Tactics recap: Juventus 2-2 Chelsea

Most journalists and commentators will likely say this evening's 2-2 Champions League draw between Juventus and Chelsea was a fair result. In truth, the west Londoners were probably a bit fortunate to emerge with the home point on the balance play. Chelsea opened a two goal first half lead through a pair of Oscar goals inside two minutes of each other. However, after drawing a goal back late in the first half through Arturo Vidal, the Italian side dominated the second and although they finished the game with fewer shots on goal, they generally looked more threatening going forward. Chelsea looked vulnerable to balls played in behind the back four throughout the evening and ultimately conceded the equalizer when Fabio Quagliarella was able to time his run and finish a break away to complete the two goal comeback.

Both teams lined up in their usual formations- Juventus set out in a 3-5-2 while Chelsea opted for a 4-2-3-1. Chelsea, reacting to the creative threat of Andrea Pirlo from the base of the Juventus midfield, made one significant lineup change we've not seen from the Blues thus far. Oscar was brought in to play centrally just behind Torres, making his first Chelsea start, while Eden Hazard was shifted to the left. Oscar is a better tackler than Hazard and more used to being given some defensive responsibility. Roberto Di Matteo clearly told the young Brazilian to remain close to Pirlo to deny Juventus the distribution channel of their most clever passer. Oscar performed the task well. Pirlo never really had a huge impact on the game.

With Juventus's narrow center back three, Chelsea were able to enjoy some success in the first half when Hazard and Ramires were able to get the ball in the channels behind Juve's wing backs Stephan Lichtsteiner and Kwaswo Asamoah. When the wing backs were able to track back Ramires and Hazard, the Chelsea wide players were able to drift inside and create space for unmarked overlapping runs from outside backs Ashley Cole and Branislav Ivanovic. The problems for Chelsea came when they got the ball in these areas and needed to find a penetrating pass. All three of Juventus's center backs were able to stay narrow near their 18 yard box and usually had only Fernando Torres to worry about defending. Asking Torres to get on the end of a cross or through ball in a 1 v. 3 situation was a difficult ask and the Spaniard had a frustrating night. Oscar's goals were from solid individual efforts but weren't attributable to Chelsea's tactics.

With Pirlo largely taken out of the game by Oscar, Juventus were forced to get their attacks started elsewhere. They enjoyed success from two main channels. With two forwards, Juve were able to occupy both Chelsea center backs. Mirko Vucinic would check back deeper to the ball, forcing one Chelsea center back to follow him and allowing Sebastian Giovinco to move laterally into the space created Vucinic. Giovinco had a poor game and was ultimately subbed for Quaglierella. Quaglierella continued making these lateral runs into space and ultimately got in behind Chelsea's defense and converted a 1 v. 1 with Cech. Juve's other main attacking threat came from their more advanced center midfielders Vidal and Claudio Marchisio. The two were able to make darting runs into towards the 18 that Frank Lampard and John Obi Mikel had a nightmare of a time dealing with. The combined well for Juventus's opening goal.

Chelsea were also particularly poor at keeping possession in the second half. They wanted to sit deep and keep possession to kill the clock off but through a combination of Juventus's higher pressing and their own sloppiness they struggled to ever dictate the tempo. Mikel's giveaway that led to Juve's second goal is noteworthy but he was hardly the only one guilty of squandering possession.

The Italians will be more pleased with the away point but they were the more impressive side throughout the bulk of the match.

Thoughts on USA 1-0 Jamaica

The US looked impressive in the first half of their crucial 1-0 win over Jamaica last night in World Cup qualifying and kept a good enough defensive shape in a nervy final half hour to hold on for the three points. Despite failing to score in the opening 45 minutes, the US dominated possession with 79 percent and created a number of fine scoring opportunities. They struck the woodwork three times and forced decent saves from Jamaican goalkeeper Dwayne Miller. The breakthrough came via a Herculez Gomez freekick in the 55th minute. In truth Miller should have made the save, but the US had been good value for their lead.

Jurgen Klinsmann made five changes to the side that started in Kingston Friday evening. Right back Steve Cherundolo returned from a calf strain to replace Michael Parkhurst, Carlos Bocanegra replaced Clarence Goodson at center back, and Danny Williams, Graham Zusi and Jose Torres were brought into the midfield while Kyle Beckerman and Maurice Edu sat. Dempsey was moved forward to a withdrawn forward position behind Herculez Gomez, demoting Jozy Altidore to the bench.

The US played something like a 4-1-3-1 with Williams playing as a holding midfielder behind Zusi, Jones, and Torres and Dempsey operating just behind Gomez. Zusi brought width and a direct vertical threat down the right, while Torres brought composed passing on the left. He also frequently drifted inside to offer an additional passing option for Williams and Jones. Both outside backs in Cherundolo and Fabian Johnson did a fine job overlapping into space when Torres and Zusi came inside. The US looked particularly threatening down the right side with combinations between Zusi and Cherundolo, though the Hanover 96 captain struggled with his final ball.

Jamaica's defense sat particularly deep, often keeping all four defenders within 12 yards of the goal when the US were in the final third. Their deep line created a great deal of space for the US just outside the 18 and the Americans looked dangerous when they made delayed runs into this area (recall Danny Williams' strike that struck the post and Zusi's volley blasted over the bar came from this area). Their movement was fluid and they were able to penetrate gaps in the Jamaican defense with relative ease. That they failed to score despite such an overwhelmingly dominant first half came down to a combination of good goalkeeping, bad luck, and poor finishing.

Perhaps fittingly, the US's revenge came in the form of a freekick, the area of the game the Jamaicans had executed so well to shock the Americans in Kingston. After going down a goal, Jamaica committed more men forward and pressed the US higher up the field. In response, the US began playing much more direct. Faced with the threat of the pacey Jamaicans aggressively closing in on the ball, Klinsmann could be seen on the sideline frantically urging his back four and midfielders to knock the ball long into the corners.

This wasn't a time-wasting strategy the US was employing to kill the clock off; they were doing it with 25 minutes still remaining. Given Klinsmann expressed desire to Latinize American soccer and the success the US had in possession in the first half (they finished the half with 79 percent possession), it seemed a little surprising that Klinsmann would resort to hitting long balls for the forward to chase while keeping a tidy defensive shape with the other players.

However, the strategy made perfect sense. At no point in the course of three halves of soccer had Jamaica shown they could break down a compact US defense in the run of play. They simply don't have the technically ability to break down a team defending with two banks of four. By playing long balls into Jamaica's defensive third, Klinsmann was minimizing the ability of the Jamaicans to force turnovers in midfield and counter at the US defense with numbers. Even if Herculez Gomez (and then Jozy Altidore) was unable to get on the end of these long balls, the US was still forcing Jamaica to patiently build from the back against a crowded defense. In effect, they were forcing Jamaica to rely on their technical ability rather than their athleticism to get an equalizer. Despite a nervous-looking finish, the Jamaicans only really troubled Howard once.

The decision of Klinsmann to move to a more direct and defensive setup after getting the go-ahead goal also signals that he's not entirely confident in his team's ability to kill off games by knocking the ball around and preventing the opponent from getting possession (the way a team like Spain would see out an important game). He was clearly concerned about giving the ball away cheaply in midfield and giving the Jamaicans opportunities to counter. That's a concern I shared and given the US's precarious position in the group going into the game, minimizing risk seemed a wise move.

A conversation on US tactics under Klinsmann and Bradley

Between the U.S. Men's National Team's away  loss to Jamaica last Friday (September 7) and tonight's match versus Jamaica in Columbus, Ohio, Daniel and I exchanged several emails discussing what went wrong in Kingston and what tactical changes the team has adopted under Klinsmann. A lightly edited version of the email conversation follows.

Daniel: I'm frankly confused as to what the thinking behind the US tactics was. Why press against Jamaica? They don't pass the ball out of the back; they just hoof it forward. Pressing isn't going to disrupt their build up. Instead, the US players are just worse off in terms of their defensive positioning. The strangest part is when Jamaica had possession in their attacking half and even third, the US continued with a press. Jamaica are weak in possession. Why not sit back, defend with two banks of four, and let the Jamaicans make the error? The result of the US pressing to try to force errors was, of course, conceding free kicks in dangerous positions. I know a lot of people, myself included, complained about Bob Bradley. His tactics might not have been terribly exciting, but they were pragmatic and made sense in Concacaf qualifying. Jurgen Klinsmann seems either naive or idealistic.

Kyle: Surely not idealistic. I thought Klinsmann said in numerous interviews he changed the team's shape based on the players available and the opposition. That suggests pragmatism to me. Did the US play ugly under Bradley? For some reason my memories are of the US scoring a lot and conceding a lot.

Daniel: I think the primary difference between the US style of play under Bradley and under Klinsmann is that the team are more likely to press under Klinsmann. I don't think the play is any prettier though. Bradley definitely preferred a 4-4-2, and that also seems to be the formation that Klinsmann has settled on. Though Klinsmann sometimes plays more of a 4-4-1-1 than a 4-4-2, and a 4-4-1-1 can morph into a 4-2-3-1 depending how high the wide players are. But, I am honestly struggling to remember how the US played under Bradley so who knows!

