Basic principles

Let’s look at ten basic principles of football and see how they connect with complexity. Here I use the term complexity as it is used in information theory and cybernetics. A cybernetic system has a higher complexity when there is more variation. With less variation there is less complexity. Variation is a measure for the number of options there are.

In terms of football: the more options you have, the more complex the situation becomes because it becomes harder to make decisions. That does not mean that it becomes more difficult to play. In fact, most of the time, having more options makes it easier to execute the decision you make. Because you have more options, it becomes more difficult for your opponent to anticipate your actions. You and your team become less predictable. So the more options you have, the less your opponent is able to resist you. Thus play becomes easier for you. To conclude with more complexity of the decision making process becomes harder and the decision executing process becomes easier.


Width is an attacking principle. With more lateral options, your opponent needs to stretch their defenders to close all avenues of attack. That means less defenders in the space of your attackers. Less defenders means more options. So width is a good principle of attack because it increases complexity. With more width your attackers get more options. The attack becomes less predictable and it becomes more difficult for your opponent to defend against you.


Players without the ball need to support the player with the ball. If the supporting players create opportunities for the player with the ball to pass, then options increase again and so does the complexity. If supporting players get close to the player with the ball but not too close, passing opportunities increase, but also the ratio between attackers and defenders in the same space tilts to favor the attack. The more options, the more complexity, the less predictability and the better your chance to score. If supporting players get too close to the player with the ball, then they start to limit the space available to that player and they start to limit the options he has. On the other hand if a supporting player runs into an open space and pulls a defender with him, he is creating more options for the player with the ball because there are less defenders to limit his options.


Support and width are attacking principles that a team can always employ independent of your opponent . Penetration is more of a goal that the team tries to achieve. Yet, your opponent can prevent you from penetrating their defensive lines. The reason why penetration is desirable, is that when you penetrate the opposing defense you are able to get the ball under your control in a space that has less or no defenders. That in turn increases your options, increases complexity and increases your chances to score.


If your players are staying at the same spot on the pitch, it becomes a lot easier for your opponent to limit the options of those players. Your opponent can block pass lines or limit the space your player is. Resulting in a decreasing number of options static players have. When your players keep moving around, it becomes more difficult to block pass lines and limit the space your players have. Preventing the decrease of options and thusly maintaining the number of options your currently has, make sure that your team is not forced to play in more and more predictable patterns. Of course, when mobility is used to support the player with the ball or penetrate the defensive lines, options increases.


A creative player has more variation in his behaviors. His play becomes less predictable because he has a larger palette to choose from. Besides having more options, a creative player also needs to have flexibility to actually use all available options to him. Finally, a creative player also must have the technique to execute the decision correctly. So adding creative players to your team automatically increases the options your team has. Nevertheless, as most of the additional options are often quite difficult to execute correctly, the risk of losing the ball also increases.

Yet, there is another source of creativity that is not linked to individual players and that is the team itself. As long as the team plays in a way that creates more options, creativity will follow automatically. We have already seen that more options, leads to more unpredictability. And creativity is defined in football as doing something unpredictable. So even without creative players, the team can still play in a way that is less predictable.


The less space between defending and attacking lines, the more compact a team is. Compactness is a defensive principle, because the aim of compactness is to try to have as many defending players as possible in the smallest possible space in order to prevent the opponent getting significantly closer to the goal with the ball. The more defending players there are in a small space, preferably the space where the ball is, the less options the attacking team has. With less options, the level of complexity decreases, so the game becomes easier. Thus it also becomes easier to defend. With less complexity there is also less risk. So the game becomes less risky. Less complexity also means that the opponent is easier to predict, again making it easier to capture the ball. 

It really doesn’t matter whether the compact block is close by the keeper or high up on the pitch. Playing high up the pitch with a compact block, makes it easier for the opponent to try to penetrate the defending block by passing in the space behind it. But that is also predictable and the defenders can be trained to anticipate that move to counter it.


