The probability of victory for teams at the Men's EHF EURO 2026

Probabilities are extremely important in professional sports where barely anything can really be planned. Instead, the mantra especially for coaches is usually to “increase the probabilities” to either score or win. These conceptual probabilities can be translated into actual numbers with statistical models. For shots this is today commonly known as “Expected Goals”.

Scoring probabilities

Expected Goals (or short xG) are mainly known from football where they are regularly shown on tv broadcasts today. In an individual game xG are basically the sum of all scoring probabilities of all shot attempts.

To get these scoring probabilities thousands of shots taken in previous games have been put into statistical models which then calculates based on variables known at the point of the shot such as the position of the shot, the goalkeeper or the other players or the type of the shot and many more the scoring probability. These variables vary from provider to provider which is why there are different xG values for each provider.

Therefore, xG are the probability of an average player scoring on the respective attempt. For a single game, xG can be interpreted as a weighted shot statistic which is more meaningful than just the number of attempts, because it includes how dangerous the attempt is. In football, for example a shot from 40 metres (around 0.01 xG) is much less dangerous than an attempt from 10 metres (around 0.31 xG).

In football, chance plays a much bigger role than in other sports. The underdog wins games much more often than in high-scoring games such as handball. Expected goals are a good indicator here of who was actually the better team, more than the actual result. Studies have shown multiple times that assessing teams after their results by xG is a much better predictor for future success than their actual results.

In handball, this is a bit different. Because of the high-scoring nature of the sport, the raw xG value is less meaningful. But still, numerous very interesting insights can be derived from xG. For example, the average of all xG values from all shots shows the quality of shots. It shows whether a team or player is able to get high quality or rather difficult attempts.

At the EHF EURO 2024 the Netherlands had the best shot quality with 67.8 per cent and were great in attack in general. Their playmaker Luc Steins stood out in particular. He did not only set up countless good chances but also had by far the best shot quality of all the backcourt players. Excluding penalty throws, only four wing players and line players were better than Steins. The second-placed backcourt player was Dane Mathias Gidsel in 20th place.

A second possibility is to look at the finishing ability or shot making, which both is just the overperformance of xG by actual goals. Interestingly, Spain performed best here two years ago with 6.1 per cent more goals than expected, but had big problems in defence.

Sweden’s Felix Claar played a great tournament two years ago as he ranks first here with 23.4 more goals than expected. It also shows Mathias Gidsel’s greatness, as he is the only backcourt player that ranks in the top 20 in both shot quality and shot making, something he regularly does in the German Bundesliga as well. Overall, just five players scored, percentage wise, more than expected than the Danish right back.

The model used here and for analysis during the whole EHF EURO 2026 includes more than 100,000 shots with data that includes the distance and angle of the shot, the position of the goalkeeper, the hand of the shooter, the type of shot and the pressure on the shooter.

Winning probabilities

Based on xG, not only individual goals can be looked at but whole games. For example, looking at EHF EURO 2024’s xG shows that based on their shot attempts, France should have won the final by an even bigger margin than they did (33:31), as they accumulated 37.4 xG compared to Denmark’s 29.1 xG. France had nine more attempts, as well as higher quality attempts, with 4.1 percentage points better shot quality, while Denmark made up for this at least in part with 18.3 percentage points better shot making.

And not only that, xG can give us probabilities of how likely each team’s win was by simulating the game based on the xG values — the scoring probabilities of the shots taken. In about 95 per cent of cases, France would then have won the final as well.

These probabilities only exist when a game is played. But there are also possibilities to determine how possible it is that a team will win a game or even a tournament before the games are played.

The most common possibility is simulating the tournament based on team strengths. Based on these, every game of the tournament is simulated and then this is repeated a few thousand times. One possibility for these team strengths are so-called Elo ratings.

Originating in chess, the methodology for Elo ratings was developed by physicist and statistician Arpad Elo in 2018. Elo ratings are used in many sports today to rank players or teams, and for most official rankings including the FIFA World Rankings in football.

In general, the system awards points to winners and deducts them from losers, while winning against a higher-ranked team gives more points than against a lower-ranked team. In addition, more recent results are more important and home court advantage as well as goal difference and the quality of the competition are also considered.

One weakness of the Elo ratings, however, is that they are entirely based on the past. If, for example, several key performers are absent, this is not taken into account in advance but only retrospectively if the result is worse than previously expected.

For handball such a ranking exists as well. On handballranking.com the Icelandic programmer Ívar Jónsson publishes a daily updated world ranking for men's national team competitions based on the Elo rating on his own. His database consists of around 10,000 matches, in all official competitive tournaments and qualifiers.

These Elo ratings as team strengths allow us to calculate probabilities for who will win each match of the EHF EURO 2026 (or if there will be a draw). These allow the simulations of all games at the championships.

Based on the ratings from handballranking.com 50,000 simulations of the competition have been made. The share of a team winning the tournament or reaching the knockout stage or the main round in these simulations gives us the probability that they will win the tournament, go at least to the semi-finals or survive the preliminary round.

Based on their past results, with gold at both the Paris 2024 Olympic Games and the 2025 IHF Men’s World Championship, it is hardly surprising that the simulations see Denmark as the big favourites. In 77.8 per cent of the simulations they win the title, which means that our model sees the probability of them winning at 77.8 per cent.

The model puts the probability of Nikolaj Jacobsen’s team making it to the final at 88.4 per cent and making it to the semifinal at 95.3 per cent. In just 0.072 per cent of the simulations they were knocked out after the preliminary round.

Even though the gap is very big to Denmark, the team with the second highest probability of winning is France at 9.1 per cent. However, their probability of making it at least to the semifinal is only the fourth highest at 49.2 per cent, thanks to the bracket they are in. All four of the highest-ranked teams according to the Elo ratings — Denmark, France, Germany and Norway — are on the same side of the bracket and will be in main round group I if they survive the preliminary round.

Because of that, Sweden’s (54.2 per cent) and Croatia’s (51.1 per cent) probabilities of making it to the semi-final are higher than France’s, while their probabilities of winning the European championship are lower at 3.0 and 2.7 per cent. But they are still higher than Germany (2.3 per cent) and Norway (1.4 per cent), even though they have lower Elo ratings, because the model sees their side of the bracket as much more difficult than the other.

Click here to read article