Thursday, 25.04.2024

How (not) to constrain your principles.

The game of football is a game with a specified purpose, laws and means. Two teams are facing off against each other with the objective of scoring a goal utilizing the ball. These rules are the same all around the globe. Similarly, the principles following these laws of the game will be the same – as will be the purpose of training. Football is a game that is trained and therefore coached, with most coaches focusing on doing so within their specific interpretation of these principles, for attacking, defending and the transition in both directions between these two.

Mostly, in order to improve the players or at least keep the level, variations of the game (instead of the actual 11v11 with all the rules intact) will be played. Some coaches focus on smaller games, some keep bigger games and change the rules of the game in order to create an incentive or constraint for improvement, while others will isolate a single aspect, often to stimulate individual behavior tendencies.

Normally every variation of a game should stem from the 11v11 with pre-planned changes to its design that, ideally, give a clear base for the application of coaching points or, at least, for the hopefully positively changed (inter-)action. Due to different sensory input (as in: perception) everything else will consequentially change, being it the integration (say: decision making) or the motor output (= technical execution) where the coach can include additional coaching points to the coaching points of the designated game.

All the changes in design in a drill (session) will thus usually impact perception, decision making and, at the end, execution of the football actions. This will always be either in regards to the last game(s) or the next game(s). In our quest as coaches to do so, we might over- or underfocus (= over- or underload) one aspect within our principles of play. This can be a good thing if done accurately to induce improvement – but it can sometimes work against what we did before (for instance, neutrals or touch rules that are used sometimes counterproductively).

A Change In Thinking

For this explanation we will use a very simple example. As some readers perhaps already gathered, on this platform we moved over the years more and more away from conceptual language to (inter-)actional language, as mentioned here with the case of man- and zonal marking. While one big advantage of conceptual language, especially for more theoretical pieces (and writing in general) is obviously the possibility to include a bundle of different situations, heuristics and patterns into a single terminological construct, it often fails short when coaching on the pitch or when trying to explain specifics. Ironically, by moving away from the output of game(s) and coaches to utilizing basics of the laws of the game and general principles of working with players a compromise can be reached where principles generated within interactional language can cover different situations.

The example to introduce this thinking will be our theoretical article on counterpressing. As was mentioned by some commenters and within our internal discussion circles back then, unlike more organized situations in defense (or offense), zonal and man marking concepts did not apply well in this case. Whilst the easy explanation would be that the phase of transition (until then rarely examined in a conceptual manner) is “just different”, the actual explanation right now would be that the concepts of zonal and man marking in organized defending are lacking and this lack was exposed in the counterpressing article unconsciously.

Instead of categorizing actions in “man marking” or “zonal marking”, a spectrum between that could be used, but zooming out it’s even clearer that the actions of “winning the ball”, “blocking passing options” or “covering team mates” would apply better. The principles in counterpressing and regular, organized pressing would be the same even, just with a different possibility to execute due to the nature of the situation. Put it this way: If (broadly speaking) the same “pressing situation” in defending would emerge as in a “counterpressing situation” in transition, would we really coach our players to solve the situation differently or would it stay the same?

The aforementioned article thus moved away from these constructs just to introduce new constructs that would just be a similar, smaller error whereas thinking in (inter-)actions and principles would be a much cleaner solution. If we have the same thinking in theory, we have the same mistake in practice – so in order to improve not only the players, but ourselves we should apply this thinking in analyzing, interpreting and teaching these situations.

In some situations in (counter-)pressing we might have to block the deepest passing option, in others we can solve the situation by immediately winning the ball from the opposing protector or we might back off to cover a team mate who is doing so – which is the same for every game played in the world, just done differently due to differences in the situation, the style of the team (incl. its coach) and the abilities of a player with the same purpose (on a general level) of not conceding a goal and getting into possession again in order to score one later.

Advantages and Disadvantages of Constraints

So we could design sessions that fit our preferred solution of a specific situation that we cut out from the game or on something that was not good and thus should be improved, right? But if the changes to our training and its reflected effect on the players’ behavior are not done well, it might become either ineffective or a circle of improving one aspect and worsen another.