Although it is of course best to regain control over the ball, interrupting the attack of the opponent is also good to do. Football is not only limited in space, it is also limited in time. The more time there is, the more options there are for the attacking team to score. The less time, the less options. So interrupting the effective use of time also limits the options for the opposing team. Of course, this is especially handy if you are actual ahead in the match. Time is a sword that cuts both ways. As you are limiting the amount of time the opponent has, you are also limiting the amount of time your team has as soon as it regains possession.

If time is running out, you see the team that is about to lose or who doesn’t want to draw, play more and more predictable. That doesn’t automatically mean that the team that is winning, will definitely win in the end. That also has to do with the game intelligence and technique of the players on the pitch. But it does indicate how important it is to score more goals than your opponent. The moment you do, not only do the number of ways they can win decrease (for instance they can’t win 0-1 anymore), but interrupting their attack costs them time. Which limits the options they have.

Less space and time

Pressure doesn’t really exist in football. There is only the amount of space and time an individual player gets to make and execute decisions. The less space and time, the harder football becomes. Unless, there is too much time and space and players start to overthink the situation. In that case that particular player can’t handle the highly increased level of complexity of his situation and he blunders. But in most cases, the less space and time, the harder it becomes.

The reason why less space and time, make it harder to play, is because as a player you get less options. So because you have less options and your actions will become more predictable, it becomes harder to beat your opponent. Of course, when you are a defender and you want to capture the ball, it is great when the attacker has less options. The more predictable the attacker is, the easier it is to win the ball.


Patience is a virtue. Patience itself won’t decrease or increase complexity. Yet, patience is needed to wait for the moment complexity has dropped so much that the behavior of the attackers has become so predictable that there is a high chance of capturing the ball. If you act too quickly as a defender, before you have shut down as many options as possible, then there are still options open for the attacker. The more options the attacker has, the smaller the chance you have to predict the behavior of the attacker and capture the ball. So patiently abiding your time till the level of complexity has dropped enough, is often the best defending strategy.


Due to the close reverse connection between complexity and predictability, I have already explained everywhere how defending is a matter of decreasing complexity in order to increase predictability, whereas the aim of attacking is to increase complexity in order to increase unpredictability. Besides everything I described so far, the defending team can also specific strategies to increase predictability. For instance, a manager can use data analysis to find out which attacker of the opponent most often loses the ball. Or is the least creative and the most predictable. Then the team can employ a strategy where the opponents options are limited when it comes to passing to other attackers, so the ball gets passed to the attacker most likely to lose the ball.

Bayesian brain

Our brain is a Bayesian biocomputer that foresees the future. Or to put it more precisely: the workings of our brain can best be described with Bayesian statistics. That is why philosophers, psychiatrists and neuroscientists now talk about the Bayesian brain. At the lowest level the workings of brain cells can best be described with Bayesian statistics. (Doya 2007, Bayesian Brain) Brain cells are triggered to fire a signal depending on a jolt of electricity. Pathways that are triggered easily and often require very little electricity. Pathways that are rare and difficult to trigger require much more energy. Yet, the energy levels don’t rise and fall lineair, but they use a certain quantum, i.e. discrete jumps or drops in energy. These quanta pretty much follow Bayes Theorem. The same goes for the neurotransmitters brain cells send the next cell. This is a kind of morse code that again doesn’t rise or fall lineair, but also uses quanta that pretty much follow Bayes Theorem.

At a bigger level the brain can be divided into two parts: (a) a bottom up part that processes sense data and (b) a top down part that processes our expectations. What makes our brain Bayesian is that our brain tries to make a prevision of what it is about to experience and then checks the incoming sense data to see whether the sense data matches what was expected. Our expectations are created through associative learning and instrumental learning. Or to put it in terms of football: game intelligence and technique.

What is interesting is that when there is discord between our expectations and our sense data, the brain tries to reason the difference away. This is mainly done by the brain to see if it can ignore the incoming sense data and stick with its expectation. This leads to inattentional blindness, fallacy and biases. This process is the main reason why players with game intelligent are often the better players. Players who lack game intelligence are literally blind to unexpected patterns. Players who lack game intelligence can literally not see opposite defenders and attackers if they do something unexpected. On the pitch this looks as if the player is blundering, but in reality his model of the game inside his Bayesian brain that produces his expectation is too poor to deal with the complexity of the game.