Kyle: What would you call the formation in the Jamaica game? 4-3-1-2? 4-3-2-1? 4-3-3? I guess they're all the same. It was a lineup void of creativity, but on paper seems like the perfect one to retain a one goal lead for 90 minutes. Playing three central midfielders, all of whom are known more for their tackling and/or defensive positioning than their creativity, in theory, seems like a formation designed to prevent the opposition from scoring. The way that formation was actually employed was a different story, and I'm not so sure what the actual plan was. I think the 4-3-3 could have been fine had it been executed a little differently. Jamaica's outside midfielders were not threatening--at no point did they show they could put a dangerous ball in the box--so I thought keeping the defense compact in the center of the field was fine. However, the weird pressing thing where certain guys were chasing the ball and others were playing deep created space between the seams for Jamaican players to receive the ball and get fouled. We know how that turned out. The positioning was downright bizarre.

Daniel: I think the US were trying to play a 4-4-2 with a diamond midfield. I admittedly didn't watch the first half, but from what I read, Beckerman was sitting at the bottom of the diamond. When I tuned in for the second half, it looked more like the US were playing, as you suggested, a 4-3-1-2. That's a pretty subtle distinction and possibly a distinction without a difference. From what I saw, Dempsey rarely defended behind the ball. He would sometimes press the Jamaican midfielders from behind, but it seemed like the US defended with a bank of four (the back four) and a bank of three (the three midfielders at the bottom and sides of the diamond). To be clear, I am not trying to blame Dempsey; it's pretty obvious that his instructions were not to defend behind the ball.

As most everyone noticed, the US midfield three (if I say midfield three, I am excluding Dempsey) seemed to pressure the Jamaican players on the ball in our own defensive half and third to try to force a turnover rather than containing, maintaining a disciplined defensive shape, and waiting for the Jamaicans to make an error.

However, I am undecided as to whether the low press by the US midfield three was the cause of the problem or only a symptom of the problem. That is, were the three midfielders told to apply pressure on the player with the ball, or did they press because of the large gaps between the players in the bank of three? If a wide attacking player cuts to the inside, Edu (who was playing on the right) and Jones (on the left) may have felt like they had to close the player down otherwise the gap was large enough between them and Beckerman that the attacking player could easily split them. That wouldn't explain Beckerman pressing from a more central position, though from the short period of time that I saw him play, he didn't seem to be pressing nearly as much as Jones and Edu.

It is obviously considerably more difficult to defend with a bank of three than a bank of four, but it's criminal to defend with a bank of three and not be compact. Jones and Edu were usually positioned very wide, which left gaping holes between the three members of the midfield bank. If a team wants to defend with a bank of three in front of a bank of four, in all likelihood the outside backs will sometimes need to step up towards the bank of three to fill in the gaps in the wide areas. Because Jones and Edu were playing so wide, this couldn't happen, and instead gaps existed more centrally. Jones and Edu attempted to fill these gaps by pressuring players on the ball, but their pressure resulted in conceding free kicks in dangerous central areas rather than turnovers leading to dangerous counterattacks.

In sum, three interrelated factors seemed to contribute to the US defensive problems: (1) defending with a bank of four and a bank of three rather than the safer option of two banks of four, (2) the lack of compactness in the midfield bank of three, and (3) the pressing by the players in the bank of three that resulted in conceding set pieces in dangerous areas.

Kyle: That does make sense. I don't think you'll find too many examples of teams out there keeping three players in front of the ball and defending against a team playing with two center midfielders and two wide midfielders (that's what Jamaica were doing right?). It's always going to be difficult for a midfield three to cover the width of the field. Like you said, in that formation, you'd think it's the job of the midfield three to stay compact and shift as a unit. I don't think there is much of a problem if, say, Edu moves towards to the right sideline to close down the left-sided midfielder in possession as long Beckerman and Jones rotate that way as well to offer cover balance. Then it becomes the job of the left back to step forward from the back four and a little bit central to take away the threat of the big switch to the right midfielder (which you mention above).

I'm afraid we've discussed tactics more here than Klinsmann has with the team following the loss. All the solutions he's discussed with the media seem like they'd be quite obvious to a professional player: "move the ball quicker," "keep possession," "don't foul around the 18," "be patient." Of course those are all things that may help you win a game, but they don't get to the root of the problem which for me was team shape.

Daniel: Yeah, it's a meaningless answer to say "we need to keep the ball better" when asked what a team needs to improve upon. The real question is what you are going to do in order to keep the ball better. (Admittedly, they did say that they need to play simpler, short passes.) Like you, I frankly wasn't that concerned about their possession. Considering the lineup they fielded (which made sense given the personnel available and that this was an away World Cup qualifier in which a draw would have been an acceptable outcome) and the quality of the pitch, the US were never going to dominate possession. Again, as you noted, the team shape was much more problematic than possession.

Ranking the Premier League's most direct teams

“Direct football” or “long ball football” has mostly negative connotations in the modern era. It has become associated with a time in English football when pitches were more mud than grass, and the dominant attacking tactic was to launch high balls into lumbering center forwards to knock down in a 4-4-2 system. Indeed, it was England’s refusal to, until recently, replace direct play with the more fluid, short passing-based systems that were being used in continental Europe as early as the 1930s that has largely been blamed for its lack of success in international tournaments. Long ball football, so the reasoning goes, requires less individual technique and less sophisticated team movement off the ball. Simply whack a ball into a big center forward and hope he knocks it down into the path of a teammate close by or hit it over the top of the defense and hope a speedy forward can get on the end of it. It’s thought to be predictable and generally not the most effective way to use the ball.

In truth however, any assertion that direct play is unquestionably inferior to short passing because it requires less individual technique than dribbling by a defender or using a series of 15 tidy one touch passes to advance the ball 40 yards up the pitch is an incorrect one. Indeed, even in the modern game long passes have often proven to be an effective way to quickly break down an opposition defense. Long balls aren’t a problem in and of themselves. They can be used to stretch a defense and create valuable space between an opponents midfield and back four. Likewise a team can use them to exploit the speed or height and strength advantage an attacker has over opposition center backs. The problem with direct play is when it is overused and becomes the only method a side relies on to advance the ball. Only then does it become predictable and easy to defend. But the same thing can be said of Barcelona’s tiki taka. Relying too heavily on long spells of possession and quick short passes can allow the opposition to restrict the space the attacking side has to play in and deny the time on the ball creative players need to open up a defense.

Of course the most effective team tactics for any given side have to do with the strengths of its players and the players and tactics used by the opposition in any given game. This post will focus on how direct the 20 Barclay’s Premier League teams have been in the first two weeks of the season, the reasons some of them have had for playing direct (or indirect), and the results that different styles of play have produced for different clubs.

When I set out to judge how direct individual Premier League teams are, I first use the average number of long balls each team played per game as a measure of directness and rank teams based on that measure. Stoke City are nearly unanimously considered the most direct team in the Premier League. They’re big and strong, lacking in creative midfield players capable of clever short passing, and in Peter Crouch have a giant of a forward favored to win aerial challenges over just about anyone. However, the data show that after two games Stoke average the 12^th most long balls in the league, a curious result given Stoke are considered the most direct team. Should we assume then that Stoke have drastically altered their playing style over the summer and become less reliant on the long ball? 

As it turns out, we should not. The long balls per game statistic doesn’t tell the whole story of how much a team relies on long passes, as it doesn’t take into account possession and the number of long balls a team plays relative to short passes. For example, team A may have 80% of possession against their opponent team B resulting in them playing 60 long balls and 600 short passes. Team B has 20% of possession while hitting 50 long balls and playing 200 short passes. In this example, team A plays 10 more long passes than team B. They are not the more direct team, however. Their advantage in number of long balls played is attributable to them dominating possession and playing more of every kind of pass. Relative to the number of short passes they play, team A is far less direct. They have a ratio of 10 short passes for every one long ball (600/60=10) whereas team B plays only 4 short passes for every long ball (200/50=4). We can use this same short passes to long ball ratio with data on Premier League teams to rank them in terms of directness. This measurement is shown in the table below. Teams at the top of the table have a higher ratio of short passes to long balls and are therefore less direct than those at the bottom.

Using this method, Stoke are indeed the most direct team in the Premier League after two weeks, playing just 3.48 short passes per long ball. By contrast, Arsenal have been the least direct team, playing 11.08 short passes for every one long ball. Neither of these facts are particularly surprising. While Tony Pulis has always focused on physicality and territory at Stoke, Arsene Wenger has molded a side of mostly creative, technical players who are often small in stature. Interestingly, both teams have struggled to find the net in their first two games. Arsenal have yet to score, registering two goalless draws, one of which was to Stoke last Sunday. Stoke have scored just once in their opening two games.

The sample size is too small to enable us to predict whether either team will struggle to score all season and there are obviously other factors besides how direct a team is that influence number of goals scored. In the case of Arsenal, one big factor may be the loss of Robin Van Persie and the lack chemistry between Arsenal’s three big attacking summer signings Olivier Giroud, Santi Cazorla, and Lukas Podolski.

