KickForm is a unique platform that allows football fans to predict individual results without mathematical skills. It’s a statistical model, developed by physics professor Andreas Heuer, aimed at creating a forecasting program that would produce accurate, repeatable, and highly profitable football predictions.
Data is taken from KickForm and compared to bookmakers' odds to find value. Its success has been revolutionary, with a proven track record yielding results even more accurate than those of many mainstream bookmakers.
How KickForm Works: 7-Step Process
KickForm, at its core, is broken down into seven steps. Now, you don’t need to have a maths degree to understand how it works, but for those of you who want a look under the hood, here’s how we do it.
Step 1: Determine Home Advantage
We start by determining the advantage the home team has in the game. Once we have this, all other metrics fall into place, allowing us to create the most logical result for the match.
This is the formula:
Home advantage = c1 (home advantage of the last three years) + c2 (home advantage of the current season)
On average, home teams tend to score more (Statistics: 1.66 goals for home and 1.20 for away).
Step 2: Calculating the Number of Goals per Match
Next is determining the average number of goals scored per match. This considers all teams in the league.
Typically, three goals are scored in a football match. However, for the sake of precision, the number of goals has increased over time and now lies at 3.13 in the Bundesliga and 2.93 in the English Premier League.
Step 3: Calculating the Performance Level and Expected Goal Difference
The formulas below calculate the difference between the performance level and the expected goals.
Performance level = c1 X1 + c2 X2 + c3 X3 + c4 X4
X1 = mean goalscoring difference (GCD) of the previous season, weighting of the last three years (0.5, 0.35, 0.15)
X2 = goalscoring difference of the current season
X3 = current fitness value (mean goalscoring difference, weighted with a decreasing exponential function)
X4 = logarithms market value
Goalscoring opportunities are much more informative for forecasting than actual goals. Good teams display a slightly better conversion of chances. The prediction becomes a lot more accurate if the goalscoring opportunities of the current and the past season, as well as the market value, are taken into account. By doing so, correlations of up to 0.67 are made, resulting in a 67% rate of correct predictions.
Step 4: Determining the Exceptionality of Promoted Teams
The performance of promoted teams is surprisingly strong. Obvious deviations from the lower half of the table (goal difference: -13 +/- 8) are therefore quite rare.
Step 5: Calculating the Expected Number of Goals
For every match, the total number of expected goals is similar; however, some high-performing teams score more goals than average.
Number of goals = c1 X1 + c2 X2 + mean goals per match
X1 = total of goalscoring opportunities in the past, with identical weighting parameters of the last 3 years
X2 = effective total of goalscoring opportunities in the current season. Here, the total of goalscoring opportunities of all teams is subtracted so that the total number of goalscoring opportunities, compared to the average, is determined.
Step 6: Calculating Expected Goals
We assess the calculated goal difference and the total goals for the respective match.
Step 7: Matchday Weighting Factor
Weighting factors for the respective matchday or for the stage of the season are completed and applied.
Forecasting Results with the KickForm Football Formula™
To further explore how our KickForm Football Formula™ works, we’ve taken a real-world example from a game between Liverpool and Manchester United in the English Premier League.
Prediction after using the KickForm Football Formula™: 1.429:1.022
Obviously, no match results in a ratio of 1.429:1.022 goals, but this is the average number of goals scored by both teams. With the help of the Poisson Distribution, we can calculate these figures for the distribution of 100% using a row of results for each team.
The Poisson Formula itself is as follows: P(x; μ) = (e-μ) (μx) / x!
The following Poisson Distribution has been calculated from the above example:
| Goals | 0 | 1 | 2 | 3 | 4 | 5 |
|---|---|---|---|---|---|---|
| Liverpool | 23.95 % | 34.23 % | 24.46 % | 11.65 % | 4.16 % | 1.19 % |
| Manchester United | 35.99 % | 36.78 % | 18.79 % | 6.40 % | 1.64 % | 0.33 % |
This example shows that:
Liverpool have a 23.95% probability of scoring no goals, a 34.23% probability of scoring one goal and a 24.46% probability of scoring two goals.
Manchester United, on the other hand, have a 35.99% probability of scoring no goals, a 36.78% probability of scoring one goal, and an 18.79% chance of scoring two goals.
The most likely result, therefore, is 1:1.
This result will occur with a probability of 12.59%.
Creating Your Own KickForm Football Formula
KickForm goes a step further and offers each user the option to customise the formula by using a different weighting according to personal predictions. As a result, football fans can, even without in-depth knowledge of mathematics, develop their own forecasts on a scientific basis.
Registered users can choose factors such as market value, possession, home advantage, favourite team or away weakness. It’s designed to allow the Average Joe to become a maths wiz and their own betting expert with as much or as little input as they want.
Example: Weighting Factor Home Advantage
Let’s assume that Liverpool scores a high percentage of their total goals at home, and you want to weight this market towards that. You set the line at +0.6, and therefore, the result changes to 2.029 – 1.022. This means the most likely results change from 1-1 to 2-1, in favour of a Liverpool win.
How Was KickForm Born?
The idea for KickForm developed at a Stammtisch (in Germany, this is where a group of friends meet regularly at a pub). Jan was looking for a way to improve his ranking in the collective betting pool and dreamed of a ‘football formula' that would give him valuable tips to place better bets.
