The kings of two footballing philosophies locked horns in the opening Group C game of Euro 2012: Catenaccio and Tiki-Taka; Italy and Spain.
The Catenaccio philosophy is one associated with a highly organised and effective defensive tactic that focuses on taming the opponent’s attacks and in turn preventing any credible goal scoring opportunities. In Catenaccio, it is typical that the role of a sweeper, libero or verrouilleur is given a pivotal role in both defensive organisation and playing the ball out from the back – a commanding general.
In truth, the profession of a libero or verrouilleur is a dying role in modern-day football tactics – the likes of Franz Beckenbauer, Gaetano Scirea and Franco Baresi are now widely regarded as football antiquity. So eye brows should have been raised when Cesare Prandelli opted to move away from his preferred 4-3-1-2 to an italian suited 3-5-2, made using two wing backs and a libero (Danielle De Rossi), while instructing Juventus attacking midfielder Giaccherini to play out of position on his international début at left wing-back (a player who only four years a go was fighting relegation in Serie C2). Much of the prematch attention should have been directed towards Prandelli’s courage to employ such a tactic against the world champions.
Italy (3-5-2): Buffon, Maggio, Chiellini, Bonucci, Motta, Marchisio, Giaccherini, De Rossi, Pirlo, Balotelli, Cassano
Yet Vincente del Bosque had other ideas. La Furia Roja looked set to take the rule book of formations and simply tear it up, by announcing a team that seemingly was set up to play a 4-6-0 formation:
“When we got into the locker room we were joking around looking at the team sheet, as we couldn’t find a striker” reveals the Azzurri’s newly appointed Libero, Danielle De Rossi.
Guillem Balague (spanish football expert) argued that del Bosque had simply taken inspiration from an already widely praised and accepted Barcelona tactic:
“How many strikers do you think Barcelona use? No strikers means that there’s not one referent point, but five different mobile reference points…set formations are an obsolete way to look at football …centre backs prefer a number 9 due to familiarity of being able to mark them”
Balague went on to claim that del Bosque had set the team up like so, as a tactical response to both Italy’s likelihood of employing a Catenaccio strategy:
“Spain will find a team that will kill spaces. You combat that with dynamism, 2v1’s, short passing and involvement, hence no striker – [this is the] logical progression of style”
Spain looked to be playing to their own strengths, as a response to both their opponent’s set up, their own abilities and as a response to a lack of David Villa. With much of the attack built around a successful Barcelona team and the influence of the Barca-influenced infamous ‘rondo’ it seemed as if it was only a natural progression of formation in a bid to force what they do best: keep ball. The Rondo is essentially a fiercely competitive game of ‘piggy in the middle’ that has become tradition at Barca’s training ground and is often used to intimidate any new arrivals at the club.
However, it seems that there was actually less preparation for this tactic than one would expect, as Javi Matallanas revealed to Guillem Balague that Cesc Fabregas was only made aware of his role as a ‘false number 9’ three hours before kick off.
With a throw-back libero, a false number 9, a Balotelli-Cassano strike partnership, the battle of the regista’s (Xavi and Pirlo) and the clash of the two polar footballing philosophies this match up was beginning to look more and more like it was going to be one of the classics of Euro history.
Spain (4-6-0?): Casillas, Arbeloa, Piqué, Ramos, Alba; Busquets, Xabi Alonso, Xavi; Silva, Iniesta, Fàbregas
The 90 minutes
As expected the spanish dominated possession (65.8%) and played 646 passes to Italy’s 292 passes. Xavi, whom Gary Breen said would “keep the ball off you in a phone box”, was involved in 197 of Spain’s passes (played or received) and created 8 goal scoring opportunities for the striker-less tiki-taka side. Busquets was another player central to La Roja’s possession play as he was involved in 183 passes and boasted a 92% pass rate (the highest of anyone on the field).
Spain’s unnamed formation allowed the spanish to play 172 passes in the final third compared to Italy’s 48, a theme than ran throughout the game.
However, Italy pressed high, boasted a 100% tackling rate (15/15) and provided Spain with a problem as they didn’t begin the game with many players who could run in behind Italy’s back line. The libero role was played well by De Rossi who relieved Pirlo of the play-making pressure, with De Rossi deciding the direction of play more often than not.
Half time came with the score at 0-0, but as one of the most technical and interesting duels of the competition so far.