The data produce some other interesting findings. Both Liverpool and Tottenham brought in new managers this summer. Brendan Rodgers and Andre Villas-Boas were expected to bring new styles of play to their respective teams. Rodgers likes to build the attack from the back with patient buildup play and linking a number of shorts passes. At Swansea last season, his team had the third highest average possession percentage behind the two Manchester clubs. Villas-Boas prefers a pressing game where players expend energy high up the field to win the ball back and then get their rest while patiently knocking the ball around in possession. Neither system relies heavily on the long ball. However, both teams are in the bottom half of the table in terms of short passes per long ball, suggesting they’ve relied on direct play more than most teams. Liverpool have played 5.96 short passes per long ball, while Tottenham have played 5.67.

The data also show that Everton and Newcastle, two teams that finished in the top 7 of the Premier League last season, are among the most direct teams thus far. Newcastle have played 5.07 short passes per long ball and Everton have played just 4.7. These numbers make sense when we consider the strengths of each team and who they’ve played in their opening fixtures. Everton started the season with a home game to Manchester United. United had three injured center backs in Chris Smalling, Johnny Evans, and Rio Ferdinand and were forced to play Michael Carrick out of position in the center of defense alongside Nemanja Vidic. In Marouane Fellaini, Everton had a tall, strong midfielder able to dominate Carrick in the air and knock balls down for his teammates. Everton tried to exploit this mismatch all evening, continually sending long balls towards the towering Belgian. The direct style worked as Everton emerged 1-0 winners. Newcastle’s frequent use of the long pass early in the season likely has to do with the fact that its forward pairing of Demba Ba and Papiss Cisse are full of pace and able to use their powerful running to get in behind the opposition back four. The Magpies have creative midfielders in Johann Cabaye and Hatem Ben-Arfa capable of getting the ball on the floor and playing, but the direct threat of the two Senegalese forwards gives their attack another dimension and they’ll likely continue to look long over the top for them this season.

Again, a sample size of two games doesn’t necessarily reflect how a team will play throughout an entire season, but if we look at data from last season we can get a good idea of how direct we’d expect teams to be in 2012-2013 (at least those teams that have kept the same managers). The figure below shows the same short passes per long ball statistic. Notice Stoke were also the most direct team last season. They also scored the fewest goals in the league with just 36. Another point of interest is that four of the teams that finished in the top six of the table last season--Manchester City (1), Arsenal (3), Chelsea (6) and Manchester United (2)--were among the five least direct teams. This isn’t terribly surprising since these are among the biggest, wealthiest clubs in the league and can afford to bring in the most technically gifted players suited to play in a short passing system. The only top six finisher among the league’s 10 most direct teams was Newcastle. Three of the bottom four finishers were among the four most direct teams--Blackburn, Bolton, and QPR. This almost certainly has to do with the inability of smaller clubs to purchase the most technically gifted players capable of playing a short passing game. 

The table may lead us to conclude that relying on short passes produces superior results to playing direct football. This is somewhat misleading. Clubs like Manchester City, Manchester United, Chelsea, and Arsenal play less direct football in the Premier League because they have technically gifted players, and they gain a competitive advantage over most of their opponents by keeping the ball moving along the ground. It wouldn’t make any sense for Arsenal to set out launching long balls forward against Stoke City--they lose their competitive advantage doing that. But, it also doesn’t make sense for Stoke to try to tiki tika their way up the pitch against Arsenal--they don’t have the quality of players to do that. Their advantage over Arsenal is in their superior size and strength, so they play direct. In short, teams adopt styles that best utilize the strengths of their players and attack the weaknesses of their opposition. Not every team can have the quality of Europe’s top clubs and where there is a gap in talent between two sides, direct play will remain a tactic teams employ. 

Links, 8/28/12 Edition

• Micahel Cox: Fabulous Falcao
• Stylistically, he's something of a throwback -- he's a pure penalty box striker, the type that has become increasingly rare as coaches demand greater linkup play from their frontman. Some of his goals are scrappy and others are beautiful, but there's an incredible efficiency about all of his strikes. He's able to power in headers from unusual angles when his body shape seems wrong, while his feet are always in the right position to exploit a loose ball inside the penalty box. 

• Michael Cox: Liverpool 2-2 Manchester City: neither entirely comfortable with new approach
• Liverpool weren’t 100% comfortable playing high up the pitch and being told to pass out of the back, City weren’t 100% comfortable playing with a back three. Rodgers won’t change his approach – Mancini might have to, because City don’t immediately appear any stronger defensively or offensively, and he is under pressure to get immediate results. Rodgers will take more positives from this game – Allen’s passing, Sterling’s wing play, Coates’ calmness.

• Jonathan Wilson: Football has gone back to the back three, but why can be a mystery 
• Three years ago, playing three at the back had all but disappeared. It had died away in the late 50s and 60s as the W-M was superseded by a back four, and re-emerged in the mid-80s, in slightly different forms, with Carlos Bilardo's Argentina, Sepp Piontek's Denmark, Franz Beckenbauer's West Germany and Ciro Blazevic's Dinamo Zagreb... Yet three at the back has started to make a comeback. It began in Italy, with Udinese and Napoli. At Barcelona, the first and most successful stage of Pep Guardiola's season-long charge backwards through the evolution of tactics was a back three. Then Wigan Athletic started doing it. Now Manchester City have joined in. In fact, in the top divisions of Europe's top five leagues over the weekend, 12 teams used the shape (eight in Italy, two in England, one in Spain, one in France and none in Germany). Three at the back is back.

• Sid Lowe: Athletic Bilbao impotent as Marcelo Bielsa's project unravels apace
•  "We were," Bielsa admitted, "impotent in the face of the dimension of our opponent." The dimension of that opponent was gigantic: Arda Turan was consistently dangerous, Falcao's brilliance was barely believable and Atlético had destroyed Athletic in last season's Europa League final too. Bielsa talked of "antagonistic styles" and took the blame for not devising a strategy to impose upon Diego Simeone and his team. But it was not just about Falcao or Simeone or Atlético; there was something else, something deeper, something a little depressing. Something in that word: impotent.An inescapable feeling that Athletic Bilbao, the team that reached two finals and destroyed Manchester United at Old Trafford, are unravelling before his eyes.

Recap: Chelsea 4-2 Reading

Despite another impressive performance from new signing Eden Hazard, Chelsea struggled mightily to break down Reading's crowded defense and were extremely fortunate to emerge 4-2 winners. Just as they often did last year, Chelsea found it difficult to create meaningful goalscoring opportunities when the opposition allowed them to have the bulk of possession and forced them to patiently pick apart seams against two defensive banks of four. The Blues finished the game with just under 72% of possession but only had 7 shots on goal. Four of those shots were goals but the first was a penalty, the second came off a serious goalkeeping blunder, the third was clearly offsides, and the fourth came late when Reading's keeper had gone forward to attack a corner kick.

Chelsea's problem in breaking down compact defenses last season was largely due to the fact Juan Mata was the only creative passer in the Chelsea attack. As I mentioned in my preview to this game in the previous post, the strength's of their other advanced midfielders/wide forwards in the 4-2-3-1 or 4-3-3 systems last season- Kalou, Sturridge and Ramires- were mainly pace and the ability to advance the ball forward with the dribble. None of the three are exceptional passers. Chelsea's deeper lying midfielders last season, Meireles, Mikel and Lampard, were not of the deep lying creator mold of a Luka Modric, Xabi Alonso, or Andrea Pirlo. Meireles is a hardworking ball winner, Mikel is a very defensive holding player, and Lampard keeps the ball moving and makes well timed runs into the box but none are known for springing attacks with their clever passing. Thus, Chelsea were left with Mata as the sole player with the creativity to cut apart a defense with a vertical pass. As a result, against compact defenses Chelsea would play horizontal balls in midfield all afternoon without ever posing much of a penetrative threat. It was methodical, predictable and easy to defend.


Hazard is a player capable of penetrating the center of compact defenses both with the dribble and with creative forward passes. He brings to the side a directness they lacked in the center of the park last season. He's not content simply keeping the ball moving from side to side but instead likes to go to goal and attack vertically. He proved incredibly effective yesterday, getting into dangerous pockets of space between the seams of the Reading defense and completing 38 of 41 attempted passes in the attacking third (two of which were assists, see the figure below).  Just as they had at Wigan Sunday, Hazard and Mata interchanged between central and wide positions, with Mata given the freedom to come infield to collect the ball. The pair combined for the two highest pass combinations of the match with 18 Hazard to Mata combinations and 18 Mata to Hazard combinations.

With Hazard and Mata orchestrating moves forward Chelsea's struggles yesterday, unlike last season, had less to do with an absence of creativity and more to do with a lack of pace in their ball movement. At 72%, Chelsea had Barcelona-like possession stats but unlike Barcelona, who rapidly move the ball from player to player, they seemed too often to take an extra unnecessary touch, allowing Reading to comfortably rotate their defensive shape. They were also uncharacteristically loose in possession, Ramires particularly guilty of some untidy first touches and passes. Hazard was exceptional in possession but once he got rid of the ball the pace of Chelsea's movement stalled. As the second half progressed with Chelsea trailing 2-1, everything went through Hazard and Mata, with the Spaniard continuing to get in central positions in an effort to get on the ball more. As a result their shape became a bit narrow with the Blues trying to force the ball through the crowded center of Reading's defense. The figure below shows Juan Mata's second half passes in the attacking third. Notice how many of these passes came from central areas, specifically ones just outside the 18 yard box. The two were dangerous on the ball in these crowded pockets of space but Chelsea were struggling to stretch the Reading defense laterally by making threatening runs from wide areas.

