finally landed against the league’s best team. By the time Game 7 ended with SGA holding the trophy and Haliburton on the bench with a torn Achilles, the model had landed differently. The Pacers’ run was the cleanest test case in a decade for “can you build a contender around depth alone?” The answer is “yes, until the Finals, against a healthy top-three team that has done the same thing better.” The lineup synergy piece covers the math of why depth produces diminishing returns in the highest-leverage games. The Pacers proved the formula. The Thunder won the title. Both things are true and they explain the gap between very good and championship-good in a way the regular-season standings could not.
Playoff stats via Basketball Reference; lineup data via Cleaning the Glass; bench impact metrics via Dunks and Threes.