The second half started, promising more of the same, only this time the first real opportunity came early in the half. A frustrated Mario Balotelli, pressing high up the field managed to steal the ball from Sergio Ramos and began the charge towards Casillas on his goal line. All of Spain and Italy held their breath as Balotelli’s charge towards the goal began to turn into a casual stroll towards Casillas as if a Playstation controller was running out of battery. As Balotelli began to see the world around him in slow motion Ramos had caught up with him to put in a last-ditch tackle. What just happened?!
Prandelli, clearly not happy with Balotelli’s performance and perhaps concerned about the booking Mario had received earlier on in the first half, decided to bring on Di Natale in his place, leaving Balotelli to think over how on earth he hadn’t had a shot moments before.
Up stepped Andrea Pirlo to stake his claim as one of the world’s best – skipping past his Spanish marker on the half way line and changing the pace of the game to feed through Di Natale, who put away his first shot of the game and made his mark immediately.
But the action packed opening to the second half was not finished just yet. Three minutes later Spain managed to get in behind De Rossi with a clever run from Cesc Fabregas who finished cooly – to complete the curse commentator Jon Champion had put on the game where he opened the game with the words “there is no striker, there’s only Fabregas” much like Tim Lovejoy’s cursed words of “Don’t worry, it’s only Ray Parlour”, moments before the midfielder hit an unstoppable shot from 30 yards in the 2002 FA Cup final win against Chelsea.
Di Natale had one more undeniable opportunity with an agonisingly missed volley inside the box from a clever Giovinco ball before the game returned to it’s status quo of Spanish control in the final third, in a duel that looked destined for a testing draw there after – aside from the two missed Torres one on one opportunities to put the game to rest, a ‘nearly-impact’ after his introduction to the spanish front line. The final score Italy 1 Spain 1.
Azzurri Key Performances
Claudio Marchisio, Antonio Cassano, Danielle De Rossi and Gianluigi Buffon all put forward their claim for Azzurri of the match. Iniesta continued to ‘wow’ plaudits for the Spanish, as others put in a nothing more than expected performance from the world champions.
Uefa named Marchisio (9.15) as their man of the match for both teams via their statistical method ahead of followers Fabregas, Iniesta, Xavi, Alonso and Busquets. The Juventus midfielder created balance in a robust midfield trio of Pirlo, himself and Thiago Motta. Marchisio showcased all aspects of his game and was unlucky not to score after unlocking the spanish control in the second half: dribbling through the open space from the half way line towards the spanish box to play a one-two, receive the ball back within the box only to hit a tame shot at Casillas. The Turin-born midfielder also recovered possession well throughout and made a key block to a goal bound Xavi effort.
Pirlo showed glimpses of mesmerising brilliance, including his assist, however was not consistent throughout the game with his passing – hopefully not a sign of tiredness after a long season with Juventus. While Danielle De Rossi had an excellent first half, the introduction of Torres showed a weakness in the player as a defender as Spain found Torres in a good position on more than one occasion:
“I did better in the first half and suffered more against Fernando Torres, who is physically strong and intelligent”
Azzurri Post Match Quotes
Danielle De Rossi: “It was natural to lose a little confidence after the Russia performance and some expected a catastrophe but that wasn’t the case”
Cesare Prandelli: “The important thing is that they did what I asked, playing well when passing vertically and gave our all. This is Italy.”
James Horncastle on Cesare Prandelli: “Nothing but admiration for Prandelli. Humble enough to put own system aside in recognition of what works here and now”
Post match the President of Italy, Napolitano went into in the locker room to congratulate many of the players on an optimistic outcome and embraced Buffon after his performance (http://www.youtube.com/watch?v=VGi0hiAyGy4).
The draw concluded a 14 match winning streak for the Spanish in competitive matches and leaves an air of optimism amongst the italians again after a recent pessimistic outlook on the Azzurri’s chances of winning Euro 2012. Both Italy and Spain showed glimpses of why they should be considered favourites for the tournament, upstaging any of the performances from early favourites in Group B.
While most of the Azzurri came out of the duel playing well, Mario Balotelli put in a below par performance in a promising Balo-Cassa striker force. The only positive to take from the game is that hopefully Balotelli will use that miss as fire for the rest of the tournament presuming Cesare Prandelli provides him with another opportunity in the next game, one that hopefully this time, he’ll take.
Were they lucky? Introduction: Analysis of PDO
Using PDO to analyse how lucky a team may have been in the outcome of a game is a concept @11tegen11 and @jameswgrayson have employed in a bid to “separate the two entities that determine who wins a football match: luck and skill” (@11tegen11). The PDO concept was first used as analysis for Ice Hockey and has revolutionised the way in which teams are assessed.