The introduction of Sturridge gave the Blues needed width on the right and a new point from which to attack. Prior to his introduction Chelsea's only point of attack was through the middle and thus the center of Reading's defense was able to stay compact and simply check the runs and close the passing lanes of Torres, Mata and Hazard. With Chelsea offering little threat from the right, left back Ian Harte was able to pinch inside and provide additional cover through the middle. Sturridge's introduction forced Harte to defend wider on the right, leaving one less defender to provide cover in the box. Immediately Sturridge was able to use his pace to get around the much slower left back and cut in towards the front post. The threat of him cutting inside from the right left Reading's center backs with an additional concern- not only did they have to check runs coming from the center, an area Chelsea were trying to overload with Mata, Hazard, Lampard and Oscar, they also had to worry about shifting to provide cover for Harte if he were beaten by Sturridge on the outside. Although Sturridge was not involved directly in the third goal, it started when Chelsea had shifted Reading's defense to the right and quickly switched the point of attack to the left allowing Ashley Cole the space to make an unchecked run from left back into Reading's weak side defense.


If yesterday proved anything for Chelsea, it was that their new look attacking outfit will almost certainly experience some hiccups as the players take time to get used to one another. Their key playmaker in Hazard has been involved in only three competitive matches with the squad and Fernando Torres, the only true center forward at the club with whom they'll rely on heavily for goals, was very much a peripheral figure at the club last season. However, there have been hints of what could prove to be a bright future at Stamford Bridge as well, particularly from Hazard and Mata. The two have shown a good understanding of one another in the first two league fixtures and their ability to interchange positions and overload different areas of the field should cause serious matchup issues for opposition defenses. Against teams that pack the defense as Reading did yesterday, Chelsea will need to offer a point of attack from wide areas as Mata and Hazard both move centrally to try to use their combination passing. Wide threats will stretch the opposition defense and allow the two creative players the space to play clever through balls as they did for Chelsea's game-winning third goal yesterday.

Reading will offer Chelsea different challenge than Wigan

Chelsea's rather comfortable 2-0 win in their opening fixture at Wigan offered a display of the pragmatic efficiency the club has become associated with in the years since Jose Mourinho's arrival. After snatching both goals in the opening 7 minutes, the Blues were content to sit deep, absorb pressure with a compact defense and very cautiously move forward. Wigan had 52 percent of the possession, controlled 58 percent of the territory, and completed 113 passes in the final third to Chelsea's 44. Frank Lampard led Chelsea with 8 passes in the final third--7 Wigan players had as many or more passes in the final third. Despite dominating many of the offensive statistics, the Latics rarely looked like threatening Petr Cech's goal, and throughout the 90 minutes the result never appeared to be in doubt. It was hardly the kind of Barcelona-esque, stylish possession-based performance Roman Abramovich wants from Chelsea but the win was as easy they come.

In many ways the possession-based attacking philosophy of Wigan under Roberto Martinez plays directly into what proved to be Chelsea's strengths late last season--sitting deep and drawing opposition midfielders and full backs forward, then regaining possession and countering into the space they left vacated. Chelsea were much less comfortable when forced to unlock compact defenses with tidy possession and clever buildup play. Even against superior opposition Wigan are not a team that sits back and defends in banks of four with ten men behind the ball. Indeed, they used the same brave 3-4-3 formation Sunday that they used during their terrific late season run last year. They like to get on the ball and get numbers into the opposition half (evidenced by their dominance of passes in the final third). Inevitably that leaves them vulnerable to quick counter attacks which proved to be Wigan's undoing Sunday.

In the second minute Wigan advanced the ball into the attacking third through an impressive interchange of passes between right wing back Anderson Boyce and right forward Victor Moses down the sideline. Boyce then played the ball to Franco Di Santo about 24 yards from goal in the middle of the pitch. Shaun Maloney had drifted centrally from his left forward position, forcing right back Branislav Ivanovic to track his run for Chelsea. Maloney's clever movement left acres of space down the left flank for Maynor Figueroa to run into from his left wing back position. The movement was impressive from the Latics, but Di Santo was caught in possession before he could find Figueroa, leaving both wing backs high up the field and out of defensive position. Ivanovic collected the ball at Chelsea's 18, played a short pass to Juan Mata and took off into the space left vacated by Figeueroa. Two passes later and Eden Hazard had turned away from Ivan Ramis brilliantly, leaving Wigan's other two center backs hopelessly exposed. He found Ivanovic to his right who tucked the finish away coolly. Wigan had been undone in the blink of an eye by four quick passes. The buildup that led to Hazard winning the penalty that made it 2-0 was similar. Figueroa advanced to the touchline before hitting a poor cross directly to Cech. Cech quickly rolled the ball to Lampard who was left with acres of space down the middle to dribble into. Hazard drifted right into the space left empty by Figueroa's attacking run. He received a pass from Lampard, advanced the ball into the box and was ultimately hacked down by Ramis. Two quick counter attacks had effectively killed the game off inside 10 minutes.

Reading at Stamford Bridge will likely offer Chelsea a very different kind of challenge. Whereas Wigan bravely attacked with numbers and left themselves susceptible to the counter, Reading is expected to get bodies behind the ball and get into a compact defensive shape before looking to spring counterattacks of their own. Under Brian McDermott, Reading are an extremely organized outfit and conceded the fewest goals in the Championship last season. Chelsea will be forced to patiently keep possession and provide the creativity to unlock a crowded defense, something they struggled with at times last season. The addition of Hazard should certainly help in this regard. Last season Chelsea's wide outside forward and attacking midfield options were mainly Mata, Daniel Sturridge, Salomon Kalou and Ramires. Of those four, only Mata is known for his creative passing ability. Kalou, Ramires, and Sturridge are known for their pace and ability to advance the ball with the dribble but aren't especially technical players. Their skill sets mean they are more suited to playing a counter attacking style where they can run at defenders with pace rather than one centered around patient buildup play. Hazard is a player with good technique who brings both the ability to beat opponents off the dribble and to unlock them with a clever pass. Having two technical players behind Torres should make Chelsea a more threatening and unpredictable side when they come up against crowded defenses. The positioning and movement of those two today should be one of the more intriguing tactical elements of the match.

It will be interesting to see if Roberto Di Matteo's team sheet is influenced by the fact that Chelsea play three games in the opening seven days of the season. They face a strong Newcastle side at the Bridge Saturday. It's possible Di Matteo could opt to rest a usual starter or two in the hopes they'll be able to defeat an inferior Reading side today without a first choice 11. Look for Reading to try to frustrate Chelsea by limiting their space to move the ball in the attacking third and crowding the box. If Chelsea get an early goal they should cruise.

Match report: Arsenal 0-0 Sunderland, 8/18/12

For a third consecutive season, Arsenal opened up their Premier League campaign with a draw. Going back to last season, Arsenal now have gone 4 consecutive home matches without a win--their longest stretch of home games without a win since August 1995.

Arsenal dominated the possession/passing battle throughout the match. The Gunners had 70.1 percent of the possession and completed 637 passes to Sunderland's 222 passes. Arsenal's passing completion rate was an impressive 91 percent, while Sunderland completed 76 percent of their passes. It would be difficult to overstate the extent to which Arsenal out-passed Sunderland. In fact, in the figure below, I break up the game into 18 different 5-minute increments (e.g., minutes 0-5, minutes 6-10) and compare Arsenal and Sunderland's passing statistics. Sunderland completed more passes than Arsenal in only a single 5-minute period of the match. The height of the red bars indicate the number of passes Arsenal completed in each 5-minute increment, and the height of the black bars indicate the number of passes completed by Sunderland in each 5-minute increment.

Of course, a disparity in raw passing statistics can be misleading if most of the passing is concentrated in the middle and defensive thirds of the field. But, Arsenal also completed far more passes in the attacking third than Sunderland: 162 passes completed out of 201 attempted for Arsenal and 35 passes completed out of 61 attempted for Sunderland. Arsenal's passing dominance in the attacking third unsurprisingly resulted in far more shots than Sunderland: 23 total attempts for Arsenal vs. 4 total attempts for Sunderland.

The match data presented above would seem to suggest a far more unbalanced game than was actually the case. Sunderland defended very deep and were difficult to break down (see Kyle's post for more on the tactics employed in the game). Thus, while Arsenal were able to complete a lot of passes in the attacking third and generate far more shots than Sunderland, Sunderland restricted Arsenal to very few genuine scoring chances. In fact, only 3 of Arsenal's 23 shots were on goal, while 2 of Sunderland's 4 attempts were on goal. Both of Sunderland's shots on target came in the first 10 minutes of the game, the most dangerous of which came from a James McClean breakaway. Sunderland's only other 2 attempts both occurred later in the first half. Arsenal's best chance came in the 83rd minute when Santi Cazorla cleverly slipped a pass between 2 Sunderland defenders to set up Olivier Giroud only 8-10 yards from goal. Giroud could not convert, as his strike sent the ball wide past the post.

