The 2025 NBA playoffs have been exceptionally marked by dramatic comebacks, a challenging feat that the Eastern Conference champions, the Indiana Pacers, appear to have mastered.
The Pacers showcased their remarkable resilience in the first round against the Milwaukee Bucks, overcoming a seven-point deficit in the final 40 seconds of overtime in Game 5 to clinch the series 4-1. They demonstrated a similar incredible ability in the Eastern Conference semifinals, pulling off the same comeback from seven points down within the last 50 seconds of Game 2 against the Cleveland Cavaliers.
These earlier demonstrations of clutch performance were merely preparation for Indiana`s most improbable escape in Game 1 of the Eastern Conference finals. Trailing by 14 points inside the final four minutes of regulation and still down by eight points in the final minute, the Pacers orchestrated a stunning rally. Key three-pointers from Aaron Nesmith, coupled with ill-timed turnovers from the New York Knicks and a high-arcing shot at the buzzer by Tyrese Haliburton, forced overtime, allowing Indiana to ultimately steal the opening game of the series.
Motivated by Indiana`s spectacular late-game recoveries and the New York Knicks winning three games they trailed by at least 20 points – the most by any team in a single playoff run in the play-by-play era (which began in 1998) – this analysis takes a closer look at the prevalence of comebacks in the 2025 playoffs.
Are the Pacers the greatest comeback team in NBA playoff history?
Pinpointing this title precisely is complicated, especially when limiting the scope to the period where comebacks can be accurately quantified. However, based on available data, the short answer is almost certainly yes. Quantifying this claim involves detailed analysis.
Mike Beuoy of Inpredictable.com, a superb resource for NBA statistics, has conducted extensive research on win probability and game comebacks. His website evaluates each game by assigning a “comeback score” based on the winning team`s lowest probability of victory during the contest. Significantly, Indiana`s three highly improbable victories in these playoffs rank among the top seven comebacks recorded in the playoffs since 1997.
Inpredictable also computes an average comeback score for each team`s wins (specifically, a geometric mean, which somewhat reduces the impact of a single extreme comeback compared to a simple average). Despite this adjustment, achieving a high average comeback score is generally easier with fewer overall wins. When visualizing the average comeback score against the total number of wins for each playoff team since 1997, with this year`s teams clearly highlighted, it becomes evident just how exceptional outliers Indiana and New York are.
In collaboration with Beuoy, we explored several methods to develop a single metric that considers both the volume and improbability of a team`s comeback wins. The most satisfying approach we identified was calculating the product of the likelihood of winning each game from its lowest point. This method inherently favors teams with a higher number of playoff victories.
Under this specific measurement, the Pacers` 12 wins (and counting) place them second overall since 1997, surpassed only by the 2011 champion Dallas Mavericks, who achieved 16 wins. The 2011 Mavericks are prominent on the chart, showing the highest comeback score among all championship teams. Meanwhile, the Knicks` 10 wins rank them seventh using this metric, a higher placement than any team prior to this year that did not reach the Finals.
Are win probability models underestimating the chances of comebacks?
Each of Indiana`s three major comebacks occurred in games where win probability estimates gave them a 2.1% chance of winning or less, including an astonishing 0.9% likelihood against New York. While not quite as rare as being struck by lightning multiple times, these events are statistically highly improbable purely by random chance.
Given these outcomes, some skepticism regarding win probability models is understandable. A contributing factor is that these estimates rely on historical data that may not fully keep pace with the evolving dynamics of the NBA, such as the increased speed of play and higher volume of three-point shooting. As comebacks potentially become more frequent due to these trends, models trained on older datasets might, to some degree, underestimate the chances of such rallies.
Another challenge lies in model calibration, particularly at the extreme ends of the probability spectrum. The difference between a 57% and 58% win probability is usually insignificant, but at the low extremes, uncertainty is magnified. For instance, a comeback from a 2% probability is twice as likely as one from 1%. Even small calibration inaccuracies can therefore significantly impact predictions for highly improbable events.
Is there a stat that quantifies the Pacers` offense and defense playing off each other so well?
A potential explanation for the increased number of comebacks this postseason, beyond factors that have been present for years, could be the relationship between offensive and defensive efficiency depending on how a possession begins. Offenses are typically somewhat more effective following a defensive rebound, as it allows for more transition opportunities. However, the specific benefit derived from a stop can vary significantly by team and season.
Data from Inpredictable.com for the current playoffs indicates a substantial difference in efficiency based on the possession`s origin. Teams are averaging 1.17 points per possession after a defensive rebound, compared to 1.07 after a made shot or a dead ball turnover. This 0.10 point difference is significantly larger than in the past couple of playoffs, where the gap was much smaller (just 0.01 points per possession better after defensive rebounds in 2022 and 2023).
One possible reason for this change might be the increased physicality allowed by referees in the playoffs over the last two seasons. Physicality tends to have a greater impact on half-court situations than on transition plays. For comparison, during the 2023 playoffs, when officiating was tighter regarding contact, teams were more efficient after made shots or dead ball turnovers (averaging 1.1 points per possession).
This shift might also tie into the heightened importance of avoiding turnovers this year. As tracked by Owen Phillips of the F5 Newsletter, the team committing fewer turnovers has an impressive 53-20 record (.726) in these playoffs, potentially representing the highest winning percentage for such teams on record for a single postseason. Last year, teams with fewer turnovers won only about 60% of the time, which is closer to the average over the past decade (62%).
Explaining why teams are scoring so efficiently off defensive rebounds this year is more challenging, though player fatigue among starters on several teams reaching the conference semifinals, who have logged heavy minutes, could be a factor.
Focusing specifically on Indiana, the Pacers derive a greater benefit from defensive rebounds than most teams. They average 1.26 points per possession after securing them, ranking third best in the NBA. While they are still third best after made shots or dead balls, their efficiency in that scenario drops by an above-average 0.16 points per possession, illustrating a significant split.
How does this relate to comebacks? A larger gap in efficiency based on the possession start type makes a team (or the league overall) more prone to streaks. Getting a stop is highly rewarding, not just by preventing opponent points but by fueling a more efficient offensive opportunity. Conversely, allowing a score leads to a less efficient offensive setup. This amplified swing creates either a virtuous cycle (stops leading to efficient scores) or a vicious one (scores leading to less efficient possessions), depending on performance.
Increased streakiness resulting from these efficiency splits makes it more likely for teams to build large leads rapidly and, consequently, for trailing teams to mount equally rapid rallies from those deficits. This phenomenon provides the fertile ground for the numerous dramatic comebacks witnessed in these playoffs, including those by the Indiana Pacers.
In contrast, the Oklahoma City Thunder, despite having some memorable games, exhibit a different pattern. While their defense is very effective after opponent scores or dead ball turnovers (ranking second), their defense is exceptionally strong after opponent defensive rebounds (meaning when OKC is transitioning back to defense after the opponent gets a stop). In this scenario, they allow significantly fewer points per possession than any other team, suggesting their defensive performance is less subject to the kind of negative swings that might contribute to large scoring runs against them.
