51.05%
1,916 of 3,753 · interval 49.45 to 52.65
The number anyone would quote. Across every bout with exactly one home fighter, the home fighter won 51.05% — on an interval that contains an even split. On its own it establishes nothing at all.
1,206 fighters measured both ways · 1994–2026 · UFC only
HometownAdvantage
Everybody quotes the same number: the home fighter wins 51.05%. A coin flip. Nothing to see. We went looking anyway — 1,206 fighters set against nobody but themselves, and the year the arenas stood empty, no ticket sold and no flag waved. Something survives it. And it is the part of a fight a fighter can actually do something about.
The question
When a fighter is fighting in their own country, do they have a hometown advantage?
The received wisdom
When fighters have to deal with travel, jetlag and foreign surroundings, their performance is likely to suffer.
TLDR; Quick Facts
46.91% — 59.72%
2,539 fights at US events — 1,214 elsewhere
Because it is two opposite effects cancelling. At events in the United States the home fighter loses. Everywhere else he wins by nearly ten points.
83% — 5.7%
share of appearances made at home · USA — Europe
And here is why they cancel. An American fights at home five times in six. A European almost never does. On a US card “home fighter” means most of the roster; abroad it means the local the card was built around.
+8.81 pts
1,206 fighters, 11,551 fighter-fights · [+6.63, +10.99]
So compare a fighter with himself instead. Home record against away record, same man, his quality held fixed. Composition was hiding this effect, not manufacturing it.
−4.93 pts
8.59 falls to 3.66 once the opponent is controlled
But home opponents genuinely are weaker — on all four measures we hold. More than half the apparent edge was handed over by the matchmaker before either fighter signed.
+1.5 to +4
most conservative estimate +1.55 [+0.11, +3.49]
Points that survive every control we can apply, including one that assumes nothing whatever about who is good. A real edge, and a modest one.
+1.09 — +19.94
split decisions — the closed-door period
Two explanations fail. In split and majority decisions, where judges have the most room, there is no signal. And the edge survived a year of empty arenas. Not the scorecards, and not the noise.
In short
The received wisdom is right, and for a reason it did not name. There is a hometown advantage, but it is worth two to four points rather than eight, and the single most quotable number on the subject hides it completely. What is left once the crowd and the judges are ruled out is the unglamorous part — the flight, the sleep, the cut in someone else’s kitchen — plus a matchmaker’s thumb that was on the scale before either fighter signed.
The pooled 51.05% is two effects cancelling
The home fighter loses in America
Take every UFC bout where exactly one fighter was in his own country and the other was not. There are 3,753 of them. The home fighter won 51.05%, on an interval from 49.45 to 52.65 — which contains an even split, and therefore establishes nothing.
That is the figure a preview would quote, and it is the least informative number in this story. Underneath it sit two populations pulling in opposite directions.
Events in the United States
46.91%
home win rate across 2,539 fights [44.97, 48.85] — clear of fifty, and on the wrong side of it
Events everywhere else
59.72%
home win rate across 1,214 fights [56.93, 62.44] — clear of fifty, emphatically
The home fighter loses at American events. Everywhere else he wins by nearly ten points. Pool the two and they very nearly cancel, which is how a real and quite large effect hides inside a number that looks like a coin flip.
The reason is not mysterious, and it is worth seeing rather than reading. It is that the word “home” does not mean the same thing for everybody on the roster.
Share of a fighter’s UFC appearances made at home
United States83.0%
Canada38.1%
Oceania32.7%
UK & Ireland31.3%
Asia20.2%
Brazil19.3%
Latin America11.6%
Europe & other5.7%
at homeaway
An American fights at home five times in six. A European fighter does it once in eighteen. So on an American card “the home fighter” describes most of the people on it — which is to say it describes an average fighter. Abroad it describes the local star the card was built around. Two completely different populations, given the same label, and then averaged together.
