The early 2000s saw baseball transformed. The Oakland Athletics, which was led by Billy Beane, used data analytics to find undervalued players. This created “Moneyball” approach, and subsequent film that detailed its revolutionary impact.
The strategy changed how the sport (and other sports) was analyzed. Yet, it never had much of as big of an impact on soccer as expected. This article looks at why, and how it may explain why soccer betting feels so different (and difficult).
Discrete vs continuous
The core difference lies in the game’s structure. Sports like baseball and cricket are largely discrete. This means they consist of distinct well-defined actions – think of this as parameters. A pitch; a bowl; a throw.
These moments have a beginning and end, and its contained repetition makes it easier to analyze with numbers, because variables can be more easily isolated. Data points are simply easier to collect, and deviations from the norm can be spotted.
In soccer, the game is more continuous. In fact, entire rule sets and referee ideologies are built on “letting the game flow”. One sequence melts into the next, and it becomes very difficult to collect data and contain the analysis. The simplified football betting odds do not do its complexity justice.
The best way to understand this is by freezing a cricket delivery – its outcomes are plentiful, but still limited. If you freeze a soccer match, it’s difficult to calculate the number of things that can happen over the next few seconds, let alone 90 minutes.
Incalculable data
As the ball moves across the pitch and time passes, the situation compounds the complexity. Pure statistical analysis (and therefore betting) is incredibly difficult and rare.
Christofer Clemens, who was head analyst for Germany’s 2014 World Cup winners, noted a perceived lack of data providing real information about game-winning factors. Metrics like Expected Goals (xG) are a good attempt to quantify scoring chance quality, but they must account for many variables. Also, there is a lot unsaid behind this data, and how different styles match up against one another.
How soccer’s nature shapes the betting landscape
No matter how data-focused the bettor, soccer often feels different. In fact, most people feel that, unlike many other sports, it’s one you shouldn’t be betting on if you don’t watch a lot of it.
This is where gut feeling, a sense of who has momentum, context around upcoming fixtures (e.g. will certain players be rested), and manager confidence (e.g. current media scrutiny) all help create a picture about what might happen in a game. To think that comparing expected goals could paint this picture would be naive.
The result is that, unlike many other sports, soccer betting hasn’t really evolved all that much given the size of the viewership. Tactics within the game have, and so has punditry, but this is qualitative, and in fact only further complicates the analysis for betting as it highlights more variables.
Is this the fate of soccer betting?
We can say with high certainty that soccer will never be “solved”, not quantified. For this to change, the ethos and protective culture around “letting the game flow” will need to be completely reversed. However, small pockets within the game can be more easily quantified, namely set-pieces, which is where we may see the data-driven market lean towards.
