Does defense win championships?

“Defense wins championships,” that annoying commentator announces. Every soccer fan has had this similar experience: the television is showing a game of the league leader. They’re leading your favorite team by a few points, and you desperately want them to lose. They’re being slaughtered, but they just won’t lose, and then in the injury time, they score a goal and take three points. “When you don’t concede, you don’t lose,” says that annoying commentator. “This is what champions do. Their defense wins them championships.” You hate him for saying something so overt.

       “Defense wins championships” is an adage for most sports. But scientists are using data and modern approaches to challenge it. Jon Wertheim and Sam Sommers questioned this belief in their book “This is Your Brain on Sports.” They found that in the NBA and NFL, better defensive teams don’t win more championships or games than better offensive teams. It’s a similar story in soccer: 41 champions of the top four leagues in Europe (Bundesliga, La Liga, Premier League, and Serie A) have the best defense, compared to 39 that didn’t. In contrast, the best offensive teams won 46 times (compared to 36 that didn’t). Therefore, winning doesn’t seem to require the best defense.

       But Defense does win championships in some places: In a league that focuses on the defense such as Serie A, 14 winners of the past 20 years have had the best defense, compared to six that don’t.

Offense doesn’t carry the same weight. Only half of the Serie A winners has the best attack. The adage is still valid in Italy.

Defense wins games

How does defense win games? We remember that commentator, “When you don’t concede, you don’t lose.” Everyone knows that! In soccer, as opposed to sports like bowling, losing requires one team to concede. You can’t lose if you don’t concede. But not conceding also gives you a better chance to win based on probability. There are three outcomes of a soccer game (in a league format): win, draw, or loss. You have an even chance (33.3%) to obtain one of these outcomes. But when you don’t concede, there are only two possible outcomes: win or draw. The odds of winning (or drawing) increase to 50%. Therefore, defense can help to win because of the probability of statistics.

       We can test a phenomenon of probability by statistics. Surprisingly, the result shows that defense doesn’t just help to win a game by removing loss as an outcome. It pushes a team toward winning.

In Serie A for the last 20 years, a team has a 70% chance of winning a game when not conceding a goal. Every year, the observed frequency of winning a game when not conceding is 15%-33% higher than would be expected based on probability. Moreover, conceding a goal pushes toward losing: when conceding a goal, a team has a 50% chance of losing the game, almost 17% higher than would be expected by probability. They are also 10% less likely to a win. Therefore, not conceding a goal helps a team to win a game, not merely by probability, but also by some unknown force(s).

You may think that winning also favors not conceding. Leading on the scoreboard (a necessary step to get a win) may discourage the opponent (psychologically) and make it harder for them to score. But the results suggest otherwise:

In the last 20 years, a winning team is 53.3% likely to not concede, a mere 3% increase of what would be expected from probability. Therefore, not conceding favors winning but the opposite is not true.

Defense tells us more than we thought

This analysis suggests that we should look at defense from a new angle, and we should study soccer data differently from the past. Defense helps winning, but we don’t know how. Another adage in soccer provides some insight; Soccer professionals and enthusiasts agree that not conceding is an essential measure of a team’s performance: you prevent conceding when you don’t make a mistake. Players need to focus and carry out their duties when defending. They need to maintain a prime mental and physical state. Such state of mind should not only benefit the defensive phase of the game. It should also impact the offensive phase of the game. Therefore, not conceding is a sign that players are playing the peak of their games. Analysts should change the ways they analyze the game data. For example, they should identify what prior conditions promote/hinder a team not to concede. Does any training method facilitate players not to concede? Is there a level of mental and physical fatigue that promotes conceding? Looking at Defense in a new perspective allows us to understand the game better and improve it further.

       “When you don’t concede, you don’t lose.” Next time you hear that you can tell the commentator that he is wrong. Defense wins games, period.