How to play against Pep Guardiola’s Barcelona

(yeah, I am not an expert, and this is more of a statistical/data survey than a tactical masterpiece)

From 2008-09 to 2011-12, FC Barcelona made the entire world sit up and take notice with some of the best football played ever by any club. Coached by Pep Guardiola, a Barcelona legend of the Johan Cruyff Dream Team glory, the team put on display the results of some of the finest coaching philosophies ever seen. Possession-dominating, ruthlessly attacking, high pressing, superb display of JdP principles – one can wax eloquent all day with all sorts of superlatives to describe the juggernaut that the team became under Guardiola’s coaching. Barcelona won 27 games in La Liga 2008-09, 31 in 2009-10, 30 in 2010-11 and 28 in 2011-12, with a win percentage of 76% (just La Liga)!! Pep Guardiola’s achievements cannot be understated: he won the league on the first three occasions and the UCL in 2008-09 and 2010-11. These are astounding feats. Coupled with his win percentage, it is extremely fair to say that the opposition teams really did not stand much chance – Barcelona swatted aside all opposition in an utterly dominant period and with impunity. So, how do you even go up to face one of the greatest coaches in modern footballing history?

Well, the short answer is: one has to be pragmatic when facing such a team. Barcelona had some of the most press-resistant players in Xavi Hernandez, Andres Iniesta and Sergio Busquets, possibly the greatest player of all time in Lionel Messi, and one of the greatest full backs of all time in Dani Alves. The team strength and quality was absurdly good every year, and the coaching made these stars even better collectively. One cannot expect to go toe-to-toe with this team in an expansive attacking sense – if you set up your team to try and high press them, you would probably not succeed(unless you are Marcelo Bielsa), if you fail to be compact while defending, your team will be ripped apart. One has to do extremely careful planning based on the weaknesses of Guardiola’s team. Yes, there were weaknesses, like any other team – let’s take a look at some of them. All of the data used in the following sections come via Statsbomb’s open data collection, in particular the Lionel Messi data biography. They have publicly released the data for all games in which Messi played a role – be it as a starter or as a substitute. In 2008-09, there were 31 such games. There were 35 in 2009-10, 33 in 2010-11 and 37 in 2011-12.

Offence

Let’s begin by looking at the shots and goals conceded by Barcelona in all those years. The following shotmaps show that the opposition teams ended up taking a larger fraction of shots from the left as compared to the right. The xG heatmaps, which show zones of the pitch from where the highest quality of shots were taken, also show a slight skew towards the left. While the amount of left-right skew varies from season to season – less pronounced in some and clearer in the others – the pattern is indeed defined. To be noted, this only looks at goals coming from open play. All freekicks and penalties have been removed from consideration.

The left-skewedness of the number and quality of chances conceded are probably explained by the back that the predominant right back of Barcelona was Dani Alves. Alves was an excellent fullback, but being super-attacking in nature, tended to leave space behind him. On the odd occasions where the opposition got to attack – didn’t happen too frequently – the left side was the preferred side.

Interesting patterns can be seen in a variety of other attacking statistics, like key passes and assists, or cut backs from wide areas, or threatening passes in general. Let’s have a look at each of them. First, we look at the heatmaps of key passes conceded:

The end location heatmaps are of no surprise: the biggest chunk of these passes ended in and around the box. The start location heatmaps are the more interesting ones. Seasons 2008-09 and 2010-11 show the biggest chunk of key passes coming from the zone left behind by Alves while attacking. 2009-10 is a curious one, with lots of central attacks. 2011-12 flips the skew and shows a higher fraction originating from the wide areas on the right.

The through balls and cutbacks say a similar story. Most of the cutbacks originate from Alves’ side. 2008-09 sees a lot of through balls through the centre but also through the wide areas or targeting the wide areas. In contrast, in 2009-10, most through balls are down the central zones.

