Tiki-taka, Numbers & Football

As another season of amateur football draws to a close here in NSW, most players, teams and clubs are glaring at their division points tables to see what needs to be done to the numbers to ensure a finals spot, or to avoid relegation. As they do, I can’t help but feel these numbers give rise to a very ugly and very useless style of football, the “Kick and Chase” game. Worse yet, I see clubs proudly parading teams that employ this style if it achieves the desired result. I contend that this style of football not only has no effect in determining a team’s table ranking, but that it has no long term worth in football. Let’s look at some real numbers to see why.

First though, let’s define the only two styles of football that are played in the amateur leagues:

  • Kick and Chase
  • Passing

The kick and chase style can be seen as an unrefined version of a counterattacking style (e.g. Italy), while the passing style can be seen as an unrefined version of the Tiki-taka style (e.g. Spain). Regardless of any other stylistic quirks, it is clear that the major characteristic of both styles is the percentage of possession employed by each. Tiki-taka aims for maximum possession, while counterattacking purposefully employs a low percentage of possession. Proponents of either style would claim the validity of that style based on the results of their team. However, I suggest that only the passing style can claim their results reflect the value of their team, and that the results of the kick and chase team only reflect the value of their opponents.

Greater Possession = Greater results

Let’s take an extreme example to look at the connection between possession and results. If you have 0% possession during an entire game, you would of course have a 0% chance of winning , but you may lose. If you maintain 100% possession during a game, you would have a 0% chance of losing, but you may win. At the very least, we can say that as possession increases (from 0% to 100%) the chances of a more favourable result increase (from definitely not winning to definitely not losing).

Now, I am aware that the chances of a favourable result may not increase linearly. That is to say, if you have 20% possession during a game then you do not necessarily have a 20% chance of a favourable result, just as having 80% possession does not necessarily mean you have an 80% chance of a favourable result. The only numbers we can be certain about are 0% and 100%, neither of which is likely to ever occur in football.

So why then should we bother to maintain possession? Well, to win a match you must score at least one goal. To score a goal you must have at least some possession. Let’s say that any team has a percentage chance to score a goal each minute they retain possession. We don’t know what this chance is, but it’s fair to a make a general claim that they do have some chance. If this is the case, then the longer any team retains possession the more likely it is that they will score a goal. Let’s say a team has a 50% chance to score every minute they retain possession. In the first minute, you have a 50% to have scored at least one goal. In the second you have a 75% chance to have scored at least one goal. The third, an 87.5% chance. While it may not be possible to determine what percentage chance a team actually has to score while in possession, we can at least say with certainty that the more possession a team has, the more likely they are to have scored a goal.

So now we know two things about possession. The more you have, the less likely you are to lose a match and the more likely you are to score a goal. These seem like good things for any football team, but how do they specifically relate to passing and counterattacking football styles?

Controlling your own destiny

We’ve talked a lot about possession, but how does a team actually acquire it? It might be too hard to determine how a team specifically wins possession, but we can say that in the strictest sense, a team has possession when the opposing team loses it.

Now consider what teams do with possession. When a kick and chase team obtains possession, they immediately and intentionally give it up by kicking the ball as close as possible to their opponents goal, hoping it will be regained by their team. When a passing team obtains possession, they actively work to retain it through further passing. Given that possession is only gained by a team when it is lost by their opponent, a passing team effectively dictates its level of possession while a kick and chase team has its level of possession dictated by its opposition.

We have already established that possession directly affects the likelihood of scoring and the chances of a more favourable result. We have also established that passing teams dictate their level of possession while kick and chase teams have theirs dictated by their opponents. Given this information, we can say by extension that passing teams dictate their chances of scoring and obtaining a more favourable result while teams that play kick and chase do not.

In short, passing teams control their destiny, while kick and chase teams leave their results to fate.

The illusion of results

But wait, why do so many teams play kick and chase and seem to get good results? How often do we seen teams in professional or amateur league resort to kick and chase football to try and stave off relegation? There must be some effectiveness to this strategy right?

