Does Home Court Actually Matter in the NBA Playoffs?

With the NBA playoffs right around the corner, the question always comes up: just how important is home court advantage?

The NBA playoffs are set up as follows:

• 16 of 30 teams make it
• Teams are split into 2 conferences, East and West, and ranked by regular season record from 1–8
• Seeding decides who you play: 1 seed vs. 8 seed, 2 vs. 7, and so on
• The higher seed gets home court advantage in a best-of-7 series, in this format:
  • Game 1: Higher Seed
  • Game 2: Higher Seed
  • Game 3: Lower Seed
  • Game 4: Lower Seed
  • Game 5: Higher Seed (if needed)
  • Game 6: Lower Seed (if needed)
  • Game 7: Higher Seed (if needed)
• First team to win 4 games moves on
• It's a bracket where the winner advances until there's 1 champion

I pulled game result data from every NBA playoff series between 2000 and 2024 using the NBA Stats API's leaguegamelog endpoint, an undocumented internal API reverse-engineered and catalogued by the community swar/nba_api project. That gives us 2,062 games across 25 seasons. All numbers in this article come from that dataset unless otherwise noted.

The Baseline: Win Rates by Location

The first thing to check is the simplest: do home teams just win more?

home_win_rates = (
    df[df["IS_HOME"]]
    .groupby("ROUND")["WIN"]
    .mean()
)

Home court advantage is real, and doesn't fade as dramatically as the rounds progress. Even in the Finals, the home team wins 60.7% of games. The gap in the Conference Finals (60.4%) and Finals (60.7%) is nearly identical, suggesting something deeper.

Does Seeding Drive It More Than Home Court?

There's an obvious confounding variable in that first table: the home team in any playoff series is generally the better team. So when we see home teams winning 63.7% of games, are we actually measuring home court or just that the home team is better?

To separate the two, I looked specifically at games 3 and 4 of first-round series. This is the moment in the bracket where the lower seed finally plays at home. Take a 1 vs. 8 series: after two games at the 1-seed's arena, the series shifts to the 8-seed's building for games 3 and 4. The 8-seed is still the worse team, but now they have the crowd.

There's also a second, rarer scenario worth isolating: series where the lower seed steals home court by winning games early. If the 8-seed wins game 1, they lead 1-0. With 6 games left (3 at home, 3 away), winning all their home games is enough to advance. At that point, the lower seed effectively "owns" home court for the rest of the series despite being the underdog.

Filtering for both cases gives us a clean look at what home court does when it's decoupled from team quality:

# Games where the home team is the lower seed (higher seed number = worse team)
upset_games = df[
    (df["home_team_seed"] > df["away_team_seed"]) &
    (df["round"] == "First Round")
]
 
home_win_rate = upset_games["home_win"].mean()
print(f"Home win % when lower seed hosts: {home_win_rate:.1%}")
# Output: Home win % when lower seed hosts: 44.1%

The result: 44.1%. The lower seed wins less than half the time even when playing at home. Home court gives them a real boost (remember, home teams win 63.7% overall), but it's not enough to overcome the talent gap. A significant chunk of what we call "home court advantage" is really just team quality in disguise.

Shooting: The Most Direct Signal

If home court has a real effect independent of team quality, it should show up in shooting, and it does. Here's where each zone sits on the court:

Zone-level shooting data comes from the leaguedashteamshotlocations endpoint on the same NBA Stats API, which has no official documentation, but is catalogued by the community-maintained swar/nba_api Python package. Data fetched for each of the 25 playoff seasons with home and road splits:

records = []
for season in seasons:
    for location in ["Home", "Road"]:
        resp = requests.get(
            "https://stats.nba.com/stats/leaguedashteamshotlocations",
            headers=HEADERS,
            params={
                "Season": season, "SeasonType": "Playoffs",
                "PerMode": "Totals", "MeasureType": "Base",
                "Location": location, "DistanceRange": "By Zone",
            },
        )
        rows = resp.json()["resultSets"]["rowSet"]
        for row in rows:
            for i, zone in enumerate(ZONES):
                records.append({"location": location, "zone": zone,
                                 "FGM": row[2 + i*3], "FGA": row[3 + i*3]})
 
agg = pd.DataFrame(records).groupby(["location", "zone"])[["FGM","FGA"]].sum()
agg["FG_PCT"] = agg["FGM"] / agg["FGA"]

Paint shots show the largest home-away gap (+1.5%) not corner threes, as conventional wisdom might suggest. The smallest gap is free throws (+0.3%), which makes sense: a free throw is the most crowd-isolated shot in basketball. Corner threes and above-the-break threes are nearly identical in their home advantage, suggesting three-point shooting as a category isn't especially crowd-sensitive.