The Bayesian brain also explains why the influence of the manager on his players can decrease over time. Over time players don’t hear what the manager is actually saying, but literally hear what they expect him to say. For more details on this phenomenon, see: bias.

What is better live, video or data scouting?

The Bayesian brain explains why you need all three forms of scouting: live, video and data scouting. The reason is that the brain of the decision maker works through Bayesian principles. That means that the decision maker takes into account all the recommendations from different sources and comes up with the final decision. The more trustworthy sources the decision maker has the less risky his decision will be. A decision maker that only uses one form of scouting will make more risky decisions because he has less data to work with.

Yet, acknowledging this also brings additional responsibilities for all scouts, no matter whether they scout live or through video or data. Scouts have to make their probability estimations explicit! A live scout can’t simply give a player an A and an advice of getting the player as soon as possible. No, the live scout also has to give a concrete number for the probability that the player is going to be a success. That is a number between 0 and 100. The same goes for the video scout. The same goes even for the data scout. Because almost all data scouts come up with a player report with lots of numbers, but almost never with a number for the probability that the player will be a success.

When all the scouts actually make their unconscious probability estimation explicit, then the decision making process can be made more rational. By building a Bayesian network that uses all the inputs from the scouts, the club can rank the players as to who would have the highest probability of success. 

More importantly, one can then track how good the probability estimations of the scouts were by giving penalty points for the gap between prediction and reality. It will take up some time to build up enough reference points. Yet, once that has been done, one immediately sees which scout is most trustworthy. That enables you to enhance your Bayesian model by giving more weight to the opinion of the most trustworthy scouts, independent on whether they are live, video or data scouts.


In football it is very important to pay attention to actual behavior rather than cognitivist constructs. Behavior can be distinguished between internal and external behavior. To check to see whether something is external behavior there is the MARCO acronym:

  • M stands for measurable. External behavior can be counted. Passing is external behavior as it can be measured.
  • A stands for active. Someone needs to be doing the external behavior. There is a dead man’s test here: if a corpse can do it, it ain’t behavior. Lying on the pitch ain’t behavior as a corpse can do it as well. Tackling is external behavior as a corpse can’t tackle someone.
  • R stands for reliable. External behavior can not only be measured, but most people will reliably come to the same conclusions when counting external behavior.
  • C stands for control. External behavior has to be under the control of the agent. “Go and win this match” is a bad suggestion for a manager to make as it suggests something that is not under the control of the agent. “Go and score a goal” is a good suggestion as scoring is under the control of the player.
  • O stands for observable. External behavior can be observed by third parties.

Here is a great example of external behavior:

Our subjective experience is internal behavior. We all understand that in order for our body to produce external behavior, our brain has to do a lot of processing. The same goes for our subjective experience. Subjective experience is not a mental state, but the end product of a lot of unconscious processing in the same way that our external behavior is not a state, but the end product of a lot of brain processing. 

Internal behaviors basically boil down to:

  • Feelings and emotions.
  • Visualizing by remembering what the past looked like.
  • Visualizing by making a memory like fantasy about the future.
  • Inner self talk.
  • Remembering what other people said or how music sounded.
  • To be complete: memories of taste and smell are also internal behaviors. But for most people these are so unconscious that they have very little subjective experiences of these and therefore we can ignore them most of the time.

For scouts it is crucial to make sure that when they report on a player they only report on external behaviors if they can’t actually interview a player to ask him about his internal behavior. For coaches and the manager internal behavior is very important. Each player has a distinct set of internal behaviors that help him achieve his best performance. This depends a lot on the biological hardware structure of his brain that most people think of as the player’s personality. For some players it is important that they feel good or that they feel relaxed. For other players, it is important that they feel tension and excitement. Knowing which player needs what kind of internal behavior makes a great manager.