Arsène Wenger and Arsenal supporters are likely disappointed with the result, but they are almost certainly encouraged by some of the players' performances, especially Cazorla. The new signing was far and away Arsenal's most dangerous player. He led all players in several statistical categories. He created 7 scoring chances (no other player created more than 2). He completed 36 passes in the attacking third (next on the list is Mikel Arteta with 25 passes in the attacking third). And, Cazorla had more shots (4) than any other player. Gervinho was also impressive, especially in terms of his willingness to take on defenders. He had 10 successful take-ons (Theo Walcott ranked second with 5 successful take-ons). However, Gervinho attempted 20 take-ons, far more than any other player (Walcott again ranked second with 7 take-on attempts).

Arsenal's three big signings--Cazorla, Lukas Podolski, and Giroud--all featured in today's game. In addition, Abou Diaby returned from injury to play over 45 minutes for Arsenal for the first time since May 2011. The game was also Per Mertesacker's first match for Arsenal since he suffered an injury in early February. Given the new and recently repaired components in Arsenal's system, it would be premature to pass judgment at this stage.

Thoughts: Arsenal 0-0 Sunderland

Arsenal failed to make their 70% possession count as Sunderland put in a gritty defensive performance to emerge from the Emirates with a deserved 0-0 draw. It was a pretty straightforward tactical battle; Sunderland defended with two lines of four to restrict the space Arsenal had to play in while the Gunners controlled possession and looked to use quick passing to find gaps in the defense. Here are some of the more interesting developments and observations I noticed.

• The Wearsiders were content to get all 11 men behind the ball and force Arsenal to patiently pick them apart. They looked to counter through Sessegnon, McClean and Campbell and did so with some success early on. However, as the game wore on Arsenal closed off their outlet pass and Sunderland offered little going forward the final 65 minutes of the game. 
• With Sunderland dropping deep in two banks of four to restrict the space Arsenal had to play in, it was important for the Gunners to find space in between the lines to unlock the compact defense. New signing Santi Cazorla did this with considerable success early on in the match, moving into pockets of space between the Sunderland back four and midfield. However, Lee Cattermole began getting tight on Cazorla and the Spanish midfielder found it more difficult to find the time for a cutting pass as the game went on. He still found ways to get on the ball, often coming deeper to receive it. He played a brilliant ball to put fellow new signing Olivier Giroud through on goal but the Frenchman, who had recently come on to replace Podolski at center forward, scuffed his effort wide. 
• Despite starting as the loan center forward, Podolski often came very deep into the midfield to receive the ball. It’s understandable why he was doing this since the the central areas around the box were so crowded. However, it left Arsenal looking like they were playing a 4-6-0. The problem with that was Arsenal didn’t have anyone to stretch the Sunderland defense vertically and create space for Cazorla between the Sunderland back four and midfield. Giroud stayed closer to the Sunderland center backs which created more space for Cazorla and second half substitute Aaron Ramsey to float into. Ramsey got a decent look at goal from 18 yards out but his effort was extremely poor.
• In the first half Arsenal looked more dangerous when they were able to win possession from the Sunderland midfield and break quickly towards goal with 3 or 4 passes. They struggled when Sunderland were able to get all 10 men behind the ball. I thought at halftime it might be a good idea for Arsenal to concede some possession to the Sunderland midfield and press a little deeper towards the midfield line, the thinking being that they could push Sunderland out of their defensive shell opening up space behind the Black Cats’ midfield to counter into with speed. Indeed, Arsenal seemed to use this strategy and the game opened up very slightly in the middle stages of the second half. However, when Arsenal were able to break they were missing the final ball or and the presence of a poaching center forward in the box to put the ball in the net.

Preview 2012-13: Arsenal

Arsenal's summer transfer windows in 2011 and 2012 are a study in contrasts. In 2011, Arsenal were paralyzed as manager Arsene Wenger's stubbornness seemed to preclude the club from conducting their necessary business. The negotiations for the sales of Cesc Fabregas and Samir Nasri carried on for several weeks and likely interfered with the club's preparations for the upcoming season. Wenger seemed reluctant to close deals on quality players early in the window (e.g., Juan Mata). Only in the aftermath of a devastating 8-2 defeat to Manchester United on August 28, 2011 (the first time Arsenal had conceded 8 goals since 1896) did Wenger seem aware of how desperate the club's circumstances actually were. With only three days remaining before the close of the transfer window on September 1, he completed deals to bring in Mikel Arteta, Per Mertesacker, Andre Santos, Park Chu-Young, and Yossi Benayoun (on loan). While some of these additions proved quite useful to the club, Wenger only acted at the end of an abysmal month in which Arsenal somehow incurred 3 red cards and captured a single point out of a possible 9.

If the summer of 2011 demonstrated Wenger's deficiencies in the transfer market, the summer of 2012 has reminded observers of his primary strength: identifying and purchasing undervalued talent. Unlike 2011 when Wenger procrastinated buying replacements for wantaways Fabregas and Nasri, Wenger quickly moved to bring in replacements for Robin van Persie, the malcontent striker with only one year remaining on his contract. Far in advance of van Persie's public declaration that he would not be extending his contract with the club, Wenger had already negotiated the purchase of Olivier Giroud who scored 21 goals in France's Ligue 1 (joint top scorer) and Lukas Podolski who scored 18 goals in Germany's Bundesliga (joint 4th leading scorer). Remarkably, Wenger purchased Giroud and Podolski for £13m and £11m, respectively. In the current transfer market, these fees are paltry for players of their caliber of talent. Giroud and Podolski are not Wenger's only marquee purchases on the cheap, however. Arsenal understood that the precarious financial situation of Malaga in Spain's La Liga presented an opportunity. Malaga were (and probably still are) desperate for cash and fast, as several players filed complaints that they are owed back wages (among other debts owed). Arsenal, with plenty of cash on hand, may have negotiated a discounted fee for Santi Cazorla by paying most of it up-front to help alleviate Malaga's near-term financial woes. (In general, most transfer fees are paid over the course of a period of years rather than up-front.) The transfer fee for the highly creative midfielder is reportedly around £16m, which is a great bargain for a player of Cazorla's quality.

Wenger and the club certainly deserve plaudits for their activity in the transfer market thus far. Even so, more questions remain than have been answered by Arsenal's transfer activity. If van Persie leaves, which admittedly seems modestly less likely with each passing day, Arsenal lose 40 percent of their league goals from last season. Can the new additions compensate for this dearth of goals? Maybe but it's a big maybe. Adjusting to the Premier League can take time, and Giroud and Podolski are good players, but they are not of van Persie's quality.

Even more problematic for Arsenal, the back four are not improved from last season. Center back Laurent Koscielny had a fantastic season in 2011-12; can he maintain that level of play? Center back Thomas Vermaelen's positioning was very poor last season. Can he improve and play at the level that people expect him to play? While Per Mertesacker is an adequate backup at center back, if he is required to play regularly, as was the case last season, opponents will likely easily exploit his lack of pace. (As an aside, for someone as tall as Mertesacker is, he is incredibly weak in the air.) Left back Kieran Gibbs is a promising young player, but he seems prone to nagging injuries. He missed nearly 4 months of last season due to a hernia and related complications. Reserve left back Andre Santos is dangerous going forward, but he lacks the sort of defensive capabilities that are generally considered requisite to be, errr, a defender. Most worrisome for the club, right back Bacary Sagna will be out for the beginning portion of the season as he recovers from a broken leg suffered at the end of last season. I think his return in February of last season was incredibly important to Arsenal's resurgence in which they took 27 out of 30 possible points between February 4 and April 11. In my view, Arsenal are a substantially better team, both in attack and defense, with Sagna out on the field. Until Sagna is healthy, I believe that Arsenal are better off playing Francis Coquelin out of position at right back over Sagna's understudy at right back, Carl Jenkinson. Even though Coquelin is an improvement over Jenkinson, he is still a huge drop off in quality from Sagna.

In sum, Arsenal have made three very astute purchases in Giroud, Podolski, and Cazorla. They are in a better position leading up to the season than most fans would have anticipated after van Persie announced his intention to leave the club. Even so, whether Arsenal's attacking players can maintain the potency of their scoring threat in the absence of van Persie (should he leave), and whether the defense can improve from last season are questions to which the answers are far from clear.

Further Reading:

• David Hytner, "Premier League preview No1: Arsenal"
• Dominic Fifield, "Olivier Giroud could make a striking difference for Arsenal"
• Raphael Honigstein: "A more mature Podolski set to prosper in Premier League"
• Sid Lowe: "Cazorla lifts Malaga's prospects"

Goalkeeper distribution trends limited in their ability to predict successful tactics in MLS

At the end of May, Peter Brownell wrote a blog post for Opta Sports that looked at data on the percentages MLS goalkeepers played the ball long versus short in the 2011 MLS season. The table above is from Brownell's original posting and shows goalkeeper distribution trends from that season for each MLS team. While his data set is interesting, Brownell uses it to arrive at some suspect suggestions about what are the most successful tactics for MLS teams to employ.