The short answer is no, the effectiveness is a complete illusion, but let’s take a more detailed look at why with the following example:

A competition is played with 10 teams

  • Team A plays passing football very well such that they win 99.9% of their games
  • Team B plays passing football very poorly such they win 10% of their games
  • Team C plays kick and chase. Their possession, and therefore their results, are dictated by their opposition
  • All other teams are listed as X and we have no data on how they play or how good they are

At the end of the season, the points table is as follows:

1st: A
2nd: X
3rd: X
4th: C
———————
5th: X
6th: X
7th: X
8th: X
9th: X
10th: B

Team A has come first, team C has obtained a finals spot by finishing in the top 4 and team B has finished last. Given the styles of play, and the impact of possession on results what can we say about how each team has performed?

strong has directly controlled their results and won 99.9% of their games through possession. They are a quality team and deserve to be in first. If you were to guess how they might perform against a random team, you could reasonably guess that they would win.??Team B has directly controlled their results but has lost the majority of their games because they are bad at retaining possession. If you were to guess how they might perform against a random team, you could reasonably guess they would lose.

About Team C, most interestingly, we can say absolutely nothing, because their results are entirely dictated by their opponents. If you were to guess how they might perform against a random team, you would have no idea as it would be dependent on the team.

You might think that coming 4th is a decent result, but lets consider how this result was achieved. Team C played against team A who has a 99.9% chance of winning, and of course lost. They played against team B, who has a 10% chance of winning, and of course won. They also played against all other teams, and on average came 4th. So where as team A can say they came first because they are the best team, team C can only say they came 4th because on average all other teams won and lost such that team C was ranked 4th.

Imagine now, that team C has to play team A in the finals of this competition. When team C plays matches against the whole competition they average out as a 4th place team. However the final is just a one off game, and as team C’s chances of winning are dictated by their opposition they have only a %0.01 chance to win the final. In other words, they will never beat team A.

So in the strictest sense, when winning the competition is the only result that matters, we can say that team B did no worse then team C. However, there is one major difference between the teams. If team B improves its ability to retain possession through practice, it will have a better chance to win the competition. However, team C, playing the kick and chase style cannot influence its chances to win the competition at all.

Play to win

So let’s do a quick recap:

  • Possession directly influences the chances of a favourable result
  • Possession directly influences the chances of scoring
  • Passing football actively works to increase possession
  • Passing football dictates a team’s quality
  • The quality of a kick and chase team is dictated by its opposition

Once we know all of this, we know what a joke it is to think that playing kick and chase football would ever be a wise decision. If it looks like playing kick and chase football has saved your team from relegation, or has earned you a finals spot, look again. You will see that all that has happened is that enough teams in your competition have failed to play with the quality required to relegate you or prevent you from making the finals.

If you want to play kick and chase, you may as well be playing the football lottery.

If you want to play to win, you better start playing a passing style.

Isn’t it a lot more fun to play anyway?

8 Comments

  1. Interesting Bayesian inference you have going on here.

    Some realistic numbers for your example; based on the end of season wins-per-games-played in the top 5 grades of our competition (http://www.shirefootball.com/), the 1st placed team will not win (at least) 25% of the time against a kick-and-chase team… interesting.

  2. JB – I think there is merit in what you are saying but it seems to me that this argument, or ideology, in its reduced/simplified form (necessary for getting the point across, I know) overlooks the defensive element of the passing and kick/chase styles.

    If I compare to AFL – and of course jumping between codes never makes a perfect analogy – I might be able to explain what I mean.

    For the most part the most successful teams in AFL are those that tend to kick long and down the line and avoid over-possessing the football with unnecessary short kicks or handballs. Not only is it preferable to kick long toward your attacking half in order to produce chances to score, but by not ‘buggerising around the flanks’, you reduce the opportunity for error and the possibility of affording the opposition a costly turnover.

    I guess what I am saying is that you have to recognise, or add in to your argument/formula, the need to be defensively sound and ultimately extremely precise with a passing game that can involve switching the ball back into dangerous positions. And at the same time acknowledge that there are ‘defensive benefits’ so to speak, by attacking with a long kick approach (if the ball’s nowhere near your goals then they can’t score).

    • I’ll concede that greater defensive pressure makes it a lot harder for a team to retain possession. That being said, the article is not about how much skill any specific team has in pressuring or withstanding pressure – only that possession yeilds greater results. If you’re suggesting that a team is as unlikely to retain posession through short passing as long passing, then my logic would not apply. However, my experience in football has taught me that a direct short pass, that is not through other players >IS< more likely to be received than a long ball. This is mainly due to the time it takes a long ball to reach your intended target, and therefor the time an opponent has to react. I would note that as AGL is a much rougher game, and passing is generall much longer (even short passing) than in football, it is not appropriate to compare.