The PDO is simply a number that determines how lucky each team has been. The PDO is set as a norm at 1000 and always combines with the opponents in a total of 2000 – is calculated as:
[Shooting Percentage (sh%) + Saving Percentage (sv%)] x 10 = PDO
Where sh% is calculated as ‘total number of goals’/’total number of shots’ x 100, and;
sv% is calculated as ‘total number of shots that missed’/‘total number of shots’ x 100
A PDO of +1001 would suggest that the team have been luckier than the norm of 1000 and a PDO of -999 would suggest a team have been unlucky compared to the norm. The combined value must always add up to 2000, since there are two teams each playing each with 1000 norm worth of luck-skill.
For example, take the Denmark v Netherlands game in Group B
Denmark: 1 goal, 8 shots. 32 ‘saves’ from 32 shots.
Netherlands: 0 goals, 32 shots, 7 ‘saves’ from 8.
Denmark PDO = 10 (12.5 + 100) = 1125
Netherlands PDO = 10 (0 + 87.5) = 875
Leaving Denmark with a PDO of 1125 and the Netherlands with a PDO of 875. Meaning that the Netherlands were 12.5% unluckier than the norm, Denmark were 12.5% luckier than the norm and a differential of 28.6% more luck to the Danes. Literally taken, Denmark won their game with a degree of luck and Netherlands lost by being unlucky.
Were they lucky? Conclusion: Italy and Spain PDO
Italy: 1 goal, 11 shots. 18 ‘saves’ from 19 shots
Spain: 1 goal, 19 shots. 10 ‘saves’ from 11 shots
Italy PDO = 10 (9+94.7) = 1038
Spain PDO = 10 (5.3+90.9) = 962
A minor differential of 76, or 38 more/less lucky for each side. Italy were 7.9% luckier than the Spanish.
So to conclude were Italy lucky to draw against Spain? – In short, Yes… but only marginally, in relation to Denmark and the Netherlands game.
If Mario Balotelli had actually managed to get his shot away this would have evened this statistic up further still, providing Casillas had made the resultant save. This would have bought the totals to 972 and 1028 in the same balance of luck and a 5.7% luck differential.
SEE LINKS for a more comprehensive understanding behind the rationale and analysis of PDO on the Dutch Eredivisie and English Premier League using the same analysis method with additional outcomes such as futuristic prediction:
@11tegen11 http://11tegen11.net/?p=1731
@jameswgrayson http://jameswgrayson.wordpress.com
TikiTaka Football www.thepathismadebywalking.wordpress.com
All statistics taken from FourFourTwo Euro 2012 Stats Zone
- Excellent
- Informative
- Awesome
- Good Read
- ok
Categories: Arsenal (NN), Chelsea, Denmark, EPL Index Featured Article, EPL Index Statistical Comparisons, Euro 2012, Italy, Liverpool, Manchester City, Netherlands, Opta Stats, Spain
Tags: 3-5-2, 352, 4-6-0, Assist, Azzurri, Balotelli, Busquets, Cassano, De Rossi, Di Natale, Draw, Euro 2012, Euro 2012 Opta Stats, Euro 2012 Stats, Fabregas, False No 9, football, Giaccherini, Group C, how do the italian's play, how many passes, Iniesta v Italy, Italy, Italy 1 - 1 Spain, Libero, lucky, Marchisio, miss, No Striker, one on one, Pirlo, Prandelli, spain, Spain 1 - 1 Italy, Spain Tactic, Sweeper, they, Tiki Taka, Torres, v Italy, were, What is Catenaccio, Xavi
This article has had 2,740 Views






Great stuff to see more people picking up on what I think should be a break-through stat in football!
PDO can be calculated either by using goal from shots on target or goals from overall shots. Since overall shots are more frequent, I prefer the latter, but in the longer run it won’t make a difference whether you’d use overall shots or shots on target.
Anyway, for the Spain – Italy match the numbers should be this:
Spain: shooting % 1/19 ; saves % 1/11 ; PDO 1000*(0.053+0.909) = 1038
Italy: shooting % 1/11 ; saves % 18/19 ; PDO 1000*(0.091+0.947) = 962
Since one team’s goals are the other team’s unsaved shots, the average of two teams over a single match would always be 1000.