He starts off the article noting that American soccer is known for the speed and athleticism of its players over their technical merits on the ball. I'd agree with this point. He uses this bit of information to hypothesize that goalkeepers in MLS will more often than not select to hit long fifty-fifty balls rather than going short and forcing a team to advance the ball with patient buildup from the back.

Indeed, his data backs that hypothesis. MLS goalkeepers elected to play the ball long 58% of the time. Only three teams played the ball short more than long- Chivas USA, Toronto FC, and the New York Red Bulls. He points out that these three teams all had relatively disappointing seasons. Chivas USA and Toronto FC finished in 8th place out of 9 teams in their respective divisions and the New York Red Bulls squeaked into the playoffs in 5th place and were disposed of quickly by the LA Galaxy. Brownell unconvincingly uses these three teams' lack of success to suggest that the best tactics for MLS teams may be to forgo the defense altogether and launch long balls from the goalkeeper. He says,

This suggests, perhaps, that playing out of the back in MLS 2011 might not have been the most fruitful strategy.  One theory to explain this is the idea that MLS players are not technical enough to competently build long, calculated passing sequences.  This is not a slight at the technique of MLS players, but rather trumpets the physically superior MLS rosters.  MLS players are athletes who are fast and fit, bring intense pressure, and close down backs quickly.  As a result, teams like Toronto FC and Red Bull New York turn the ball over in their defensive thirds and give up cheap goals (watch Tim Ream's gaffes against the Philadelphia Union and Real Salt Lake for proof).

By implementing a 4-3-3 and trying to play the beautiful game beautifully, Aaron Winter (Toronto FC's manager) is fighting a losing battle.  The numbers say so at least.  It is easier, safer, and more reliable to forget the backline, smash the ball forward and attempt to win the territorial battle.  Getting big-bodied strikers and midfielders into the box and battling for set-pieces is regarded as ugly soccer, but it can bear results.

Brownell fails however to mention the relative lack of success in the league of teams at the other end of his table, those that elected to play long from the keeper the highest percentage of times. Of the five teams that played the ball long from the keeper the most, only Columbus made the playoffs and they were beaten in the opening round. His suggestion that patient buildup from the back may not be the most fruitful strategy for MLS teams would be much better supported by possession data. While the percentage of balls played long from the keeper may be indicative of the importance a team places on keeping possession, I don't think that stat in and of itself is very telling in explaining a team's success. After all, how many goals begin with moves that start from the goalkeeper? I'd guess that the vast majority of the time goals come from intercepting the ball from the opposition, not from a team's own keeper.

I do think Brownell's data could be put to good use since I believe how often a team plays long from the goalkeeper is a good indicator of how much importance they place on winning the possession battle. It would be interesting to see how Brownell's data correlated to average possession data in MLS and then how average possession data correlated to league success. This information would be more telling of whether direct or possession-style tactics created more successful teams.

Is Clint Dempsey undervalued in the transfer market?

The often unreliable Daily Mirror reported earlier this week that Clint Dempsey is on the verge of a move to Liverpool for a £7.5m fee. A common response to this transfer rumor has been that £7.5m seems like a very low fee for a player of Dempsey’s quality. Dempsey was the joint-4th leading scorer in the 2011-12 Premier League with 17 league goals. Perhaps even more impressive, Dempsey was involved, by either scoring or assisting, in 54 percent of Fulham’s non-own goals in the 2011-12 Premier League—that is the highest share of involvement in a club’s non-own goals in the Premier League's 2011-12 season.¹ Moreover, Dempsey’s 23 goals for Fulham in all competitions during the 2011-12 season was hardly a fluke. During the 2010-11 season, Dempsey scored a very respectable 14 goals in all competitions, 12 of which were Premier League goals. 

While it’s clear that Dempsey is a very talented player, actually calculating a player’s value in the transfer market is a convoluted exercise. Two primary factors are driving down his cost in the transfer market: his age (29.3 years) and the one year remaining on his contract. Prior to the implementation of UEFA’s Financial Fair Play (FFP) rules,² wealthy clubs like Chelsea and Manchester City were much less concerned about a player’s value in 3-5 years. But, as the rules are implemented, these clubs are now forced to consider a longer time horizon for player purchases. If Chelsea buy a player who is age 21 for £35-40m, in 4 years that player is still in the prime of his career. Consequently, Chelsea could sell the player and likely collect a substantial fee or avoid paying a hefty transfer fee for a replacement. On the other hand, if Chelsea buy a player who is age 29 for £35-40m, in 4 years it is unlikely that the player will still be able to play at high level. Chelsea probably could get at most £2-3m for a player who is 33, and they would need to spend even more cash for a replacement. While this example is presented as a hypothetical, Chelsea spent £35m on 21-year-old Eden Hazard earlier this summer. During the summer of 2006, Chelsea bought then 29-year-old Andriy Shevchenko from AC Milan for £40m. Shevchenko had a terribly disappointing spell with the club. He returned to Milan on loan for the 2008-09 season and then left Chelsea for Dynamo Kiev on a free transfer during the summer of 2009.

Simon Kuper and Stefan Szymanski summarize the basic economics behind buying and selling older players in their book Soccernomics (57-58):

Second, older players are overrated. "I've noticed over the years how often Liverpool sell players as they near or pass their thirtieth birthday," notes Taylor in his book. "Bob Paisley [Liverpool's then manager] believes the average First Division footballer is beginning to burn out at thirty."  Taylor added, rather snottily, that that was true of "running side like Liverpool," but less so of a passing one like Forest. Nonetheless, he agreed with the principle of selling older players.

The master of that trade today is Wenger. Arsenal's manager is one of the few people in soccer who can view the game from the outside. In part, this is because he has a degree in economic sciences from the University of Strasbourg in France. As a trained economist, he is inclined to trust data rather than the game's received wisdom. Wenger sees that in the transfer market, clubs tend to overvalue a player's past performance. That prompts them to pay fortunes for players who have just passed their peak. Probably because Wenger was one of the first managers to use statistics to assess players, he spotted that older players declined sooner than was conventionally realized. […]

Wenger often lets defenders carry on until their midthirties, but he usually gets rid of his midfielders and forwards much younger. He sold Patrick Vieira for $25 million (age twenty-nine), Thierry Henry for $30 million (age twenty-nine), Emmanuel Petit for $10.5 million (age twenty-nine), and Marc Overmars for $37 million (age twenty-seven), and none of them ever did as well again after leaving Arsenal.

The Taylor that Kuper and Szymanski reference above is Peter Taylor who, along with Brian Clough, assembled the Nottingham Forest teams in the 1970s that won two European Cups on a very limited budget.

A simple method to gauge Dempsey’s value in the transfer market is to compare Dempsey to “similar” purchases. To do this, I searched through the current rosters of Manchester City, Manchester United, Arsenal, Chelsea, Tottenham, and Liverpool. From the current rosters for only these clubs, I compiled a list of the players whose clubs purchased them on or after their 28th birthday:

Of the 13 players who satisfy the above criteria, four players were free transfers, five players cost less than £10m, three players cost £10-12m, and one player cost £16m. Interestingly, Manchester United’s roster does not contain a single player who was purchased by the club on or after his 28th birthday. The two largest fees, those of Kolo Toure and Gareth Barry, were purchases of Manchester City during the summer of 2009 prior to the start of FFP implementation. Meireles, Arteta, and Benayoun seem like more comparable purchases, but I believe all three had longer than one year remaining on their contracts at the date of their transfer. When players have a single year remaining on their contract, their transfer fee is obviously deflated since the purchasing club could have the player on a free the following summer. How much is having Dempsey for the 2012-13 season worth? This question is somewhat misleading since a purchasing club would also risk losing Dempsey to another club the following summer. Arsenal reportedly offered €7m for Marouane Chamakh in the summer of 2009, and Bordeaux turned down the offer. Chamakh subsequently signed with Arsenal on a free transfer in the summer of 2010.

My personal view is that a player of Dempsey’s quality with a proven track record in the Premier League is likely worth more than £7.5m even at 29.3 years of age with one year on his contract. That said, I think there is little evidence to suggest that he is severely undervalued. He is probably worth in the range of £1-3m more than the rumored £7.5m price tag, but his price very well could increase before any deal is done. If he had more than one year remaining on his contract, I'd probably add another £1-2m to the fee.

---

1. Dempsey’s 54 percent edged out Robin van Persie in this category who was involved with 53 percent of Arsenal’s 2011-12 Premier League goals.
2. While UEFA will not assess a club’s finances until the 2013-14 season, they will analyze a club’s finances for that season and the two prior seasons (2011-12 and 2012-13).