8,867 of the sport’s 17,025 fighter-appearances have been made at home — 52% of them. That is what makes the pooled figure sit so close to fifty and be so useless: it is very nearly a comparison of the roster with itself.
Stop comparing different people
The same man, nearly nine points better at home
Comparing home fighters with away fighters compares different people. The fix is to compare a fighter with himself: his record at home against his record abroad, so his own quality is held fixed and only the venue moves.
1,206 fighters have fought both at home and away — 11,551 fighter-fights, 6,669 of them at home. Measured that way the gap is +8.81 percentage points [+6.63, +10.99].
Note which direction that correction ran. The pooled figure was +1.04 points and could not clear zero; the same-fighter figure is nearly nine and clears it comfortably. Composition was hiding this effect, not manufacturing it.
| Group | Fighters | Difference | 95% interval |
|---|---|---|---|
| All | 1,206 | +8.81 | +6.63 to +10.99 |
| Men | 1,079 | +8.59 | +6.57 to +10.90 |
| Women | 125 | +10.91 | +1.44 to +19.89 |
| Fighters from outside the US | 541 | +13.61 | +10.54 to +16.56 |
| American fighters | 665 | +5.06 | +2.01 to +8.20 |
Americans do have a hometown advantage. It is simply that their home is also where almost everybody else fights, so it buys them less than it buys anyone else.
Justin Gaethje
14 June 2026 · UFC Freedom 250 · Washington
Gaethje stopped Ilia Topuria in front of an American crowd. He was the home fighter; Topuria had flown in — and he is from the one group on this page a home fight is worth least to. American fighters gain +3.14 points once the opponent and the era are controlled, against Brazil’s +11.34, because an American’s home is also where almost everybody else fights.
So the most familiar image in the sport — the local man, the American crowd, the finish — is a picture of the smallest edge in the story. The photograph at the top of this page is the other end of it.
One fight is an anecdote either way. And the database knows only countries, never home towns: Washington is Gaethje’s country, not his city.
The matchmaker got there first
Home cards are built softer
Before crediting the time zone or the crowd, there is a duller explanation to rule out: promoters build home cards around local fighters, and they do not book them against the hardest available opponent.
That turns out to be true, and it is measurable. Comparing the same fighter’s home and away bouts, the opponent waiting for him at home is worse on every measure we hold.
| Opponent measure, at home | Difference | 95% interval |
|---|---|---|
| Prior UFC win rate | −4.35 pts | −5.72 to −3.00 |
| Is a UFC debutant | +3.67 pts | +2.11 to +5.22 |
Half a UFC fight less experience, a win rate four points lower, more likely to be debuting, and rated weaker. So a good part of those eight points is not an advantage the fighter earned by being home. It is an advantage the matchmaker handed him.
Pay off every duller explanation in turn
One and a half to four points survive
Each step below adds a control and keeps everything above it. The question is not whether the effect is large — it is how much of it is still standing once every duller explanation has been paid off.
The last row is a check on the machinery rather than a result. We shuffled which country each event was held in, at random, forty times over, and ran the whole thing again. If this method invents home advantages, that row is where one would appear.
Home advantage in percentage points · men’s divisions, controls added in turn
Same fighter, nothing else controlled +8.59
Clears zero
…plus opponent record, career stage, era +5.65
Clears zero
…plus a strength rating for the opponent +3.66
Clears zero
Every fighter’s strength fitted jointly +1.55
Clears zero
Placebo: countries shuffled within year +0.95
No effect
← away betterhome better →
Four estimates and one check. The ladder descends as each duller explanation is paid off, and the last rung — a model that fits every fighter’s strength jointly with the home term, and so needs no assumption about who is good — still sits clear of the line. The placebo, which shuffles event countries at random within each year forty times over, does not. 14,787 fighter-fights, 1,079 men; the joint model, 8,351 fights.
The honest answer is a range rather than a number: somewhere between one and a half and four points, and we quote the floor rather than the flattering end. It is enough to move a close line and not enough to move a mismatch — real, and modest, and nothing like the eight points the raw within-fighter comparison suggested.