To complete the numbers:
Holland: Sh% 0/32 ; Sv% 7/8 ; PDO 875
Denmark: Sh% 1/8 ; Sv% 32/32 ; PDO 1125
For overall shots aren’t the save percentages incorrect because the only shots that should be counted are the ones that are on-target?
A shot off target is not saved so should not count in the save calculation. I can understand why total shots wish to be used as it does indicate the amount of attacking play from one side however for saves it should only be shots on target. Am I understanding this correctly?
Very interested in using PDO and maybe help getting this out there into the public domain (although it may already be there) because it’s the first I’ve seen of it.
I think perhaps ‘save percentage’ is something that may be misleading if you are going to use the overall shot count. You have to use either overall shots or overall shots on target throughout the calculations. It’s acceptable to use either – 11tegen11 uses overall shots where as JamesWGrayson uses shots on target only.
The calculations above are now 100% correct, however – the variable comes into whether you want to use shots on target/overall shots in the calculation.
I prefer to use overall shots as it seems to give a better overall picture of the game… but like 11tegen11 says in the long run it won’t make a difference… it’s just because this stat has been analysed for one game.
It took me a while to get my head around PDO but once you’ve got it – it all becomes a little clearer.
For those not wishing to understand the calculations and only wanting to see the end result, it works on that front too.
I came across PDO on 11tegen11.net’s analysis of the dutch Eredivisie and I think it gives a remarkable analysis to something that is largely ignored in stats… who was more lucky? A lot of people tend to look at either who had more possession or who had more shots to conclude who deserved to win… so much more to it.
Overall though – I thought this was an excellent article. Thanks for posting it.
thanks Satto – I think the PDO was always going to be a risky analysis to take on – especially over a single game analysis. My strengths have always been tactical analysis rather than statistical – as a football coach. Statistics as a way of ‘proof’ has always intrigued me though, so made a good attempt at it!
PDO for the Barca Vs Chelsea – in the 2-2 (2nd leg):
Barcelona 22 shots, scored 2 goals; 7 shots against, 5 saved.
Chelsea 7 shots, 2 goals; 22 shots against, 20 saved.
therefore,
Barca PDO = 10 (9.1 +71.4) = 805
Chelsea PDO = 10 (28.6+90.9) = 1195
Here’s the issue I find tactically however…as explained by JamesWGrayson –
“imagine this game between team a and b who are equally talented. Team a score early and as the match progresses become more defensive to protect the lead. Team b get more desperate and shoot from further and further from goal, providing relatively easy saves. Team a are heavily outshot but win, they may even score a second on a break away chance late in the game as team b press even more. Team a end up with a huge PDO but in truth were likely worthy of at least a draw anyway. These games happen plenty during the season but tend to do a decent job of cancelling out in the long run”
So over the course of 6 games, whereby a team defends after taking a lead – the PDO analysis isn’t as useful. However in the Italy v Spain game, I guess you could say that Italy didn’t have time to defend their lead as they conceded so quickly. Where this may work on a single game basis is where both teams are going for it…
But over the course of an entire season (as Grayson mentions) these games that involve mass defending are ‘cancelled out’ over the long run.
While I’m not entirely convinced that Luck/Skill can be determined from the two components of sv% and sh%, it’s a damn good start and gives a brilliant insight into Luck to a certain extent. I quite like it as a form of analysis.
But for a game like Chelsea v Barcelona… perhaps its a stat that won’t really tell you much.
As Rodgers says: “I like to control games. I like to be responsible for our own destiny. If you are better than your opponent with the ball you have a 79 per cent chance of winning the game…for me it is quite logical. It doesn’t matter how big or small you are, if you don’t have the ball you can’t score.”
I’d like to know how that percentage was precisely figured out. Probably by retrospective analysing.
I hope someone on this site takes the PDO and puts it to good use… I can see from 11tegen11 and JamesWGraysons work that you can tell quite a lot from it.
Barcelona 5 shots on target, scored 2 goals; 3 shots against on target, 1 saved.
Chelsea 3 shots, 2 goals; 5 shots against, 3 saved.
[USING THE ALTERNATIVE SHOTS ON TARGET]
sh% = goals/S.O.T
sv% = saves/S.O.T
Barca PDO = 10 (40+33.3) = 733 ≈ 730
Chelsea PDO = 10 (66.7+60) = 1267
… the point is, while the results using this method are wider apart – over the course of the season (again) the differences are cancelled out – you just have to be consistent with which you you use (shots on target or total shots)…
I’m confused as to why this calculation actually figures to more than 2000… I’m sure I’ve done the maths right here… maybe someone who uses this method will know
730+1270 = 2000
Yeh but 733 doesn’t and it’s not 730 hence the approx. 733 would mean it was 3 over and the definitive rule is that it should equal 2000….