News in brief, 5/28/12

Eden Hazard has quite respectably quashed all speculation about his future by tweeting earlier today, “good afternoon guys. i made up my mind. see you later. thanks.” You’re welcome and see you later as well, Eden. Also, we all really appreciate your earnest attempts to quell rumors about whether you’ll end up at Chelsea, United, or City. Muckraking journalist Luis Saha indicated that Hazard is “[a]pparently on his way to Chelseafc.” But, in a shocking turn of events, the usually reliable Saha backed down from his original well-sourced report: “Im just like everyone. Only him and clubs know. We, Fans of football have to wait and see.” Journalistic integrity seems to be an anachronism in the age of digital media. Saha’s report may pan out, however, as Hazard has reportedly hopped on British Airways flight BAW395 from Brussels to Heathrow for a physical with Chelsea. In actual news, Italian police raided Italy’s Euro 2012 camp as part of a serious and widespread match-fixing investigation in Italian football. Defender Domenico Criscito has been dropped from the squad due to the investigation. Tom Kington of the Guardian reports: “In the same operation, police arrested 17 people, including 11 players, among them Stefano Mauri, the captain of leading Serie-A side Lazio, and placed Antonio Conte, manager of champions Juventus, under investigation.” The investigation may have made Italian manager Cesare Prandelli’s life a bit easier, as he only needs to drop one further player from his now 24-man roster. While some pundits might speculate that the match-fixing probe will be a distraction for the Italians and perhaps a bad omen, history suggests to the contrary. Simon Kuper notes, “Whenever Italy have a big match-fixing scandal, they win the tournament (1982, 2006).” ESPN The Magazine’s recent story on match fixing now seems especially topical. In other actual news, Serbia’s manager Sinisa Mihajlovic sent home and suspended midfielder Adem Ljajic for refusing to sing the Serbian national anthem for personal reasons. It’s been a tough month for Ljajic in terms of altercations with managers. Earlier in May, his club manager Delio Rossi quite literally assaulted Ljajic on the bench when Ljajic gave Rossi a sarcastic thumbs up after being subbed out. Lasana Liburd has uncovered documents implicating former FIFA official Jack Warner in further corrupt dealings. Gareth Barry--who was impressively speedy at the 2010 World Cup--is out of Euros after suffering a muscle tear in his abdomen during the friendly against Norway. Roy Hodgson has replaced Barry with Phil Jagielka.

Does having less of the ball mean more success against Chelsea?

Dan informed me yesterday of an interesting defensive strategy employed by Newcastle in their 2-0 win over Chelsea at Stamford Bridge that I'd not noticed while watching. Newcastle forwards Papiss Cisse and Demba Ba would press Chelsea's back four high up the pitch when Newcastle lost possession while their defenders and midfielders dropped in to form two lines of four. Once the defense and midfield had gotten into position Cisse and Ba would stop applying pressure and also drop in behind the ball. I assume the initial pressure from Ba and Cisse was meant to prevent a quick forward transition pass, allowing the Newcastle defense to get in its proper shape and preventing Chelsea from countering. Their aim was not necessarily to win the ball back but rather to slow down Chelsea's transition into offense. As the game progressed into the second half with Newcastle enjoying a one goal advantage, the Magpies' began to defend deeper and deeper, restricting the amount of open space Chelsea had in the attacking third and forcing them to be creative enough to score on a crowded defense. Chelsea ultimately failed to do so.

The strategy seemed to make perfect sense. Aside from Juan Mata, who started the game resting for the FA Cup on the bench, Chelsea lack midfielders that are able to open up a crowded defense with incredible vision and passing ability. Their central midfielders in Raul Meireles, Jon Obi Mikel, and Frank Lampard (who was also rested at the start yesterday while Florent Malouda was given the start in his place) are not considered to have the same passing ability as central midfielders like Andrea Pirlo, Andres Iniesta, and Xavi. Meireles and Mikel are better known for their work rate and defensive positioning while Lampard is famous for his well timed runs into the box. Likewise, two of Chelsea's attacking wide players, Ramires and Daniel Sturridge, are not especially gifted passers. Both prefer to receive the ball in open space where they can use their exceptional pace to run at defenders and cut inside. Juan Mata is the only Chelsea attacking player likely to carve open a defense with a penetrating pass. What all of this seems to suggest is that Chelsea are a better suited to dropping deep and allowing teams to come at them and then exploiting open spaces left by the opposition defense on the counter. When the opposition chooses to drop deep themselves and force Chelsea to come at them, their lack of creative passing midfielders indicates they are likely to struggle to score goals.

To test whether Chelsea enjoyed more success when they were allowed to defend deep and counter rather than playing a team that defended deep themselves and forcing Chelsea to score on a patient buildup I looked at their possession percentages and results throughout the season in all competitions. I assumed that when Chelsea were outpossessed by the opposition they defended deeper and were able to counter. I guessed that because Chelsea are not set up to break down crowded defenses while enjoying long spells with the ball the data would show they actually achieved better results when their opponent outpossessed them. Importantly, the data only included what I defined as elite opponents. Elite opponents were defined as teams in the top 8 of the Premier League (with the inclusion of Fulham because they are level on points with 8th place Liverpool) plus all Champions League opponents (minus Genk who I thought were not on par with a top 8 Premier League side). Thus, the data consist of games played against Manchester City, Manchester United, Arsenal, Tottenham, Newcastle, Evertion, Liverpool, Fulham, Bayer Leverkusen, Valencia, Napoli, Benfica, and Barcelona. I elected to only look at elite opponents because I assumed against lesser sides Chelsea would inevitably enjoy more possession and they would score simply because they had superior players than their opponents. For example, Chelsea will almost always outpossess Wolves and Blackburn and will likely win those games because they are more talented. I wanted to look at results against teams that could come somewhat close to matching up with Chelsea talent-wise. Outside of matches against this group of elite opponents Chelsea were only outpossessed once in 29 matches (they had 49% of the ball in a 1-1 draw against Wigan). The data includes FA Cup and Carling Cup matches against teams that are currently in the top 8 of the Premier League.

The data show that in the 16 matches against elite opponents in which Chelsea enjoyed less than 50% of possession they averaged 1.69 points per game (7 wins, 3 defeats, 6 draws). In the 11 matches Chelsea had 50% or greater possession they averaged only 1.00 points per game (3 wins, 6 defeats, 2 draws). The three games Chelsea had their lowest possession percentage were all in the Champions League against Spanish opposition and they enjoyed arguably their season's most successful results in all of those games. They beat Valencia in the final group stage game 3-0 with just 31% (a game they had to win to advance to the knockout stage), beat Barcelona 1-0 at home with 21% and tied Barcelona away 2-2 with just 18% despite being a man down to advance to the Champions League final.

The data suggest that the means for teams to achieve success against Chelsea is to allow them to have more of the ball. Chelsea are not effective when they are forced to score from patient buildup play. When forced to score patient goals Chelsea often tend to bring their two fullbacks forward (Ashley Cole and either Jose Bosingwa or Branislav Ivanovic). This leaves Chelsea's two center backs vulnerable to counter attacks and, given John Terry lacks for pace, can often mean opposition forwards can get in behind the Chelsea defense for breakaways on goal. Perhaps had Barca been willing to compromise their philosophy and ceded some possession to Chelsea, like Newcastle yesterday, they may be playing for a Champions League final. Chelsea had 63% of the ball against Newcastle.

(I think Chelsea have a decent chance of winning the Champions League final. I'm guessing Bayern Munich will try to take the game at Chelsea, allowing Chelsea to defend deep and counter. Bayern Munich's defense looked leaky early on against Real Madrid and Chelsea may be able to exploit them early then park the bus).

Links, 4/14/12 edition

• John Bennett: The extraordinary life of German goalkeeper Lutz Pfannenstiel
  
• He was locked up in a Singapore prison, he was declared dead on the pitch, he was signed by 25 clubs in 13 countries and he is still the only professional to have played in all six of Fifa's confederations.
• Miguel Delaney: Dispelling the early season 'truths'
• Spurs, for example, haven't just gone off form; they've almost collapsed, earning a full point per game less. As it stands, that is one of the worst drop-offs in league history. Liverpool, meanwhile, aren't far behind. Remarkably, it's Fulham who are the most upwardly mobile - and most underappreciated - risers.
  
• The Secret Footballer: Moneyball works better at Stoke than Liverpool
• Stoke City's style is the most rudimentary example of soccernomics on a football pitch. Each full-back generally looks for Peter Crouch on an angle (final third entry) and, in turn, the striker will attempt to cushion the ball down into the penalty area (penalty box entry) for his partner or a midfield runner. It goes almost without saying that the higher these two statistics are over a season, the more likely Stoke are to end up with a shot on goal.
• Ryan Bailey: Peter Crouch probably doesn’t play a lot of golf
• It was the England Foundation Charity Golf Day today, which gave Peter Crouch the perfect opportunity to show how ridiculous he looks holding a driver.

MLS wages and performance for 2011

I've had the MLS on my mind as of recently with the 2012 season kicking off tomorrow. Inspired by Simon Kuper and Stefan Szymanski's work on evaluating the best managers in English football, I've put together a simple analysis that attempts to answer the question: which are the best run clubs in the MLS? The idea behind Kuper and Szymanski's analysis is that a team's wage bill explains much of the variation in a team's performance. Thus, managers whose teams consistently perform better than would be predicted by wages are effective managers.