Two more ways of asking the same question, both of which could have killed it and neither of which did. Throw away every bout against a UFC debutant, on the grounds that feeding a local a debutant is the crudest form of the matchmaking effect: +3.42 [+1.44, +5.46]. Restrict to the 366 fighters with at least three bouts on each side, so nobody is carried by a single lucky night: +5.92 [+3.02, +8.98].
A note on scope. This ladder is fitted on the men’s divisions, which is where the sample is. The women are reported separately further down, and the short version is that they agree with the men and cannot prove it.
If it were the judges, this row would be the biggest
Split decisions show no home edge
The obvious mechanism is a partisan crowd leaning on three scorecards. If that were it, the advantage would concentrate where judges have the most room to be swayed — in the close fights, the ones that go to a split.
It does the opposite.
Home advantage in percentage points · by how the fight ended
Split & majority decisions +1.09
No effect
Unanimous decisions +4.87
Clears zero
All decisions +4.51
Clears zero
Finishes +3.51
Clears zero
← away betterhome better →
The row where the crowd-and-scorecards story predicts the largest effect is the only row with no effect in it. Split and majority verdicts are also the smallest group here — 1,110 fighter-fights against 5,578 unanimous ones — so the wide bar is partly thin data. But a mechanism that works through judges cannot be absent exactly where judges matter most and present everywhere else.
Home fighters are not more likely to reach the judges either: decisions run +4.51 and finishes +3.51, on intervals that overlap almost entirely. Whatever is happening is happening in the fight rather than on the cards.
9 May 2020 to 23 April 2021, no spectators
The edge survived a year of empty arenas
Between 9 May 2020 and 23 April 2021 the promotion ran without spectators. If the crowd were the mechanism, the effect should have vanished in that window.
It did not.
Closed doors, no spectators
+19.94
points [+2.28, +37.18], on 45 fighters. Larger rather than smaller — but on an interval thirty-five points wide, which is far too loose to take the size seriously.
With crowds, 2019 and 2022 onwards
+5.64
points [+2.03, +8.92], on 647 fighters. The two intervals overlap almost completely. What the window supports is a direction, not a difference.
So the closed-door period does not show that empty arenas helped home fighters. It shows that removing the crowd did not remove the advantage, which is the only claim being made here and the only one 45 fighters can support.
One caveat that belongs with it: in that window the promotion ran in just two countries — 706 fighter-appearances at events in the United States and 277 in the United Arab Emirates. The closed-door result is therefore mostly a statement about Americans fighting at home, and it should be read as one strand of evidence rather than a clean experiment.
With the crowd ruled out and the scorecards ruled out, the remaining candidates are the unglamorous ones: travel, time zones, sleep, cutting weight away from your own kitchen, and a camp that did not have to move. That is also, conveniently, the received wisdom — which turns out to have been right about the mechanism, while the only number anyone quotes pointed the other way.
Which is a large claim to rest on one join between two tables. Before the map of who it helps most, three things this data cannot do.
Three ways this could still be wrong
A Brazilian who lives in Nevada is not away
Nationality is not residence. A Brazilian who lives and trains in Nevada is not “away” in Las Vegas in any sense that matters, and the database has no idea. We probed it: fighters from outside the US who mostly compete abroad show +10.06 [+6.40, +13.73], while those who fight mostly in the US — so probably live there — show +4.88 and do not clear zero [−5.74, +12.95]. The misclassification therefore pushes the estimate down, which makes the headline conservative rather than generous.
The join was broken, and quietly. Event countries and fighter
nationalities come from different sources with different vocabularies. Matching them by
simple equality finds 1,236 home fighter-slots where there are really 8,865 — 14% of
them — because every event says USA while 1,350 fighters say
United States, and 31 UK events face 106 fighters who say England, Scotland,
Wales or Northern Ireland. Any analysis of this question that skips the mapping silently
deletes almost every American and British home fight, and those are the majority of the
record.
