For most of football's history, home advantage has been one of its most dependable patterns. When the Football League began in 1888, home teams won well over half of top-flight matches. In the 2020/21 Premier League season, played almost entirely without supporters, away teams won more matches than home teams for the only time in the history of the English top flight.
Two separate stories sit behind those figures. The first is a slow decline over more than a century, which no single cause explains. The second is a sudden test case. When Covid-19 emptied stadiums across Europe, researchers could compare matches with and without crowds on a scale that had never been possible before. Together they answer a question that matters to anyone who studies football, bets on it, or wonders why a team with a strong record can struggle at its own ground: what does home advantage actually consist of, and why is there less of it than there used to be?
This analysis uses match results for every English top-flight league game since 1888, drawn from an open historical results database compiled by space-weather scientist Sean Elvidge. It adds the main peer-reviewed research on matches played behind closed doors. The figures cover every completed season up to and including 2025/26.
How to measure home advantage
Home advantage can be measured in several ways. They do not always move together, so the choice of measure affects the conclusions.
Home win percentage is the share of matches won by the home team. It is easy to understand, but it ignores the difference between a draw and an away win.
Home share of points is the percentage of all league points won by home teams. A value of 50% means no home advantage. Researchers in the tradition of the sports statistician Richard Pollard have used this measure for decades because it is comparable across leagues of different sizes. The figures in this article apply three points for a win to every season, including those played under the old two-point system, so that all eras are measured on the same basis.
Home goal difference per match is the average number of goals by which home teams outscore away teams. It is less sensitive to random variation than win percentage, because each goal counts rather than only the final result.
Throughout, it helps to remember that one season of 380 matches is a small sample for this purpose. With a true home win rate of about 45%, the statistical uncertainty in a single season is around 2.5 percentage points. Year-on-year changes of four or five points can be random variation alone. Long-term trends and multi-season averages are more informative than any one season.
A long decline in the English top flight
Grouping the top flight by decade shows the shape of the long-term trend.
html
<table>
<thead>
<tr>
<th>Decade (English top flight)</th>
<th>Matches</th>
<th>Home wins</th>
<th>Draws</th>
<th>Away wins</th>
<th>Home goal difference per match</th>
<th>Home share of points (3 points for a win)</th>
</tr>
</thead>
<tbody>
<tr>
<td>1890s</td>
<td>2,366</td>
<td>58.5%</td>
<td>19.3%</td>
<td>22.2%</td>
<td>+1.00</td>
<td>69.4%</td>
</tr>
<tr>
<td>1920s</td>
<td>4,620</td>
<td>54.0%</td>
<td>24.2%</td>
<td>21.8%</td>
<td>+0.82</td>
<td>67.5%</td>
</tr>
<tr>
<td>1950s</td>
<td>4,620</td>
<td>51.4%</td>
<td>22.5%</td>
<td>26.1%</td>
<td>+0.66</td>
<td>63.7%</td>
</tr>
<tr>
<td>1970s</td>
<td>4,620</td>
<td>48.9%</td>
<td>29.0%</td>
<td>22.1%</td>
<td>+0.57</td>
<td>64.9%</td>
</tr>
<tr>
<td>1990s</td>
<td>4,128</td>
<td>45.9%</td>
<td>28.2%</td>
<td>25.9%</td>
<td>+0.43</td>
<td>61.1%</td>
</tr>
<tr>
<td>2000s</td>
<td>3,800</td>
<td>47.2%</td>
<td>25.7%</td>
<td>27.1%</td>
<td>+0.42</td>
<td>61.0%</td>
</tr>
<tr>
<td>2010s</td>
<td>3,800</td>
<td>45.7%</td>
<td>24.6%</td>
<td>29.7%</td>
<td>+0.36</td>
<td>58.7%</td>
</tr>
<tr>
<td>2020s (2020/21 to 2025/26)</td>
<td>2,280</td>
<td>43.0%</td>
<td>23.6%</td>
<td>33.4%</td>
<td>+0.22</td>
<td>55.1%</td>
</tr>
</tbody>
</table>The trend is clear on every measure. In the 1890s home teams won roughly one more goal per match than their opponents and took almost 70% of available points. In the 2020s the home goal margin is less than a quarter of a goal and the home share of points is about 55%. The advantage has not disappeared, but by these measures it is around a third of its Victorian size.
