Every season sparks the same heated debates: Who are the best teams in the world? Which league truly reigns supreme in the increasingly-competitive global football landscape? With so many elite teams, leagues, and competitions across every continent, it's getting harder to settle the argument.
But how do you rank teams and leagues that rarely, if ever, face each other? How do you compare a team from Brazil to one from England or a league from the USA to one in Germany?
The answer lies in a carefully crafted ratings methodology that takes into account everything from a team's form and historical performance to the relative strength of their domestic league. Global Football Rankings' (GFR) system doesn't just look at wins and losses; it evaluates how well a team performs relative to expectations and how they compare against others on a global scale.
Let's break down the methodology that powers the rankings, giving you a clearer picture of how we can objectively determine which teams and leagues are at the top of world football.
At the heart of GFR's rankings is an Elo-based algorithm, adapted for football's unique characteristics. Originally developed for chess, the Elo system adjusts teams' ratings based on match outcomes, comparing actual results to expected outcomes. In our system, we've introduced additional factors to account for the complexities of global football:
1. Match Importance: High-stakes games—such as cup finals or relegation deciders—carry more weight in rating adjustments.
2. League Structure Variations: The system accounts for leagues with different structures or those that have expanded or undergone recent format changes, like the 2024/25 Champions League.
3. Global Comparisons: To address the challenge of comparing teams across different continents and leagues with limited intercontinental competition, we use a 'True Elo' score. This score combines our 'raw' Elo ratings of teams, leagues, countries, and continents, then scales them to allow for meaningful global comparisons.
GFR's ratings represent a team's form and ability on a scale of 0 to 100. On this scale, 50 represents the average global team performance, with higher numbers indicating stronger teams. The highest-rated teams typically score in the 80-100 range, while the lowest-rated professional teams might score in the 20-30 range. Currently, we cover 75 of the top leagues in the world, with the lowest rated teams in our rankings coming in around the average.
These ratings are calculated using a combination of:
1. Historical match results
2. Simulated match outcomes based on team statistics
3. Adjustments for league strength and competition level
At the start of each season, we recalibrate a team's rating based on several factors:
- Previous season's final rating (primary factor)
- Player transfers: We assess the impact of key player arrivals and departures
- Coaching changes: New management can significantly affect team performance
- Squad health: Teams starting with injuries to key players may be adjusted downward
- Market value changes: Significant changes in overall squad value are considered
- League context: We account for changes in other teams' strengths within the league
Throughout the season, ratings are updated weekly. Here's how:
1. Before each match, we simulate the expected outcome based on current ratings
2. After the match, we compare the actual result to our simulation
3. Ratings are then adjusted based on the difference between expected and actual performance
- Exceeding expectations leads to a rating increase
- Underperforming leads to a rating decrease
- The magnitude of change depends on the degree of over/underperformance and the match's importance
This continuous adjustment allows our ratings to reflect a team's current form and long-term ability, capturing the dynamic nature of football performance.
Our system uses two types of Elo ratings: 'raw' Elo and 'True Elo'. Raw Elo ratings are calculated separately for teams, leagues, countries, and continents based on their respective performances. The 'True Elo' score is then derived by combining these raw ratings and applying a scaling factor. This process allows us to make meaningful comparisons between clubs from different regions of the world, accounting for the varying strengths of different leagues and continents.
When simulating match outcomes, GFR uses a Poisson distribution to model the expected goals (xG) for and against each team. The Poisson distribution is particularly well-suited for modeling rare events in a fixed interval, making it ideal for low-scoring sports like football. In our model, we use each team's historical xG data as the mean for the Poisson distribution, allowing us to generate probabilistic outcomes for each match. This approach accounts for both the quality of chances created (xG) and the inherent variability in football scores.
Each team's simulated results are compared across a number of possible match outcomes, from which we calculate the probabilities of wins, losses, and draws. These calculations factor in:
- Match importance
- Home/away advantage
- Team's current form and historical performance
Because most matches are played within domestic leagues, GFR also assesses league strength to contextualize team performances. Starting with a global average, leagues are adjusted up or down across a strength scale, ensuring that a team's rating reflects both its own ability and its competitive environment.
It's important to note that our published league ratings are simply averages of all the team ratings in the league. Given the extensive nature in which we derive team ratings, we have found that the average of all teams in a league is an accurate representation of that league for ranking purposes and passes the 'sniff test' when each league's rating is stack-ranked against other top leagues.
We are constantly working on our ratings to ensure that we account for the necessary variables found in the modern, ever-evolving game. In the near future, we will publish a dev log that offers visibility into our updates.
If you have any questions, please contact us.
The form has been successfully submitted.
Our team will contact you soon!
We will review and publish your platform soon!
We'll be in your inbox soon!
Our excellent customer support team is ready to help.