Let us stay in the counterpressing example: In training we could use a rule that the opposing player winning the ball has to keep it a certain amount of time afterwards to receive a point – thus the opponents would have to win the ball directly from him to prevent this. With this constraint we might work against other principles – in the actual game the opposing player might play past our counterpress. If this happens, the training could be changed that the opposing player who won the ball has to pass into a different zone or into minigoals – which could lead to our players become more passive and only block the passing options. This would be principles working against each other within a single situation (or team task or game phase).

Another example could be principles working against each other over situations, team tasks or game phases: If we create a drill that focuses a lot on counterpressing, very often we might become unrealistic with our defending (making it too easy to enable the counterpress) or our attacking (making it too hard or too vertical to provoke the counterpress) and harm our development there. But if we play a 6v3-Rondo where the three in the middle can switch out by 1) a player going out by passing into a small goal outside of the pitch 2) two players going out by dribbling over the line and controlling the ball or 3) three players going out by passing to each other without interruption we might not compromise any behavior while enforcing it much more without not doing so in a regular 6v3-Rondo where a touch might be enough to become an outside player.

One heuristic: Never compromise a principle for another. Either bypass this effect by another rule or prevent it by not using rules at all and only coaching what you want at times. Still, keep in mind that your coaching can be like a rule and lead to similar consequences. Ask questions that allow players to find their solutions and not close off too much with our coaching. Just highlight certain principles within session design but ideally without compromising behavior in other phases of the game.

By Author RM.

Constraints: What are those?

Constraints have been defined as boundaries which shape the emergence of behaviour from a movement system, for example, the learner, seeking a stable state of organisation. (Newell 1986). These unique characteristics should be viewed as resources for each individual that channel the way in which the learner solves particular task problems or characteristics that can lead to individual-specific and context-specific adaptations. It is apparent that movement solutions will vary as each individual strives to satisfy the unique constraints upon themselves.

The key takeway here is that the ‘abilities’ (i. e., technical skills, game understanding, intentions and emotional skills) that players are ‘allowed’ to develop are determined by the environments and contexts to which they are exposed over time. The relational properties of rules created by constraints and opportunities afforded has been oft captured as seen in Rovegno, Nevett, and Babiarz (2001), they stressed the importance of situating learning in the game, not only to learn off-the-ball movement skills, but also to learn the tactics that go alongside skills. What complicates this, is that from an ecological dynamic’s perspective a player’s behaviour is attuned to their own action capabilities, as well as their teammates and the opponents. Therefore, it is vital when designing constraints, we are aware of whom and HOW the constraint is impinging.

Selecting the appropriate constraints:

As coaches we must be very aware of the situation, context and constraints we are creating. If you are asking your players to play two-touch only in the opposition third, but not explaining why you are asking them to do so, are they really learning the effectiveness of quick-layoffs to bypass pressure, or are they simply doing it because the coach has asked them too? Sure, they may learn from this situation, but the coach has to reinforce the learning.

Touch-limits for example create situations that coaches often expect players to learn from, but if we do not explain why we are using the limit, or what the limit leads to, the player will not become adaptive to the situation, and simply resort back to old methods once the constraint has been lifted. We use touch-limits to encourage an outcome, if the outcome is achieved for the coach but the player doesn’t understand why, but just thinks they have been asked to play with a touch limit, then the outcome will probably not appear in a game situation without the constraint or it will appear wrongly to how the coach envisaged.

I think these kinds of limits are good in practices that focus on combinations and timing of runs, they are arguably the best way to implicitly create quick combinations, whether this be deep-combinations in early build-up or final third combinations.

It is important to note the certain types of behaviours promoted from the selection of rules when designing practices, the coach needs to be aware of the difference in reward and restriction rule design. By promoting a behaviour (you get two goals, if you score after a pass from full-back to full-back) you will get a very different practice outcome to a rule of you must pass from full-back to full-back before you can score. Both types of reward/restricition have a time and place in which they are useful for rule creation/session design, and it is up to the coach to skillfully select the appropriate constraint. 