The figure for the 2020s includes 2020/21, the season played behind closed doors, and that one season pulls the decade's average down. Taking the four full seasons after it, from 2021/22 to 2024/25, plus 2025/26, the picture is less extreme but still lower than in earlier periods.
html
<table>
<thead>
<tr>
<th>Premier League season</th>
<th>Home wins</th>
<th>Draws</th>
<th>Away wins</th>
<th>Home goal difference per match</th>
<th>Home share of points</th>
</tr>
</thead>
<tbody>
<tr>
<td>2015/16</td>
<td>41.3%</td>
<td>28.2%</td>
<td>30.5%</td>
<td>+0.28</td>
<td>56.0%</td>
</tr>
<tr>
<td>2016/17</td>
<td>49.2%</td>
<td>22.1%</td>
<td>28.7%</td>
<td>+0.39</td>
<td>61.1%</td>
</tr>
<tr>
<td>2017/18</td>
<td>45.5%</td>
<td>26.1%</td>
<td>28.4%</td>
<td>+0.38</td>
<td>59.4%</td>
</tr>
<tr>
<td>2018/19</td>
<td>47.6%</td>
<td>18.7%</td>
<td>33.7%</td>
<td>+0.32</td>
<td>57.4%</td>
</tr>
<tr>
<td>2019/20</td>
<td>45.3%</td>
<td>24.2%</td>
<td>30.5%</td>
<td>+0.31</td>
<td>58.0%</td>
</tr>
<tr>
<td>2020/21</td>
<td>37.9%</td>
<td>21.8%</td>
<td>40.3%</td>
<td>+0.01</td>
<td>48.7%</td>
</tr>
<tr>
<td>2021/22</td>
<td>42.9%</td>
<td>23.2%</td>
<td>33.9%</td>
<td>+0.21</td>
<td>54.8%</td>
</tr>
<tr>
<td>2022/23</td>
<td>48.4%</td>
<td>22.9%</td>
<td>28.7%</td>
<td>+0.42</td>
<td>60.7%</td>
</tr>
<tr>
<td>2023/24</td>
<td>46.1%</td>
<td>21.6%</td>
<td>32.4%</td>
<td>+0.32</td>
<td>57.4%</td>
</tr>
<tr>
<td>2024/25</td>
<td>40.8%</td>
<td>24.5%</td>
<td>34.7%</td>
<td>+0.09</td>
<td>53.3%</td>
</tr>
<tr>
<td>2025/26</td>
<td>42.6%</td>
<td>27.4%</td>
<td>30.0%</td>
<td>+0.30</td>
<td>56.9%</td>
</tr>
</tbody>
</table>Grouping seasons into five-year blocks smooths out the noise. From 1992/93 to 2011/12, every five-season block of the Premier League produced a home win rate of 44.7% to 47.4% and a home goal margin of about +0.41 to +0.44 per match. The blocks covering 2012/13 to 2016/17 and 2015/16 to 2019/20 fell to a margin of about +0.35. The five seasons from 2021/22 to 2025/26, all played with crowds, produced a home win rate of 44.2%, an away win rate of 31.9% and a home margin of +0.27 goals per match.
The 2024/25 season is a notable outlier. Home teams outscored visitors by only 0.09 goals per match, the second-smallest margin of the Premier League era after the empty-stadium season. In that context, the low home win rates widely reported during 2025/26 are a continuation of the trend rather than something new.
What empty stadiums showed
The long decline has many possible causes and none can be isolated cleanly. That is why the Covid-19 period mattered so much to researchers. For the first time, the most obvious ingredient of home advantage, the crowd, was removed while most other factors stayed the same. Teams still played at their own grounds, slept in their own beds and avoided travel when at home.
The largest European study
The most widely cited analysis is a study by Dane McCarrick, Merim Bilalic, Nick Neave and Sandy Wolfson, published in Psychology of Sport and Exercise in 2021. It covered 4,844 matches from 15 leagues in 11 countries and compared games played with crowds earlier in 2019/20 with games played without them after the restart. Home advantage in both goals and points fell significantly. Home teams also created significantly fewer attacking chances without supporters. The authors observed that the reduction came mainly from home teams performing worse, rather than away teams performing better.
Their findings on refereeing are just as important. Away teams still received more fouls and yellow cards than home teams, but the gap was smaller. Once the analysis accounted for which team was dominating the match, the apparent referee bias was largely diluted. No effect was found for red cards.