Their analysis is much more thorough than the analysis below, and they focused on managers in the English game. This analysis is focused more on clubs than managers (or coaches), since most MLS coaches have limited control over which players are on their roster. The complex salary cap rules also make interpretation of the MLS data difficult. The figure below displays the cost (in wages) for every 3 points earned in the league. Since a club earns 3 points for each win, it is more simply thought of as the cost per win (or alternatively 3 draws). Clubs that over-perform (i.e., have a low wage bill and perform well) have a low cost per win, while clubs that under-perform (i.e., have a high wage and perform poorly) have a high cost per win. In other words, a shorter bar in the figure below indicates better performance given the level of wages.

Data on points earned are from the MLS, and data on wages are from the MLS Players Union. I use 2011 guaranteed compensation to calculate each club’s wage bill. The wage column in the above figure is in millions of dollars.

As can be seen in the above figure, Seattle has the lowest cost per 3 points earned, as each win cost the Sounders only $161,911 in wages. This suggests that Seattle is an effectively managed club in terms of player personnel decisions (e.g., signing bargain players who are undervalued) and/or tactics and strategy (i.e., getting the most out of the players available). On the other hand, the New York Red Bulls have the highest cost per win, $871,810 in wages, which is over 5 times more costly than Seattle.

It's important not to make too much of the above data. The LA Galaxy had a fantastic season in 2011, and yet, judging by the above metric, they performed poorly. I would argue that the law of diminishing returns applies to soccer in the same way that it applies to most firms. In the context of soccer, at a certain point, each additional good player added to a roster yields a lower return than the player before. For instance, if a soccer club has no good strikers, it probably would reap substantial rewards by signing a top striker. The return on signing a second top striker is also likely quite high (though possibly lower than the first). But, the return on signing a third striker would almost certainly be lower than the second (though it still might be considerable), and signing a fourth striker would likely yield an even lower return than the third. In other words, at some point, the return from each additional dollar spent on the club is lower than the last. So, once the club is past this point, the cost of earning an additional point (our measure of production) increases; this is known as increasing marginal cost.

One method to account for diminishing returns is to use a logarithmic functional form. I replicate my analysis from above, but instead of calculating wages per 3 points earned, I use the natural log of wages, ln(wages), per 3 points earned. Interpreting the actual values for log wages per win is not very intuitive, so I present the rankings using each of the two methods in the table to the left.

Even after taking into account diminishing returns with log wages, Seattle remains atop the rankings. Amazingly, the LA Galaxy jumped 15 spots from 17th based on wages to 2nd based on log wages. New York improved in the rankings, but the Red Bulls only jumped to 12th. Aside from LA and New York, no other team improved more than 2 places. Portland experienced the largest drop in the rankings falling from 4th to 11th, while New England fell 4 places and Chicago and Vancouver each fell 3 places.

Manchester United vs. Tottenham Hotspur, second half

Tottenham dominated most of the first half, and they continued to control the game in the opening period of the second half. Unfortunately for Spurs, Ashley Young scored goals in the 60th and 69th minutes against the run of play to give Manchester United a 3-0 lead. Young's first goal was a difficult volley into the far, lower corner. For his second goal, he received the ball from Evra with plenty of time to turn, took two touches, and then beautifully curled the ball into the far, upper corner of the goal. As nice a finish as the second goal was, Younès Kaboul stepped back rather than closing down Young, leaving Young a massive amount of space. Both goals were well taken, but very much against the run of play. The passing visualization below shows each team's passing in their attacking third from half time through the 69th minute (when Young scored United's third goal). Tottenham clearly continued to dominate in that opening period of the second half even as Manchester United scored their second and third goals.

ATTACKING THIRD PASSING: HALF TIME TO MINUTE 69

After the third goal, United were content to pass the ball around their defensive third and the middle third of the field, and Tottenham--deflated from conceding 3 goals during a long spell in which they dominated--were content to let United pass the ball around. Jermaine Defoe scored a consolation goal in the 87th minute after a United defensive lapse, but it was of little consolation to Tottenham supporters. As the passing visualization below shows, United completed 148 total passes to Tottenham's 80 passes from the 70th minute (just after United's third goal) through the final whistle.

TOTAL PASSING: MINUTE 70 TO FULL TIME

In the end, much like Arsenal vs. Liverpool, the team that took their chances secured 3 points, while the team that dominated the run of play left the match with nothing. The visualization below shows that United scored 3 goals from their 6 attempts, while Spurs managed only 1 goal from their 18 attempts.

SCORING ATTEMPTS: WHOLE MATCH

Manchester United vs. Tottenham Hotspur, first half

Manchester United lead 1-0 on a Wayne Rooney goal headed from a corner just before the end of the half. Tottenham fans undoubtedly think that United are undeserving of a lead, especially after Adebayor's goal was called back for a hand ball. It was a very difficult decision for Martin Atkinson, as the ball clearly hit off Adebayor's arm (after deflecting off his stomach), but there didn't appear to be intent nor an advantage gained (the ball would have fell to Adebayor even if it hadn't hit his arm). Even so, Tottenham were the better side in the first half, as the figures below of passing in the attacking third indicate. The first 15 minutes of the half were relatively even, but Spurs dominated the remaining two-thirds of the first half.

ATTACKING THIRD PASSING: MINUTE 0 TO MINUTE 15

ATTACKING THIRD PASSING: MINUTE 16 TO HALF TIME

Michael Cox looks at relationship between EPL goals per game and performance of English clubs in UCL

We finished our last blog post mentioning Michael Cox's proposed relationship between goals per game over the course of a Premier League season and the performance of Premier League teams in that season's Champions League competition. In his article for the Guardian on Monday, Cox provided anecdotal evidence that as goals per game in the Premier League increases from season to season, the performance of Premier League teams in the Champions League falls during that time. It turns out Cox took a closer look at the data and published his findings on his Zonal Marking blog the following day. Plotting 3 year goals per game averages against a 3 year average Champions League performance metric, Cox finds that there does appear to be a relationship. You can read his post and look at his graphs here.

Defending a lost art in England? Comparing GPG in Premier League with European domestic leagues

Yesterday Michael Cox's article in the Guardian described the tactics of top Premier League clubs this season as "anarchic." He suggested that when England's top 5 sides have competed against one another this season, none have employed defensive, reactive tactics even after taking leads in these games. Rather, they have opted to push numbers forward freely, creating remarkably open games with defenses left vulnerable at the back. As a result, these games have been high scoring, often with the team that scores first failing to take control and kill off the game despite having a lead.

Cox contrasts these run-and-gun tactics of the top Premier League teams with those of Spain, Italy, and Germany. He suggests that in Spain there is a well-established hierarchy and teams desperately try to avoid open games against Barcelona and Real Madrid. As a result, opponents of the two Spanish giants are forced to react to the technical superiority of their opponents, play more defensively and keep the space Barca and Madrid are allowed to play in compact. In Italy, the tempo of games is much slower with teams patiently trying to break one another down. In Germany, games between top sides are cautious, tactical, and involve a lot of counterattacking.

The more cautious, tactically-minded approach of top Spanish, Italian, and German sides would seem to indicate that games played between the best teams in each of La Liga, Serie A, and the Bundesliga would produce on average fewer total goals than games played between the Premier League's top teams where games are very open. For the sake of this posting, we have arbitrarily defined "top teams" as those that currently sit in first through fifth place in their respective domestic league. Indeed, the top left graph above shows that games between the top 5 English clubs have produced an astonishing average of 4.86 goals per game. Games between the top 5 Spanish clubs have produced an average of 3.92 goals per game; games between the top 5 Italian clubs have produced an average of 2.33 goals per game; and games between the top 5 German clubs have produced just 1.93 goals per game. The data suggest the less tactics-conscious approach of top EPL teams results in more goals in matches contested between those teams.

The graphs above also provide statistics for the average number of goals conceded by the winning team per game and average goals scored by the winning team per game in matches between top 5 clubs in the four domestic leagues we've looked at. These statistics give us an idea of how cautious and defensive each league's approach to games are. In leagues that are more cautious and defensive, we'd expect the winning team to both score fewer goals and concede fewer goals to their opponent since teams in these leagues will be hesitant to push men forward and leave themselves exposed at the back. Games that ended in draws were not included in these statistics.

Incredibly, in the 11 games played between top 5 Bundesliga teams that have produced a winner (3 ended in ties), the winning team has conceded only once for an average of 0.09 goals conceded by the winning team. However, the winning team has scored only 1.82 goals per game. This is in stark contrast to the Premier League where the winning team concedes on average 1.25 goals per game and scores 3.75 goals on average. In Italy the winning team concedes only 0.43 goals per game on average and scores 2 goals on average. In Spain the winning team concedes 0.78 goals per game and scores 3.33. The data suggest games between top German sides are the most tactically defensive and those between top English sides are the least tactically cautious.

Cox ends his article pointing to evidence that English teams' refusal be submissive and focus on defending has contributed to their poor form in the Champions League. Manchester City and Manchester United failed to advance to the knockout stages. Arsenal trail AC Milan 4-0 after the first leg of the round of 16 and stand virtually no chance of advancing. Chelsea trail Napoli 3-1 going into the second leg and also face an uphill battle to advance. Cox notes:

When the Premier League enjoyed its best run in Europe, between 2004-2005 and 2008-2009, the goals per game average in the Premier League was down at around 2.5. In the previous five seasons, when England had no European Cup finalists, it was over 2.65. Since 2005, it has risen to 2.8.