How and Why: Is it important?

The aspect of knowing both the “how” and the “why” is vitally important for coaching as coaching can be described as the process of applying contextual knowledge in action, (Lyle, 1999, Cote et al, 1995). It’s important to note how a coaches’ effectiveness and efficacy when implementing an approach can effect the delivery of the approach. Does the coach understand complex systems theory and Non-linear pedagogy? Understanding Complex Systems and NLP will allow the coach in theory to, explain, describe and predict how their athletes learning system (Chow, 2008) adapts and changes to the stimulus provided by the selected approach.

Sensemaking in coaching & Knowledge Construction:

Some research highlights that the nature of elite coaching can often be an inexact science and that it is imperative to understand that coaching is an interpersonal process, (e.g. Abraham & Collins, 1998; Strean, 1998; Lyle, 1999), using action-language as stated throughout this piece can go a long way to improving understanding between the coach and the footballer, helping to clear up and remove ambiguity from the complex and dynamic coaching relationship.

By developing a greater understanding in our players through clearer coaching actions, we can hopefully increase what Foucault would deem our ability to “sensemake” as coaches.

When coaching and attempting to further the understanding of our players it is important to understand what kind of “knowledge” of the specific context you are coaching you want your players to gain. Coaching can often be broken down into two areas of “knowledge”, procedural knowledge (or implicit knowledge) the knowledge of how of related to methods & procedures and the declarative knowledge (or propositional knowledge) which allows one to describes things, their attributes, and their relation to each other. Both types of knowledge are essential to deep understanding, and I challenge coaches to think of how their practices and interventions are shaped by which kind of “knowledge” they are developing in their participants.

When considering asking players questions during a coaching activity, I try to create a framework to help me cover as many different types of critical thinking skills as possible when questioning players. I understand with the complex nature of the coaching process, may different outcomes will occur per question, but I use this as a framework to rationalise a starting point.

 I outline them as:

-A knowledge for understanding (What led to this happening?)

A knowledge for critical evaluation (How could you have this specific situation better?)

-A knowledge for action (Why did this happen?)

A knowledge for reflection (Next time, what will you try differently?)

For players to be independent decision makers it is vital that coaches are skilled with their interventions, questions and overarching choice of language. As well as skillful session design, it is imperative the coach knows how to ask questions that allow players to find their own solutions without impinging too much on the players own decision-making process, by harnessing effective question design and football-action language. Lemke’s (1990) work is a particularly useful starting point for this because of its focus on the language of lessons, in his case science lessons, and the kind of learning that takes place. He is interested in how students learn to do science, that is, to observe, describe, hypothesise, question, design experiments, follow procedures and so on. His starting point is that language is not simply ‘vocabulary and grammar’ but ‘a system of resources for making meaning’. For me, this is an important moment of framing for the coach, if you can switch your viewpoint as the coach from language simply being “words” to a valuable resource which can affect each player, both in what is said, and the way it is said, the coaches efficacy is sure to improve.

Action Language and the need for it?

Football tends to sit in its own lexical circle of language, language is not clearly defined within football. Having a clear and defined language inside a football club leads to better communication and understanding, thus making everything simpler. Coaching is about the transfer of knowledge, with clarity in terms of understanding the coach should be able to transfer knowledge easier to his players.

 With clear language and structure, we can often go further in depth because we are avoiding misinforming and solidifying understanding. This is based loosely on the „Language/Action Perspective“ by Fernando Flores and Terry Winograd from 1987. Their main messages can be summarized as „All Information Is Communication“ and „Language Is Action“; similarly to a more philosophical foundation with the quotes of Wittgenstein “The limits of my language are the limits of my world” (1922) or „The Theory Of Communicative Action“ by Habermas (from 1981). As we know “we can not not communicate” and every output, not just utilizing language, of ourselves as a coach will be leading to an input in our players’ brains.

