The 2025 NBA playoffs are marked by a comeback, and no one masters the arts as the Eastern Conference champion Indiana Pacers.

The Pacers emerged from the seven-point deficit in the last 40 seconds of overtime to complete a 4-1 series victory in Game 5 of the first round before achieving the same feat in the final 50 seconds of Game 2 defeating the Cleveland Cavaliers in the Eastern Conference semifinals.

All of these are appetizers for Indiana’s most unlikely Houdini performance in Game 1 of the Eastern Conference Finals. The Pacers dropped 14 points in the four-minute regulatory mark, eight at the last minute, and forced overtime to work with a bunch of Aaron Nesmith’s three-pointers, an inappropriate New York Knicks turnovers and Tyrese Haliburton’s treble shots on the buzzer and eventually stole the first game of the series.

Inspired by the late Indiana rally and New York winning three games, winning at least 20 points, a team in a playoff game-by-game (starting from the 1998 playoffs), let’s take a closer look at how the comeback dominated the 2025 playoffs.

Are the Pacers the greatest comeback team in NBA playoff history?

The short answer limits it to periods where we can actually quantify comebacks, almost certainly certainly. Quantifying the long answer to this title is complicated.

No one has more chances of winning and comeback work than Mike Beuoy of Inpredictable.com (an amazing NBA resource). The site is rated on a comeback score based on the probability of a winning team's low, while Indiana's three-shot win is the victory of seven best comebacks in the playoffs since 1997.

Incredibly, it is also focused on the team's average comeback scores per game. (Technically, this is a geometric mean, which is greater than the weight taken by the traditional mean than the weight recovered.) Despite adjustments, it has been easier to establish a high comeback score rather than a few comeback scores so far. When we take average comeback scores per playoff team since 1997 – this year’s team highlighted in primary colors – it’s clear what the outliers are for Indiana and New York.

Working with Beuoy, we tried several different ways to find a single comeback method that both illustrates the amount of comeback and its impossible performance. The most satisfying thing we found was to place the probability of each victory at the lowest point - that is, the chances of a team winning all the games in the playoffs.

It's all this in this way, whether through a comeback or not, because no match is a 100% chance of winning the jump. Still, the Pacers’ 12 wins (and counts) have ranked second in the group since 1997, trailing only 2011 champion Dallas Mavericks, who won 16 wins. You can see on the rankings Dallas is the comeback score of any champion winner.

Meanwhile, the Knicks ranked seventh in 10 wins, higher than any team before this year.

Does the victory probability model underestimate the chances of a comeback?

All three Indiana comebacks are winning the probability of winning, with an estimated 2.1% chance of chance occurring or less, including 0.9% chance of New York happening 2.1%. The lightning was not totally shocked (based on the 80-year lifespan, the National Weather Service estimated the chance of happening at one point was .000065%), but it was absolutely impossible by random chance alone.

Based on this, you can forgive some doubts about the probability of victory. Part of the challenge is that these estimates are based on historical data that may not always keep up with the rapidly evolving NBA. For example, ESPN's model was built in 2017 based on training data from the first seven or so seasons. As a comeback is becoming more common, I wrote this trend in 2019 with ESPN’s Baxter Holmes, due to faster race speeds and 3s increase in volume, so it’s possible that we’re underestimating a certain level of opportunity.

Another problem is calibration. All models have uncertainty, but in most practical cases, the difference between the 57% and 58% chance of winning is not related. In extreme cases, uncertainty is amplified, as the 98% chance of a victory is twice as likely to make a comeback. From 99% of 99% to make a comeback (one in 100 odds) is 10 times more likely than 99.9% (one in 1,000). Therefore, even small calibration problems are important.

Is there a statistic that quantifies the Pacers' offense and defense that are so good?

There may be an explanation for why this playoffs made a particularly comeback when most of the factors cited have been around for years: the relationship between offense and defense. Crime is often more effective after a stop because it can provide more opportunities for early offensive and defensive cross-match, but the benefits of stopping (and vice versa) may depend on various factors that change by team, and into the season.

Often, these playoffs vary greatly in efficiency, depending on whether the offense starts on the defensive rebound or pulls the ball out of the net after the rim. Back at Inpredictable.com, their data showed that the average defensive rebound was 1.17 points per game after a shooting or dead ball turnaround was 1.07. (Average steals or real-time ball turnovers, 1.23 points per game is much higher.) This is a change compared to past few playoffs, when the differences in the start type were much smaller - 0.01 points per game in 2022 and 2023.

As for why the change has occurred, I point out that in the past two years, referees have allowed referees to increase their bodies. Inevitably, physically more of a problem in half-time than in transition. In the 2023 playoffs, when the whistle is tighter, the team averages 1.1 points per game after shooting or dead ball turnaround.

I think it might have something to do with the reason why avoiding turnovers is crucial in this year's playoffs. As Owen Phillips of F5 Newsletter tracked, teams with fewer turnovers went 53-20 (.726), which would be the highest winning rate for such teams. Last year, teams with fewer turnovers won only 60% of the time, about the average of the past decade (62%). The winner of the turnover was better than the .500 in 2018-19 (41-37).

It's hard to explain why teams have scored so effectively on defensive rebounds this year, although fatigue may be a factor in the starters of several teams that have entered the semifinals of the conference, recording a lot of time.

The Pacers specifically shifted our focus to Indiana, which gained more benefits than most teams with defensive rebounds. Their average average of 1.26 points per game, which is unpredictable in the NBA, third place. Although Indiana remains third after shooting or dead balls, their efficiency dropped to an average of 16 points higher.

On the other side of the court, we see a similar big split. The Pacers' defense is ranked tenth after a shot or dead ball turnover, worsened in each game after defensive rebounding, falling to 14th place.

Now, what does this have to do with a comeback? The larger the gap between parking and scores on the other end of the court, the difference in teams (or leagues) is likely because the size of each property is enlarged. Stop is not just stopping opponents from scoring, but also juicing the team's offense and vice versa - according to your point of view, it's a virtuous cycle or viciousness.

The more games the more likely the team is to build a big lead and the more likely the opponents are to rallied from them. Added up, you've got Indiana comeback recipes.

On the other hand, despite losing the fourth quarter lead in their first game of the series and losing the game against the Denver Nuggets and making a comeback in half against the Memphis Grizzlies, the Oklahoma City Thunder did not rely on their offense to defend. Oklahoma City has gone through a transition from a shot or dead ball (second in every game after the Detroit Pistons), but of all teams that start the defensive rebound, each game is 0.08 points less than any other team.