Instead of looking at such specific passes, we can also look at all passes, and all ball carries. Karun Singh, in his wonderful article here(https://karun.in/blog/expected-threat.html), introduced the idea of threat values of such passes and carries. Briefly, the idea is to look at the probability of taking a shot and scoring as soon as possible after passing or carrying the ball. Based on passes/carries originating from different zones, they get classified as passes/carries of different threat values depending on how soon they can lead to a shot. As a basic estimate, the closer to goal a pass is made, the higher its threat is. Here is a heatmap of the starting locations of all passes and carries made by Barcelona’s opposition, weighted by the total threat value they produced:

This heatmap also reinforces some of the patterns noticed earlier: 2008-09 and 2010-11 conceded high threat passes and ball carries from the left wide zones. In 2011-12, the right side conceded most threats. In 2009-10, the threats came mostly from central zones or middle third of the pitch.

So, the biggest takeaway from this section is:

If you are up against 2008-09 Barcelona, the best zone to target for attacks is the space left behind by Alves. Same holds true for 2010-11. In 2011-12, you would be better off attacking from the right side. If you are facing the 2009-10 side, you would have loads of opportunities to counter through the center, as we will see next.

Goal and Shot types

Targeting the unguarded zones is not enough against Guardiola’s Barcelona. To score, one has to target all possible sorts of weaknesses present in the team. Here we are going to take a look at the variety of shots and goals that were conceded by Barcelona under Guardiola. The focus will be mainly on goals coming from set piece play patterns like corners and freekicks, and also from counter attacks. First up, shots and goals from the 2008-09 season:

Not a lot counter attacking shots were conceded, and zero goals resulted from those few shots. However, 11 of the 29 goals came from set pieces – that’s 38% of all goals coming from set-piece situations!! Moving on, next we have the 2009-10 season:

Once again, 7 of the 21 goals – i.e., a third of the lot - came from set pieces. That’s an interesting trend emerging. But perhaps more telling is the fact that there were larger number of counter attacking shots conceded per game – 20 in 35 as opposed to 10 in 31 in 2008-09. And a whopping 25% of such counter-attacking shots resulted in goals. The counter-attacking goals also took up ~25% of all goals conceded.

Next up, season 2010-11:

The number of counter attacking shots and goals go down again (7 from 33 games, only 1 goal). However, the number of set piece goals conceded is 7 out of 18 – 39%, the highest so far.

Finally, we have the 2011-12 season:

The fraction of counters is slightly higher(21 in 37 games) and 3 goals were conceded from counters. 7 out of the 26 goals conceded came from set pieces – the lowest percentage of all seasons.

So, we do have a very clear story here. Across all seasons under Guardiola, Barcelona suffered from set pieces. The biggest chunk of goals conceded came from play patterns originating from free kicks or corners. 2008-09 and 2010-11 saw the highest fraction of such goals scored.

On the other hands, the team in 2009-10 conceded a lot of counters as compared to the other seasons. There was moderate amount of success enjoyed by the opposition in 2011-12 as well when they counter-attacked. This information is useful in squad building to face Guardiola’s Barcelona. You would preferably want to have strong runners/ball carriers to lead counter attacks, like Real Madrid under Jose Mourinho. You would also want to train your team on set pieces, because that’s the best opportunity you will get to score from. Having set piece delivery experts as well as good headers of the ball seem to be crucial.

Defence

While defending against Guardiola’s Barcelona, the key is not to get dragged apart. Compact shapes in the center of the pitch is crucial to holding Barcelona’s holy trinity in the midfield at bay. Also, pressing too high up the pitch will lead to disaster in all probability, so defending and pressing mainly in the middle and defensive thirds is the way to go. As a case study, we will focus on all the El Classicos played in La Liga across these 4 seasons. First, we take a look at the pressure applied by Real Madrid in these games:

There are two things that emerge as broad patterns here: firstly, whenever Real Madrid pressed in relatively deeper zones than higher up the pitch, they enjoyed more success or at least respectable scorelines in defeat. Secondly, whenever Madrid played at home, they tried to play a more expansive game, perhaps in trying to please their fans, and suffered most of the times(2011 being the only odd one). For eg, in the 2011-12 Classicos, Madrid applied 52 percent of their total pressure in the middle third and 16% in the final third in the away victory at Camp Nou. In contrast, they went for a slightly higher press in the home game at Santiao Bernabeu, with 57% in the middle third and 20% in the final third, and lost 3-1.