Germany as a case study
A study of the Bundesliga by Alexander Dilger and Lars Vischer, published in the Athens Journal of Sports, used a neat design. It compared 83 matches played behind closed doors after the 2019/20 restart with the 83 matches between the same pairs of teams played earlier in the season with fans. The home win rate fell from 48.2% to 32.5%, and away wins rose from 32.5% to 44.6%. The fall in home wins was statistically significant; the rise in away wins and draws was not. The authors also found that without crowds, referees gave away teams significantly fewer yellow and red cards and less additional time.
The yellow card effect
A study by Alex Bryson, Peter Dolton, J. James Reade, Dominik Schreyer and Carl Singleton found a large and statistically significant change in yellow cards in matches without spectators, with away teams booked less often. Of all the effects measured in research on empty stadiums, the change in disciplinary decisions is among the most consistent.
Not every league moved the same way
A statistical analysis by Luke Benz and Michael Lopez, covering 17 professional leagues, reached a more cautious conclusion. They showed that common methods tended to overstate how precisely home advantage could be measured over a single season. With a more suitable model, a bivariate Poisson regression, the results were mixed: some leagues showed a clear decline, while others showed none, or even a possible increase. For readers, the takeaway is that home advantage is not a fixed property of football. It varies between competitions, and the evidence from one league cannot simply be transferred to another.
The Premier League puzzle of summer 2020
The English data contains a detail that most summaries of the Covid-19 period leave out. The Premier League resumed behind closed doors on 17 June 2020, with 92 matches still to play in 2019/20. In those 92 matches, home teams won 46.7% of games, with a home margin of +0.37 goals. That is slightly higher than in the 288 matches played with crowds earlier in the same season, where home teams won 44.8% with a margin of +0.29.
The full season that followed told a different story. In 2020/21 home teams won only 37.9% of matches, away teams won 40.3%, and the home margin was +0.01 goals per match, almost zero.
There are several reasonable explanations, and none can be proven from the results alone.
Sample size. At 92 matches, the statistical uncertainty in the home win rate is around five percentage points. The difference between summer 2020 and the earlier part of the season is well within that range.
Context. Several of the matches after the restart had little at stake, because Liverpool had all but secured the title and some teams were safe mid-table. A sample like that does not represent normal conditions.
Adaptation. The 2020/21 season was played under empty-stadium conditions for its full length, with only brief periods of limited attendance. Any effect from crowd absence, on players, referees or tactical approaches, had a full season to develop. A few weeks of football in summer 2020 may not have been enough to show it.
The lesson for any analysis of home advantage is that a striking short-term figure, in either direction, is rarely conclusive. The 2020/21 season is a better test than the summer of 2020, and the European studies with thousands of matches are better still.
What makes up home advantage
Research before and after Covid-19 usually breaks home advantage down into several components. Their relative weight is still debated.
The crowd and the referee
The pandemic evidence points strongly to referees as part of the mechanism. When crowds were absent, disciplinary decisions became more even across studies. The likely explanation is psychological rather than deliberate: referees under pressure from tens of thousands of people may give marginally more of the close decisions to the home side.
The size of this effect has limits. In the McCarrick study, much of the referee bias disappeared once match dominance was taken into account. Home teams commit fewer fouls partly because they have the ball more and attack more. Some of what looks like bias is a consequence of how matches are played.
VAR as a natural experiment
If refereeing decisions explain a large part of home advantage, the introduction of video assistant referees should have reduced it. The evidence so far does not support that. A study by Camilo Abbate, Jeffrey Cross and Richard Uhrig in the Southern Economic Journal used the staggered introduction of VAR across top divisions in 16 countries between 2009 and 2019. It found negligible effects on home advantage and on most key match statistics, although yellow cards and offsides declined. A study of the Turkish SΓΌper Lig by Ali IΕΔ±n and Qing Yi, covering 1,838 matches, reached the same conclusion: home advantage in points and goals was present with and without VAR.
The most plausible reading is that VAR reviews only a small set of decisions: goals, penalties, direct red cards and mistaken identity. Most of the marginal decisions where crowd pressure could matter, such as free kicks in midfield, minor yellow cards and added time, are outside its scope.