Foul Differential: The Quiet Edge

Shot quality only tells part of the story. Foul calls are where home court shows up most consistently, and most controversially.

foul_summary = (
    df.groupby("IS_HOME")[["PF", "FTA"]]
    .mean()
    .round(2)
    .rename(index={True: "Home", False: "Away"})
)

Home teams get +1.84 more free throw attempts per game than away teams, a direct result of away teams being called for more fouls. Over a 7-game series, that compounds into roughly 13 extra free throw attempts for the home team. In playoff games decided by 2–3 points, that's the series.

Whether this reflects referee bias is genuinely contested. A 2022 study in the Journal of Sports Economics analyzing 30,695 foul plays from 1,679 NBA games (2017–2021) found that crowd support did not significantly cause referees to treat home and away teams differently in crucial late-game situations, contradicting earlier research. The foul differential is real in the aggregate; what drives it remains an open question.

But with average margin of victory over 11 points per game, this shouldn't be heavily considered.

avg_mov = df[df["IS_HOME"]]["PLUS_MINUS"].abs().mean()
# → 11.5 points

The Altitude and Arena Effect

Not all home courts are equal. To measure which arenas actually produce the largest home advantage, I calculated each team's home win rate minus their away win rate across all 2,062 playoff games, a direct measure of how much better each team performs at home versus on the road.

home = df[df["IS_HOME"]].groupby("TEAM_NAME").agg(
    home_games=("WIN", "count"), home_wins=("WIN", "sum")
).reset_index()
 
away = df[~df["IS_HOME"]].groupby("TEAM_NAME").agg(
    away_games=("WIN", "count"), away_wins=("WIN", "sum")
).reset_index()
 
stats = home.merge(away, on="TEAM_NAME")
stats["home_pct"] = stats["home_wins"] / stats["home_games"]
stats["away_pct"] = stats["away_wins"] / stats["away_games"]
stats["hca_delta"] = stats["home_pct"] - stats["away_pct"]

Top 5 by home-away win rate delta (minimum 30 home playoff games, 2000–2024):

Denver comes in 10th at +27.8%, a real edge but not the dominant outlier the altitude narrative might suggest.

On Denver's altitude: The physiological effect is real and documented. ESPN tracked visiting player movement data during the 2023 Finals and found the average speed of players on visiting teams dropped from 4.20 mph in the first quarter to 3.89 mph in the fourth, and the share of time spent walking or standing still rose from 69.1% to 73%. A 2017 academic study estimated Denver could expect to win 66.1% of home games compared to a typical home-court baseline of 62%. The fatigue is physiological, not psychological. It compounds through the game rather than shrinking.

On crowd noise: Systematic decibel measurements across NBA arenas don't exist. Most published figures come from team press releases or journalists with handheld meters. The one verified data point: on November 15, 2013, Kings fans at Sleep Train Arena (now Golden 1 Center) hit 126 dB to set a Guinness World Record for loudest indoor crowd roar, verified by an on-site Guinness official. What the actual game data shows is which arenas produce results, which puts Houston, Utah, and OKC ahead of the arenas most often cited in the crowd noise conversation.

Series Length: Does Home Court Hold Up?

A common assumption is that home court matters most in short series. If a team can survive long enough, the theory goes, the advantage evens out. The real data is more nuanced.

# For each series: did the team with home court advantage (higher seed) win?
series["HCA_WON"] = series["HCA_WINS"] == 4
 
by_length = series.groupby("TOTAL_GAMES")["HCA_WON"].mean()

A few things stand out. First, sweeps are not all higher seeds: 24% of sweeps are pulled off by the lower seed. Second, 5-game series have the highest HCA win rate (78.1%), not the lowest. A team that wins in 5 typically dominated, and dominant teams usually have home court. Third, even in 7-game series, the HCA team still wins 65.3% of the time, far from a coin flip.

The narrative that "long series neutralize home court" isn't supported by the numbers. What the data actually shows is that the HCA team wins about 70% of series regardless of length.

What the Data Actually Says

After 2,062 real games, the picture is clearer than the conventional wisdom:

• Home court is real at every round and doesn't fade the way people assume
• The home team wins 60–64% of individual games from first round through the Finals
• The HCA team wins 69% of series overall, and that holds across all series lengths
• The edge shows up in shooting (+1.3% FG%, +0.7% 3P%), foul differential (+1.84 FTAs per game), and physiological fatigue at altitude (Denver visitors average 3.89 mph in Q4, down from 4.20 in Q1 per ESPN)
• What looks like home court in the raw numbers is partly team quality in disguise: when worse teams host, they win just 44% of the time