This process of communication is very important as we are coaching players not only on, but also off the pitch. A more conceptual language can lead to a simpler process in other departments, but within the coaching of players action language should be preferred. Ideally, this can either evolve into two different pathways: 1. An objective language that derives from the game itself and stems from very simple language based on the football actions and thus becoming an “interactional language” where there can be no misconceptions and each action being tied to the actions of the other players on the pitch 2. Creating your own terminology and teach it to the players, where you can assign useful keywords to actions, ideally associated with the preferred volition or emotion to enhance the process of creating a playing style or even a team identity Both approaches are valid and understandable.

Considering that communication is not just linguistic, but – especially within the context of football – is derived from the interaction between the laws of the game and the means of the game (ball, player and opponent) and shares a lot of relational information due to its physical nature owns a non-verbal language of itself, so to say. Whereas nonverbal communication (like posture, mimic, gestures, etc.) in everyday life can differ in detail from culture to culture, football has a nonverbal communication that stems from the laws of the game and the possibilities of the means of the game themselves.

For instance, if you are not able to lift your leg up without losing the ball, you are communicating to everyone on the field that you will not be able to execute a long pass. Especially utilizing the first approach and using football action language for all the possible interactions that emerge non-verbally seems like the most holistic way to go right now. Whilst on the first look this approach might seem simplifying, the infinite amount of possible interactions makes it very complex and there are tools to specify either about what the action language is used for or how it is used.

Problem Description:

A tool that I have used countless times from the WFA is using the Five W’s to describe problems:

  1. What goes wrong in relation to the objective of football?
  2. Where on the field does this take place?
  3. Who played a role in it?
  4. When does the problem occur?
  5. Which specific elements are present that influence the players, the game or the circumstance?

Using this matrix, we are able to define the problem in action terminology and the players are able to understand the problem within their own context on the pitch, which leaves little room for misunderstanding. I think this tool is a fantastic reference point for thinking about football problems from within in a football context. Additionally, the WHY and the HOW of each question help on a deeper level to reflect on both, the problem and the solution. These are connected to the space-time components of every physical action: Position, Moment, Speed & Direction.

To summarise, action language is about saying what you see. To use an example, it is important to clearly define exactly what you are saying to each player. If you as the coach notices that your left-sided centre back is not stepping out to press the ball carrier quick enough, and stop the practice to tell him to be “more aggressive”, have you laid the structure for the player to understand what “aggressive” means in this context, the player may not understand that you are asking him to step out and win the ball higher, he may take this as aggressive in the challenge and may attempt a sliding challenge (thus leaving him on the floor and unable to complete further football actions). It is here that we must be careful in our wording. When talking with players it is important that we generate feedback objectively, that we focus on the decision and it’s space-time components being…

  • Position
  • Direction
  • Speed
  • Moment

These characteristics allow us to guide our thinking and correctly diagnose the problem within a football action context.

Once we explain to the player what we want from them, it is fine to use terms such as “aggressive” because that term has been defined within this coaching context between the player and the coach, up until that point, “aggressive” is just a subjective term leaving much room from misinterpretation.

A rule of thumb to implement here could be, when defining terms for group understanding of the game model, uses verbs rather than adjectives, adjectives can then be used later to embellish once the understanding has been created.

By Author GJ

Practical Examples:

6v3 Rondo

Design of the drill

In this 6v3 rondo the key rule is when the red players lose the ball to the blues. Oftentimes in a rondo one would see players change roles as soon as the defensive players in the minority (in the above case blue) touch the ball. Whilst this might be a simple and enjoyable exercise it completely eradicates transitions from the game. In one way with the rule of changing roles as soon as the ball is touched, we are encouraging undesirable behavior in the transition moments.

Variations through rules and their effect

To mitigate this, a typical constraint in a rondo would be for players to be allowed to leave the defensive role if they manage to dribble out of the field. This means that a transition moment is now added to the game. This implicitly encourages a counterpressing action from the majority as well as on-the-ball actions from the minority to mitigate that.