Ah yes, didn’t see that. But…
10*(66.7+60) = 1267
Ah yes there’s the maths error! Cheers mate.
Im no sure whys it’s x10 to be honest mate that’s something you’ll need to ask they guy in the states who invented this… Its been taken very seriously in I’ve hockey and I’m sure by researching the two guys linked you’ll find more answers.
As I’ve said above I too think there are a number of flaws in this and with its current aim I think it should be using goal scoring opportunities and not shots on target or shots – but then comes the problem of people arguing over what is a goal scoring opportunity and what happens when someone scores from a corner for eg.
It’s not perfect but it’s an interesting attempt at tackling one of the less analysed areas I sport.
I believe that the guy who came up with PDO somehow came to the conclusion that luck of winning is down to goal scoring chances v chances conceded. Which I guess if a team has had lots of Goal scoring opportunities and u have only had one then the other team should have won. But like I said at the moment it focuses on shots and ignores tactical strategies.
The next time I plan on using this I’ll be addressing some of the issues above
But thanks for your input as it shows that it clearly isn’t for everyone and in some cases statistics all together isn’t favoured – Roy hodgson recently saying he doesn’t do stats for eg
Applogies for any errors in this post it’s been posted via my mobile.
Not* and *ice
Shots on target is better. When I checked this a while back, SPG explained 47 % of the variation in the number of points per team in the EPL. The explanatory power of possession and shots conceded was similar. SOTPG explained 65 % of the variation. I also looked at 46 Arsenal games this season, and found that shots on target minus shots on target conceded explained 26 % of the variation in results. If you rather look at shots, that explained 20 % of the variation. I suspect looking at clear-cut chances would be even better. It’s about quality, not quantity.
Great to hear about more research done at great depths – have you got a blog that you’ve published anything on by any chance ? Would love to have a look at some of this
There are so many errors in this post it is unreal. How can you be 7.9% luckier, you mean 7.9% more efficient. Your PDO is 0%!!
bit harsh isn’t it? ok, in hindsight i may have used the wrong terminology… but being new to PDO I wasn’t expecting it to be perfect. I used the word ‘luckier’ purely to emphasise that the PDO was to measure luck
and regards to errors – all stats were taken from fourfourtwo’s stat zone… and every stat database varies. if you’d like to point out any mistakes, i’d be extremely grateful
Some excellent comments here, I have seen PDO in hockey for a good few years now, I have tracked it for 2 seasons now in football and a post on 2011/12 EPL pdo is written and scheduled for publish on Tuesday (had to wait post Eng-Fra) on Bitter and Blue.
Is it purely luck if a team has an excellent PDO? sometimes, but not really. Man City had league’s best with 116.2 (don’t know why hockey doesn’t use the decimal but I am). From 1-20 league position PDO tends to decrease with uniformity save a couple of teams,
Arsenal 93.8 with a terrible save% despite a low number of shots against on target per game
Liverpool 92.4, league average save % but 0.224 scoring%
Stoke with an EPL status saving 109.0 PDO, league’s best scoring % at 0.382 and a save % of 0.708. Stoke were the 3rd worst team in the EPL regards sotpg +/-
I agree with frode shots on target all the way. why even use shots off target unless you want to indicate how inefficient a team is in terms of total shots/shots on target. I see no value in working out goals from total shots metric unless it is to be used in a negatively reinforcing way.
Amazing stuff! I look forward to reading your article! Would have thought Liverpools sh% would pull their PDO down – always seem to get the opportunities and seem unlucky not to score more!
Something Frode said above about measuring clear cut chances would be interesting to look in to as there still seems like a lot of ‘hidden’ information using shots on target or shots – as neither inform you if either was a goal scoring opportunity in truth,
Look forward to your article any how. Will b sure to ask a few questions !
Would be great if we could distribute that post here too in a week or so?
I admire the attempt at some form of statistical football analysis but this “PDO” has to be one of the silliest things I have ever seen. What it suggests is that the only difference between a shot on target and a goal is “luck”. If you are going to “separate the two entities that determine who wins a football match: luck and skill”, I think you first need to find a better way to define “luck”.
On a more pedantic note, why on earth is there a superfluous factor of 10 in the equation?
Not* and Ice*