Home-Court Advantage in Tennis: Re-measuring an Assumption from the 2026 Australian Open

On February 12, 2026, the state of Victoria entered a five-day lockdown. At Melbourne Park, the Australian Open — which had begun on February 8 — continued, but the stands were closed to fans. Rod Laver Arena stood empty for the days that followed. Same court, same wind, same field of players; only the crowd variable was switched off mid-tournament. For a sports data analyst, it was close to a natural experiment: a Grand Slam running on without home-court advantage in the usual sense.

I reopened my spreadsheet that same night. For years I had followed tennis with a question hanging over it: where does home-court advantage actually come from, and how large is it? The pandemic answered part of it, but the answer was not as pretty as many would like.

In football, home advantage is well quantified, typically around 0.3 to 0.5 goals per match. Tennis works differently. It is an individual sport: fans cannot touch the ball, cannot run for the player, and since electronic line-calling replaced most line judgments, one important channel of influence has narrowed. Tennis studies usually estimate home advantage at only a few percentage points of win probability, far smaller than in team sports.

When the crowds disappeared for five days, I split the matches into two groups. The crowd group covered the early rounds before February 12 and the late rounds after February 17. The no-crowd group was the lockdown window. I tracked four metrics: first-serve percentage, points won on first and second serve, double faults, and break-point conversion.

The first thing I noticed: the gap between the two groups was smaller than expected. First-serve percentage barely moved. Double faults even dipped slightly during the no-crowd window, though the sample is far too small to conclude anything. Second-serve points won — the metric most sensitive to psychological pressure — shifted only within the margin of error. If the roar of the crowd were the main technical driver, we should have seen a clear drop. We did not.

What stands out is not the serve numbers, but how we interpret results when the crowd is absent. Ash Barty, the top seed and the host nation's biggest hope, walked into her quarterfinal against Karolina Muchova during the no-crowd window. She took the first set 6-1, then lost the next two. I would not attribute that result to the absence of fans; one match is not enough to conclude. But it forced me to separate three mechanisms usually lumped together under home-court advantage.

The first mechanism is familiarity: a home player trains on that surface year-round, knows the bounce, the wind, the rhythm of daily life. The second is the crowd: feeding energy to the home player, pressing the opponent. The third is logistics: sleeping at home, no jet lag, no adapting to unfamiliar food. During the five no-crowd days, the second mechanism was removed entirely, while the first and third remained intact.

That is why I was not surprised that the technical metrics barely moved. A home court is not just geography, until it disappears. Most of the advantage, if any, lies in familiarity and logistics — things a lockdown does not take away. The applause, it turns out, was only the surface layer.

For a cross-check, I looked at the 2026 US Open, played in full from August 31 to September 13, 2026, without fans. There, too, no wave of home players winning abnormally appeared. If anything, some players known for rising to the occasion in front of a home crowd performed more quietly. The sample is still small, but two independent tournaments pointed in the same direction.

Based on my experience watching matches across many seasons, I have seen prediction models treat home advantage as an added constant. The mistake is that this variable is not fixed. It depends on the tournament, the player, where the opponent comes from, and how much they have already played. Misreading one variable is like losing your bearings for an entire year.

Here I have to be most cautious. Five days is too small a sample. The matches in the lockdown window were mostly third and fourth rounds, where opponent quality is more even and the skill gap is compressed. Any conclusion drawn from that must carry a very wide confidence interval. Worse, even if a home player wins more at home, we are not allowed to read that as causation. They may simply get a friendlier schedule, more rest, or an opponent coming off a marathon. Correlation is not causation, and in tennis the selection effect is strong enough to produce a home-court advantage that looks real.

When the lockdown lifted and fans returned in limited numbers, the tournament closed with two champions who were not home players: Novak Djokovic in the men's draw and Naomi Osaka in the women's. That says little about home advantage, but it reminded me that a Grand Slam holds too many variables for a single season to tell the whole story.

Some analysts even argue that home advantage in tennis is overstated and that it is mostly a difference in scheduling and familiarity with the court. The 2026 experiment does not fully refute that view, but it does not confirm it either. It only shows that if home advantage exists, it is not where the media usually points.

Before trusting a number, ask where it comes from. A pretty home-win rate can come from an easy schedule, a familiar court, or simply luck in a small sample. Numbers whisper. Those willing to listen hear an entire match — and the places where the data stays silent.

When the next major season begins, I will not add a ready-made home-advantage constant to my model. Instead, I will track three signals: rest days between matches for home players versus opponents, second-serve points won in deciding games, and break-point conversion in the third set. If home advantage truly exists, it will leave a trace there — not in the applause.