This also piqued the interest of Jörg, who thought it would be exciting to develop an algorithm that would allow for scientifically sound predictions of football matches. As a result, the idea for an online platform was born: one that could not only offer a football formula to predict matches but also allow each user to create their own formula, directly incorporating knowledge of goalscoring opportunities, possession, market value, home streaks, and so on.
Jan and Jörg then approached physics professor Andreas Heuer, author of the book Der Perfekte Tipp (The Perfect Bet) and an established expert in the field of football predictions. He immediately liked the idea and lent his support to the founders as an academic advisor.
Professor Heuer was approached specifically because his predictive model for football matches has been proven to be exceptionally accurate. In fact, compared to the odds of bookmakers, it was found that Heuer’s predictions were better than those of many of the top bookies you see online today. Mightily impressive stuff, we’re sure you’ll agree.
Meet the Team Powering KickForm
It’s impossible to predict how a football match will end with complete certainty. It’s one of the many reasons why this sport is so enthralling, and exactly why it is such enormous fun to analyse matches or to place a bet. The combined expertise of Professor Heuer and the team has developed a method for drawing accurate conclusions from statistics and understanding phenomena such as streaks in home games and the longevity of football managers in their roles.
Andreas Heuer is a Professor of Physical Chemistry at the University of Münster (Germany) and an expert in the theory of complex systems. Heuer has dedicated himself to these big football questions, such as whether the results of a tournament are predictable if a change of manager makes sense, and what impact chance has on the outcome. He’s been working on solving them with the help of science, and his findings from his studies are available not only in his book “The Perfect Bet” but also on KickForm.com.
Julia Benzing is a sports statistician from the Technical University of Dortmund and is one of the most vital members of the KickForm team. When she is not developing algorithms for KickForm, she grapples with questions like “Do the achievements of Borussia Dortmund have an impact on the quantity and quality of freshmen at the Technical University of Dortmund?” and other interesting topics. In fact, her master’s thesis tackled the relationship between football predictions and statistics (Statistical Methods for the Prediction of Football Matches), showcasing her dedication to footballing statistics and methodology.
Johannes is a student of mathematics at the Free University of Berlin (Freie Universität Berlin) and a football statistics enthusiast. His bachelor’s thesis (titled “The Optimal Football Bet”) was an intensive study of football betting. His theoretical calculations for precisely estimating the probability of betting events and determining the optimal wager to maximise capital with the lowest possible risk are also put into practice at KickForm. Johannes utilises KickForm's Football Formula with the Kelly Criterion Calculator against historical odds of up to eight years. At the end of this simulation, capital more than doubled on average per season.
KickForm Predictions FAQs
Scientists from the Wissenschaftler of Münster have thoroughly analysed the statistics of the German Bundesliga and characterised a football match as a Poisson process, making football results calculable. The crucial finding of the scientists is that goals in football are a product of coincidence. Goals, however, aren’t purely random but are influenced by the technical abilities of the players — the so-called performance level — of both teams.
Home teams score more goals on average — 1.66 home goals versus 1.20 away goals. There is an overall home advantage, but no evidence to suggest that teams perform especially well at home. Thus, home strength is a myth.
In the English Premier League in 2024/25, there was an average of 2.93 goals per game. In the German Bundesliga, the average was slightly higher, at 3.13.
Not necessarily. 46% of all wins are within a one-goal margin. The most common scorelines are 1-0 and 2-1.
No. Only 19% of Bundesliga and 24% of Premier League games in 2024/25 ended in a draw; the rest were all either home or away wins. As a rule, the draw is the least common result from a 1×2 market.
Yes. However, since the mid-1980s, the number of away wins has been constantly increasing. During the 1970s, around 20% of matches resulted in an away win; that figure is now at 36% following the 2024/25 season.
On the last two match days, about 20% more goals are scored than on average compared with the rest of the season. This means you can bet on higher-scoring games from your predictions.
The goal difference of previous matches is crucial. It is significantly more informative than the number of points. The expected number of goals, on the other hand, is quite similar among all teams.
The market value of a team, as determined before the start of the season, has a strong positive correlation with performance. Rule of thumb: doubling the market value of a team equals 10 additional points or a goal difference increase of 16.
Performance level = goal difference against an average opponent.
Coincidence averages out during the season. The longer the season progresses, the more reliably goals and opportunities affect team performances, thus the stronger KickForm becomes.
If coincidence were removed, the difference in goalscoring opportunities between the two teams would perfectly predict the performance level.
86% of goals scored is deemed to be a coincidence (per match day) or 29% (per season) on average.
Fluctuations in performance levels from match day to match day are not statistically relevant. Changes in performance levels typically occur during the summer break and only rarely otherwise.
In the context of statistical precision, the conversion of effective goalscoring opportunities is identical for all teams. For that reason, goalscoring opportunities are significant for predicting goals and should be taken into account.
With the help of market value and the effective difference in goalscoring opportunities, you can come close to a perfect prediction in the second half of the season.
No. In fact, only in half of the cases does the best team win the league and become champion at the end of the season.
No. Statistically, there are no bogey teams. The effect is under 10%.