However, when just having the dribbling rule in the rondo we might create unrealistic counterpressing moments, over-focused on the ball-carrier. Thereby one might discourage principle behavior when it comes to closing down as a unit.

Therefore, to highlight all aspects of a successful counterpress we have created the following rules:

1: Blue wins the ball and pass into one of the 4 mini-goals. One player (the one that passed it into the goals) becomes a possession player whilst the other 2 remain as defenders.

2: Blue wins the ball and dribbles out of the field. 2 players become possessing players (perhaps the dribbling player can pick one to join him or if someone passed him the ball before his dribbling, it automatically becomes that player).

3: Blue wins the ball and make 2 passes between the 3 players. In that case all 3 players can now become possessing players.

Coaching points and their formulation

By joining the above 3 rules together we highlight every aspect of a counterpressing action. Ball-pressure, closing options and avenues int the goals. Making for a more balanced approach in the transition moment from the red team with all players partaking being of crucial importance for success-stability.

In terms of coaching, an approach of highlighting the importance of the different elements of the counterpress via using questions would be favorable. Asking players how they can go about covering the different escaping options together and how that would transfer into a big game in a transition moment can create deeper understanding and thereby a more stable approach.

End to end game 5v5 +6 Neutrals

Design of the drill

In this game the green team play against the green team in the field. Each team looks to transfer the ball from one neutral red team to the other without the opposition intercepting it. Every time this is done successfully a point is awarded, making for a directional positional game.

In the above version there are 3 central corridors. When the ball is transferred to the red player in the same corridor on the other side the successful team earns 1 point. If they transfer the ball from a side channel to the neutral on the other end in the opposite side channel, that is worth 2 points. If they transfer it from a side channel into the neutral in the central channel that is worth 3 points.

Variations through rules and their effect

By putting the above rules in place, one is not restricting the possession teams options of point-scoring, thereby allowing players to make decisions specific to their context and not having an appropriate decision taken away from them at any moment. However, certain behaviors (playing diagonally from outside to the inside as an example) are weighted more significantly thereby tilting decision making in their favor, all whilst not taking away other possibilities.

Another way of constraining the game would be by not allowing neutral players to be tackled and forcing them to stay in a zone outside the main field. Whilst this might make teams have more stable possession and highlight certain pressing triggers it would greatly take away from the game in other ways. For example, dribbling into space, drawing pressure to the release the ball as well as forward defending moments would be taken away that could otherwise provide very desirable situational solutions. Therefore, as mentioned above one would be highlighting parts of the game whilst compromising others. Creating a potentially positive effect in one way whilst creating another problematic one.

Coaching points and their formulation

In this game questions by the coach about when to play straight vertically as opposed to switching lanes makes sense. Asking questions such as where do I play if the pass from red straight to red is covered, should highlight the use of switches of play.

It could also be particularly useful to highlight the importance of playing from the outside to the inside and why follow up actions can often be simpler when doing so. Asking questions such as, where can I pass to when the opposition come and press on the side. Also by allowing the red players to dribble in, one can easily ask them when it makes sense to do so as opposed to just passing it. Finally, the lanes can be used as reference for spacing. What happens if too many players from the possessing team are in one of the 3 lanes? What happens to the oppositional cover when all 3 lanes are occupied by the possessing team.

Counter-attacking game 2v3 into 5v4

Design of the drill

This final game is an adaptation to an exercise witnessed elsewhere. In that exercise 3 player would play a 3v2 rondo in a square around the halfway line. If the 2 would win the ball they were allowed to break out and attack the goal in that half with another attacker. Making for a 3v2 attack towards goal.

The main thought behind this exercise was to highlight counter-attacking with superior numbers towards goal after winning the ball. However, to create more repetition unfavorable behavior was encouraged from the 3 players in the initial rondo. At certain moments they were forced to give up the ball on the coaches’ command and remain passive whilst the 2 defenders broke out of the initial area. Also, if they did lose the ball in the 3v2 they were not allowed to counterpress the lost ball, in order to not compromise attacking repetition.