What we cannot see. Purses, travel distance, time zones crossed, altitude, short-notice replacements, camp disruption, and whether a fighter slept. The residual two-to-four points is a container for all of them, and this data cannot open it. We can say the advantage is real and roughly how big. We cannot say which of those things it is made of.
It is worth most where it is rarest
Brazil gains eleven points, America three
Held to all of that, the advantage is still there — and it is not evenly distributed. Split by nationality, three groups are individually solid: Brazil at +11.34 [+6.63, +15.82], Canada at +8.18 [+0.06, +15.53] — only barely — and the United States at +3.14 [+0.27, +6.03].
The other five all point the same way and none of them clears zero on its own. Asia has the largest point estimate of the lot at +12.95, and an interval that runs from below zero to twenty-six. That is too few fighters, which is not the same finding as too small an effect.
Asia+12.95 *
Brazil+11.34
Latin America+8.67 *
UK & Ireland+8.39 *
Canada+8.18
Europe & other+6.86 *
Oceania+6.33 *
United States+3.14
Percentage points of win rate gained at home, same fighter compared with himself, opponent record and era controlled. * marks a group whose interval crosses zero — too few fighters to separate it from no effect, which is not the same as no effect.
Read that chart against the earlier one and the pattern is very nearly its inverse: the groups who rarely get to fight at home are the ones it is worth the most to when they do. The United States, which fights at home 83% of the time, gets the least out of it of anybody.
Chartered travel has not removed it
+8.70 points in 2006–12 [+3.22, +13.71], +4.50 in 2013–19 [+1.36, +7.44], and +7.35 in 2020–26 [+3.05, +11.91]. Three decades of the sport professionalising, of chartered travel and full-time camps, has not removed it. The earliest era, 1994–2005, gives the largest point estimate of all at +13.00 — on 42 fighters and an interval that crosses zero, so it is quoted here for completeness rather than as evidence.
125 fighters against 1,079
The women agree with the men, and cannot prove it
The women’s game shows the same thing. Their within-fighter advantage is +10.91 points [+1.44, +19.89] against the men’s +8.59 [+6.57, +10.90] — a larger point estimate on an interval more than four times as wide, resting on 125 fighters against 1,079.
The two are entirely consistent with being the same number, and that is the useful part: whatever the away corner costs, it costs it in two games that share almost none of their other statistics.
One thing we will not do is quote a controlled figure for the women. The ladder further up is fitted on the men’s divisions, deliberately — 125 fighters cannot carry four successive controls and return an interval worth printing. The gap between the two point estimates is a sample-size difference, not evidence of a difference between the sexes, and it should not be read as one.
What this means for your training
The fight starts at the airport
The advantage survived a year of empty arenas, which means the noise was never the thing. What is left is logistics — and logistics is the half of this you actually control.
-
Land early enough to sleep
If the crowd and the scorecards are both ruled out, the residual two to four points lives in travel, time zones and rest. Nothing in this data says how many days are enough. What it does say is that the part of the disadvantage you can move is the part that happens before you get in the cage, so treat arrival as a training variable and not a booking detail.
-
Rehearse the cut somewhere unfamiliar
A camp that did not have to move is one of the few concrete differences between the two corners. Your own scales, your own kitchen, your own sauna and your own food are all part of what makes a home cut routine. If you only ever make weight in your own gym, the first time you do it in a hotel is a live experiment run on fight week.
-
Expect the easier fight at home — and do not bank on it
More than half the apparent home edge is the opponent you were given, not the country you were in: half a UFC fight less experience, a win rate four points lower, more likely to be debuting. A home booking is quietly a softer booking. That is worth knowing before you read too much into a home record, yours or anybody else’s.
And in reverse, for anyone cornering or reading a card: at an event outside the United States the local fighter has a real edge, worth one and a half to four points of win probability once everything else is paid off. At a US card “home fighter” tells you almost nothing, because it describes five appearances in six.