Visual internal behavior works differently. When we create memory like fantasies about the future our conscious mind thinks we are contemplating that future. But our unconscious mind sees those memory like fantasies as instruction videos to be carried out in the future. Therefore it is important that players refrain from visualizing mistakes or setbacks that make them feel bad. It is much better to first visualize setbacks and see upfront that even though there is a setback, the player keeps on feeling whatever feeling helps him perform at his best. After having visualized these conditional negative scenarios that are less negative as the player sees himself reacting correctly to setbacks, it is time to visualize positively. It helps players to achieve their best performance to actually see them doing inside their mind’s eye before they actually go out and do it.

Inner self talk seldom helps players achieve their best performance. Players need to learn to trust their body and unconscious mind. All those training hours improve their technique and game intelligence. During the match for most players it is best to actually be in the moment and perform rather than overthink situations.

Avoid mental constructs

Behavioral analysis as described above was mainstream in the period of the late nineteenth century up until the mid sixties of the twentieth century. Around that time behavioral analysis was replaced by cognitivism. Cognitivism took the functional behavioral approach of cybernetics and started to create mental constructs inside our mind. Examples of mental constructs are:

  • Motivation.
  • Will power.
  • Intuition.
  • Respect.
  • Trust.
  • Confidence.

Current thinking in football is heavily influenced by cognitivism. Yet, cognitivism is failing because neuroscience is not finding brain cells that produce these mental constructs. So rather than using mental constructs and fooling yourself into believing that these mental constructs explain anything, it is much better to go and look at the actual external behaviors combined with internal behaviors if you are capable of interviewing the player.


Everyone is biased. There are explicit biases and implicit biases. Explicit biases are at least recognized by the people surrounding you and if you are honest with yourself also by you. Implicit biases are unconscious and, by definition, cannot be known. Though there are tests that will show you how your implicit biases influences your decisions and actions.

Biases are closely related with the expectations that form an important part of the brain. The brain has basically two systems: one to process sensory data and one our process our expectations. Knowledge and skills are stored in the brain through associative learning and instrumental learning. These learnings can best be described as a model of the world. In football terms: players learn technique (instrumental learning) and game intelligence (associative learning) and create a model of the beautiful game, of their team, their teammates and themselves. Based on this model the brain calculates what it expect to sense next.

The theory about how the brain works, is called Predictive Processing. The idea is that there is way too much information in the environment for the senses to take in. So rather than see everything, our senses only process relevant changes. What happens is that the brain creates an expectation of what it is going to sense next and compares that expectation with what the data the senses provide. As long as the sense data is in line with the expectations the brain is happy. Yet, if there is a conflict between what the brain expects to sense and the sense data, then things become interesting. In most cases the brain will reject the sense data and hang on to its expectations. Or to be more precise: the brain will reason the differences away. But if the sense data persistently and impactful differs from the expectation the brain created, then the brain is forced to update its model of the world.

This is the reason why players who have less game intelligence than average literally don’t see opposing players or the ball. The manager might ask himself: “How come that my defender didn’t see that attack coming?” The answer is: the ray if lights of the attack did reach the eyes of the defender and were processed by the brain of the defender, but because the defender did not expect the attack to happen the sense data was overruled because it conflicted with the expectations of the brain. As a result the sense data of the attack was deleted and the defender literally didn’t see the attack happen. That is the reason why an attack with little predictability has a better chance of success. If the unexpected happens, players with less game intelligence than average, will fail to see patterns developing. 

Another example of how expectations and implicit biases work, is when a manager after a few years of doing great things with the team, suddenly is unable to influence his players anymore. Most of the time this happens when he wants to change the system he is using because the results are less than satisfactory. But no matter how much he tries to explain his new ideas to his players, it seems as if they don’t listen to him. In fact, they literally don’t hear the new things he is saying. Instead they hear what they expect him to say. This is one of the reasons why Louis van Gaal has a reputation of doing weird stuff in the locker room. When he feels he can’t reach his players, he does something completely out of the ordinary to shock the players so they stop prioritizing their expectations and instead let their sense data through.