As stated before, this might create a certain amount of repetition. It however also stops players from performing desired actions when loosing the ball of immediately pressing or defending toward their own goal when that pressing is unsuccessful.

When keeping the said heuristic in mind it is quite simple to modify the said exercise to still have plenty of attacking repetition to goal whilst allowing desirable behavior in transition.

Variations through rules and their effects

The game above starts with white attacking a mini-goal on the halfway-line in a 2v3 (depicted with red attacking white in the first picture above). In a slightly funneled area. If white wins the ball they look to break out toward the bigger playing area. When successful in doing so, they can attack the big goal with their 2 attackers against a further 2 defenders in a 5v4 (depicted in the 2nd picture above). Allowing for a counter-attacking-like scenario with proper defending from the initial players.

In a 2v3 it is possible to score on the mini-goal, if the ball is lost one can win it back and if that is not successful one has to press backwards. For the initial 3 players when they win the ball they can break out of the area via a dribble or a pass into the 2 attackers or a localized combination amongst one another. Linking this exercise closely to the initial rondo in terms of possible outcomes.

In terms of having competition in this particular game after a ball is played on one side of the field with a particular team playing 2v3 and another 5v4, the same would now start on the other side of the pitch with opposite roles. Within this game it makes sense to switch players roles in breaks frequently as well.

Coaching points and their formulation

In this game, questions about how to press in 2v3 initially as well as certain scenarios after winning the ball make sense. One could ask players how to react after playing the ball deep towards the target for the phase of 5v4 on attack. One could question the 2 initial attackers about how they could go about counter-pressing in a 2v3 against them, as well as how that might link with the actions of the 2 defenders in behind. Particularly if they manage to restrict passes between the 3 players that won the ball from them.

By Author MK.


The goal of this article was to give an introduction in different concepts like football actions, football action language, the learning theory behind constraints and possible drill designs including their coaching points but also to open a discussion. Which characteristics should constraints have? How could they look like? Also the main point of this article was to give a very simple heuristic: We should not enforce principles in a way that they compromise other principles.

Basically, we should not hinder our development in order to accelerate our development somewhere else. Not always will that be easily possible, yet what we should keep in our thinking is that we want to constraint actions – we do not want to constrain our principles. The players should be able to follow the principles always and utilize them in the best possible way, be challenged by them – not working against them because our coaching was not able to find a better solution. Constraining actions should support the accelerated learning and utilizing of principles and not being used as a coaching tool for the sake of it.

Brian May 18, 2021 um 11:18 am

This has to be the most important article. Coaches like Pep, Bielsa, Klopp are very vague in their explanations of how to approach training and matches and it’s sad. They, like Kasparov fall into self-preservation where they explain a lot and you still end up understanding nothing. I would recommend that aspiring coaches and players seek out solutions in game situations instead of obsessing on positional play that literally needs billions worth of talented players to pull off and still focus on winning important trophies. It’s not always about using logic all the time, sometimes you just need structure and organisation to develop and maintain players levels and form.
PS maybe formations are not just telephone numbers


Chris May 2, 2020 um 8:28 pm

Hi, thanks for a good article. One question about conceptual vs action based – is there a situation when conceptual is better then action based? I’m thinking about situations like defending against crosses and that a more strict framework can favourable in those special situations. What’s your thoughts?


Mikolas April 13, 2020 um 4:18 pm

You mentioned the difference between reward and restriction rules and that both can be useful at certain times. I can clearly see the advantages of rewarding with eg. double points when the preferrable or desired action is executed while keeping the decision making up to the player, but I cannot imagine a drill or a situation when applying restrictions over rewards could be useful. Could you maybe give an example of such drill and why exactly restrictional approach could be advantageous in the drill? Thanks


Hinterlasse eine Antwort

Your email address will not be published. Required fields are marked *