In 2010-11, Madrid racked up 77% of their pressing in the middle and final thirds and got whipped with a Manita. They toned it down slightly to 74% in the return fixture and managed a draw.

In terms of results, particularly stark is the 2008-09 season. In the 6-2 drubbing received at the shell-shocked Bernabeu, Madrid attempted 73% of their pressures in the middle and final thirds. Having learnt the lesson, they only attempted 48% pressures in the middle and final thirds in the reverse fixture and lost by a much respectable margin of 2-0. Defending deep did not necessarily prevent Madrid from losing games, but at least they didn’t lose in devastating fashions in those games.

Let’s have a deeper and more detailed look into the 2011-12 Classicos. Here are the more detailed/grainier versions of the pressure heatmaps of the two games:

In the away game at Camp Nou, Madrid restricted most of their pressing activities to central zones, not letting their compact shape be disrupted easily. In contrast, in the home game at Bernabeu, they tried to press all over the pitch. In other words, they allowed themselves to get stretched apart, lost their defensive structure and lost the game.

More can be inferred from these games. Statsperform(previously Opta) defines a quantity called pitch tilt, defined as the fraction of final third passes a team makes divided by the sum of final third passes made by both teams(https://www.statsperform.com/resource/how-we-measure-pressure/). This gives us an understanding of the territory gained by teams with their passing. How to use this quantity? In conjunction with possession percentage. A team can have a lion’s share of the possession, but if they don’t have a pitch tilt, they didn’t gain a lot of territory with their passing – they were basically restricted to passing around meaninglessly in the defensive and middle thirds. And that’s good from the opposition perspective. You would prefer to make the possession team stay away from your defensive third as much as possible. Let’s look at the numbers from the 2011-12 Classicos again. Barcelona had a lion’s share of possession as expected – 65% to 35% at Camp Nou, 57% to 43% at Bernabeu. However, the pitch tilt in both games was roughly 57% to 43% in Barcelona’s favor. Even though Barcelona enjoyed an increase in possession share at Camp Nou, they couldn’t increase their territorial advantage. And even though they were still better at entering the opposition third, they could not take full advantage of it. Not to mention, aimless possession in the middle of the pitch did not add to their cause.

Ok, so we have sort of firmly established that pressing Barcelona high up the pitch and getting stretched in the process is not a good idea. Restricting their entry in the final third as compared to overall ball possession is useful. However, in the off chance that you do press them high, where should you do it?

Here is a plot of all the high turnovers Barcelona conceded to their opposition:

A couple of things immediately pop up: the most success you would probably have in 2008-09 is pressing the left side. That pattern sort of holds true in 2010-11 and 2011-12(a bit more equal spread in those two seasons, but still slightly left-skewed). Also, the highest fraction of turnovers leading to shots came in 2009-10(44%), followed by 2011-12(38%), 2010-11(36%) and finally 2008-09(33%). So, if you do choose to press high, you will enjoy the most against the 2009-10 side.

Conclusion

To summarize all the ideas, facing against Guardiola’s Barcelona is an extremely daunting task. They will probably win(76% chance). One has to go in with a slightly pragmatic mindset. Off the ball, the team needs to stay compact, and preferably press in the deeper and central zones. In possession, counter attacks against a Barcelona team with high line, or set pieces, are good avenues to score from. The team should be built around fast, technical and good ball carriers and runners to counter, and train specifically for set pieces. The defence and the midfield need to be disciplined to stay compact and defend deep. In my opinion, this is not a visually appealing way to play, but special games do require special measures if the objective is to come away with a respectable or even a positive result. Barcelona under Guardiola were special beyond words, and it will require a tactician of the highest order (the likes of Mourinho or Bielsa) to even put a scratch on that team, leave alone denting them.