The most common biases

So now that we have seen the underlying structure of biases, we can look at the most common biases:

Confirmation bias

Confirmation bias is a bias where the brain embraces information that support the current beliefs of the person and at the same time dismisses information that contradict the person’s current beliefs. Confirmation bias can easily be explained with Predictive Processing. Information coming in through the sense that confirm to the expectation of the person are taken in by the brain, whereas information that contradicts the expectation is deleted. Even though confirmation bias is well known, many people who know about confirmation bias, still make mistakes due to confirmation bias. 

The reason is that confirmation bias is so much built in the way the brain works, that conscious understanding is simply not enough to get rid of it. It is an unconscious process. The best way to countermand confirmation bias is the use of statistics. But only statistics that shock you now and then. Because if you do not strongly disagree with your statistical data, chances are that you selected that data source because it confirmed your biases.

Survivor bias

We all love Messi. Let’s research how Messi learned to play football so well and apply those lessons to all other players. Sounds like a great idea, but the idea suffers from survivor bias. Maybe Messi’s development is a lesson for many, but as long as don’t know how many players tried to do the same and failed, we don’t know whether the path Messi took is any good. Theoretically it could be that only 1 in 100.000 players actually learns to play football at a professional level following Messi’s development plan. 

In football many people suffer from survivor bias because a lot of attention is paid to players who make it. We all love a positive story. But the negative stories of players who did not make it, are at least as interesting, if not more. Focussing only on the positive is called the Via Postiva, the positive road. Many people prefer to take that road. But the Via Negativa, the negative road, is in many cases more important.

Another example of survivor bias happens in data scouting of players. If a player breaks through, data scouts start looking in the history of the player, to see whether there is a specific pattern that already signalled his greatness years earlier. The idea is that if this pattern is recognized it can be used to find future talent. But often, the data scout is uncareful with how many years he goes back. For player A he finds a pattern three years ago, and then he confirms his find for player B, even though he has to go five years back. So here you see both Survivor Bias and Confirmation Bias at work at the same time. Even worse, the data scout never looks at how many failed players had the same pattern.

Selection bias

Survivor bias is a form of selection bias. In the case of survivor bias the selection bias is selecting only survivors. But there are many more selection biases. For instance in the Eredivisie Dutch clubs have a strong selection bias to prefer players who have played in the Eredivisie. One can argue that for Dutch clubs to prefer to have Dutch players is okay. But the selection bias of Dutch clubs goes beyond that. They also have a strong preference for foreign players who have played in the Eredivisie. The Eredivisie is a relatively weak league. So it is strange to prefer foreign players who done (relatively) well in a weak league over foreign players who have done (relatively) well in a stronger league.

Belief in truth

Belief in truth in itself is a bias. If you would define bias as a systematic deviation from the truth, even though you are trying to recognize your own biases, you fall for the belief in truth bias. Truth can only be defined in formal logic and mathematics with the use of truth-tables. There is no truth outside of mathematics and logic. Everything empirical has some measure of uncertainty.

Habermas, one of the most stern supporters of truth outside of mathematics, defines true statements as statements that will hold up for everyone, everywhere and always. That means that if something is true, it is true for us and it is true for people living in Australia. It is true for everyone who will ever be born from this day forward. In fact it is even true for everyone who has ever lived so far. If aliens from outer space visit our planet, then they too will recognize a true statement as being true. While this holds for mathematical and logical statements, it fails for empirical statements. The bar for truth in the empirical world is too high.

Nor is truth needed in the empirical world. Probability gives you everything you ever wanted out of truth. Rather than saying that X is true, it suffices to say that X is highly likely. The self defeat tests of Bayesian statistics allow you to define bias much more elegantly. A bias is a probability estimation that leads people to systematically lose bets or gain unnecessary penalty points in a loss function.

Business Insider has a nice overview of 20 biases:

All of these are correct, except placebo. Placebo is the natural rate of healing in humans. Medicines need to be better than this natural rate of healing. Strengthening believes by talking to people is hypnosis or the use of hypnotic language patterns in communication.