The crowd and the players
Crowd effects on the players themselves are harder to measure than effects on referees, but the pandemic studies offer indirect evidence. In the European sample, home teams created fewer attacking opportunities without supporters. That suggests at least part of the effect works through the home team's behaviour, whether through motivation, energy or confidence, rather than only through officials.
Familiarity and travel
Before the modern era, pitches varied enormously in size and surface. Early versions of the Laws of the Game allowed pitch lengths of roughly 91 to 119 metres and widths of roughly 46 to 91 metres. Today's international standards, used by UEFA for its competitions, require 100 to 105 metres by 64 to 68 metres, and top-level grounds are close to uniform. Pitch familiarity has therefore lost much of its value.
Travel has changed just as much. A long rail journey on a match day in the early twentieth century and a short charter flight with a sports science team today put very different demands on a squad. Better recovery, nutrition and medical support all reduce the physical cost of playing away.
Tactical and structural changes
Two more developments probably contribute, though the evidence is mostly indirect. Elite teams now play with consistent tactical models and possession-based structures, whatever the venue, which reduces the influence of local conditions. Financial concentration in the Premier League means the strongest clubs often win at almost every ground, so the result depends more on squad quality than on location. These explanations are plausible, but neither has been tested as cleanly as the effect of crowds on referees.
What the evidence supports and what it does not
It helps to separate what the evidence supports strongly from what remains plausible but unproven.
html
<table>
<thead>
<tr>
<th>Claim</th>
<th>Strength of evidence</th>
<th>Main basis</th>
</tr>
</thead>
<tbody>
<tr>
<td>Home advantage in the English top flight has declined over the long term</td>
<td>Strong</td>
<td>Complete results since 1888; consistent decline in home win rate, home share of points and home goal margin</td>
</tr>
<tr>
<td>Home advantage fell in many leagues when crowds were absent</td>
<td>Strong for European football overall; mixed for individual leagues</td>
<td>Multi-league studies with thousands of matches; league-level variation in later statistical work</td>
</tr>
<tr>
<td>Crowds influence referees' disciplinary decisions</td>
<td>Moderate to strong</td>
<td>Yellow card changes in several empty-stadium studies</td>
</tr>
<tr>
<td>Referee bias explains most of home advantage</td>
<td>Weak</td>
<td>Bias diluted after controlling for dominance; VAR introduction did not reduce home advantage</td>
</tr>
<tr>
<td>Home teams play better in front of their own supporters</td>
<td>Moderate</td>
<td>Fewer attacking chances created by home teams without crowds</td>
</tr>
<tr>
<td>Pitch standardisation and easier travel have reduced home advantage</td>
<td>Plausible but not directly tested</td>
<td>Historical change in pitch rules and travel conditions alongside the long-term decline</td>
</tr>
<tr>
<td>A single low season shows that home advantage is disappearing</td>
<td>Weak</td>
<td>Single-season variation is around 2.5 percentage points; five-season averages are more reliable</td>
</tr>
</tbody>
</table>Why this matters for reading results and markets
For anyone analysing matches, the practical conclusions are fairly direct.
Old rules of thumb need updating. A rule like "home teams win about half of their matches" described the English top flight in the 1950s, not the Premier League today. Over the last five seasons, home teams won about 44% of Premier League matches and away teams almost 32%.
Use the right measure. Home goal margin and home share of points are more stable than win percentage, because they capture more information per match. Comparisons between seasons or leagues should use those measures, ideally averaged over several seasons.
Expect variation between competitions. The pandemic research shows that home advantage depends on the league and its context. A figure from the Premier League says little about the Bundesliga, a lower English division or international tournaments, where the home side is often only nominally at home.
Treat individual teams with caution. A team that wins rarely at home over half a season is usually showing a mix of fixture difficulty, finishing variance and squad problems. Assuming there is something wrong with its home advantage is rarely justified by such a small sample.
The advantage that remains
Home advantage has not disappeared, but it has become modest. In the English top flight it has fallen from around a goal per match in the 1890s to about a quarter of a goal today. Empty stadiums confirmed what researchers had long suspected: crowds influence referees' decisions and, through the home team's own play, the flow of matches. They also showed that crowds are not the whole explanation, and that the size of the effect varies from one competition to the next.
The decline has several causes, and they reinforce each other. Uniform pitches, easier travel, more professional refereeing, consistent playing models and financial concentration all reduce the factors that once made home grounds distinctive. Home advantage is still worth something, but it is a smaller factor than most football folklore suggests.