      That all being said – who knows, maybe if an AFL team worked insanely hard to play short passing – they might actually dominate the competition. People might have doubted it in football too, until barcelona showed unparralled domination.

  3. I just think Barca and Spain have to be considered somewhat as exceptions. Other teams without the same level of skill (ie. nearly everyone) cannot maintain the same fidelity to a tiki taka ideology. In some ways I share the wariness of Roy Hodgson touched upon here: http://soccernet.espn.go.com/blog/_/name/espnfcunited/id/412?cc=3436 (I also found the mention of his ‘inverse wingers’ interesting).

    In AFL there is probably one team, Hawthorn, who could be most closely compared to the high possession soccer-style. They are known for short, and accurate, foot passing (interestingly, they have an unusually high proportion of left-footers). They are also probably the best team at the moment and most likely to win the premiership. However, you just have to watch some of the crapper teams with young players, eg Port Adelaide or GWS, attempt to over-possess the ball and produce countless turnovers, to see the high level of skill required and risk involved.

  4. JB, this is a fairly shallow analysis. As Rositz suggests, possession is just one aspect of the equation. There is little doubt that possession and field position are both influential factors in determining the potential of a team to score.

    A crude, but more appropriate, equation would be:

    Possession x Field Position = Goal Scoring Potential

    I’m sure you would agree that a player possessing the ball on his own goal line has a negligible chance of scoring with his next kick. Similarly, as a player approaches the goal, the chance of that player scoring certainly increases (at least linearly; probably exponentially).

    Certainly, as you say, if a team has 0% of possession, they have 0% chance of winning the game. Now, consider the following; any time the ball is in the opposition half, that represents a positive percentage of field position. Is it not also true that if a team has 0% field position (i.e. the ball does not enter the opposition half), they also have 0% chance of winning (irrespective of possession)?

    In contrast to Tiki-taka, a kick and chase approach simply attempts to modify the other variable in the aforementioned equation (i.e. field position). In your analysis, you assume that, by kicking long, the defending team will always be handed possession – this is not the case. I will concede that, in most cases, the defending team is more likely to gain possession; but kick-and-chase tactics acknowledge this:

    Let’s assume that in any given kick-and-chase situation, the defending team has an 80% chance of possession and the attacking team a 20% chance of possession. If the ball has been kicked deep into opposition territory, the defending team has a high-likelihood of gaining possession, but there is little chance of executing a scoring play. The attacking team, on the other hand, has a relatively small chance of gaining possession, but a high likelihood of scoring if they do. Moreover, the attacking team understands that, even if they do not score, there is a reasonable chance of earning an attacking set-piece (i.e. corner, penalty or throw in).

    In summary, the kick-and-chase tactic, whilst not pretty, can be an effective strategy, provided that the team utilising it understands the dynamics of this approach (as outlined above).

    • No. I don’t agree that its a shallow analysis. It’s just a mathematical analysis using booleans to prove a point – greater possession = more favourable results. The purposeful reason I don’t mention things like field position and player skill and how fit you are and how hot it is and a myriad of other impossible things to calculate is because there would be no point in arguing something so ridiculously subjective. Your equations of possession x field position = goal scoring potential is this kind of subjective. What is “field position” how is it determined? What values are you giving it. What about goal scoring potential. What does that mean? Is it a percentage and does it translate over time, 1 second, 1 minute? You see, it’s not a foolish equation, its just impossible to say anything about. That’s the entire reason I used boolean cases as those are the only cases we can concretely say things about, and then we can extend a continuum from there. As mentioned even I couldn’t say how the continuum developed, only that on approach to 100% you increased the likelihood of a favourable result. I purposefully avoided the term winning in the initial example also.

      Also I don’t assume that when kicking long, the other team will gain possession, just that you will lose it – in the interim at least. I cannot account for who gains the ball back in a 50/50 and wouldn’t try. But you can say that instantly losing it, as opposed to deliberately retaining is different.

      And finally, I didn’t say kick and chase wasn’t effective – only that your results with it were not representative of your teams skill. Which I stand by.

  5. I like your thinking but I think we forgot one parameter – rich Russian owners!

    Just kidding but you could argue the more wealth a team has – the more technically proficient players they could attract – which would mean they are capable of successfully passing and also keeping possession!

    There is a great film called Money ball about the use of statistics and how it changed the face of baseball! i.e a poor player who can get on base is all that matters!

Leave a Reply

Your email address will not be published.


*