Hull City tended to opt for a back 3 during most of last season, and to start this season, they went with a 3-4-2-1 in their opener, versus QPR. Robert Snodgrass’ knee injury forced Hull to change their system quite a bit during this season (and maybe this would have happened regardless of that). This season, they have gone with a 3-4-1-2, a 3-5-2, a 4-4-2, and a 4-diamond-2. Along with the lack of predictability with the system they will use, Hull’s performances have also displayed a significant level of inconsistency, both between matches and within matches. Therefore, predicting how Hull will play in any match seems like a lot like throwing darts blindfolded.
Maybe we can gain the most insight into how Hull City will play against Arsenal by taking a look at how they set up for the 2014 FA Cup Final. In that match, Hull went with a rather defensive 3-5-1-1. In front of the midfield, Stephen Quinn had the task of getting the ball up to the center forward and doing as much as he could to harass Mikel Arteta. In midfield, they had three central midfielders playing combative defensive roles. Jake Livermore, Tom Huddlestone, and David Meyler formed quite an effective screen in front of the back three. Looking at the back three, Curtis Davies looked to man-mark Olivier Giroud when he could. This left the two other central defenders free to take proactive roles in defense. If Arsenal looked to make a pass into the space in front of the back three, one of the two central defenders had the freedom to step up to make an interception, a tackle, or apply pressure on the receiver. This greater ability to cover space vertically means that Hull’s midfielders did not have to be as static with their positioning. With greater confidence (less uncertainty) that the team had enough resources to defend the space behind them, midfielders could take more risks to pressure the ball or make interceptions, even if that caused them to concede more space behind them.¹
For the first 60 minutes of the match, Hull did well to defend the center of the pitch, and while Arsenal did have a potential advantage in wide areas, Hull did well to funnel the play into the center of the pitch. This helped to cut Giroud off from the rest of the Arsenal team, preventing Arsenal from playing their normal game involving Giroud’s hold-up and link-up play.
Let us pick up the game right after Hull City scored their second goal.
We see Kieran Gibbs with the ball on the left, having just received the ball from Aaron Ramsey. Mesut Ozil is ahead of him in the space between the two lines, but Gibbs would have to have great confidence with his right foot to play a pass into Ozil, as Livermore is in a good position, even though he slipped, to prevent a pass (particularly a left-footed one) from getting to Ozil. Ahmed Elmohamady blocks the passing lane to Lukas Podolski. If there is space to move the ball into, it is on Arsenal’s right side, as David Meyler has followed Santi Cazorla into a central position. Had Kieran Gibbs’ first instinct been to make a pass to the center rather than look up-field, he may have seen a potential passing lane, between Stephen Quinn and Matty Fryatt, to Ramsey. A successful pass could allow Ramsey to potentially move the ball to Bacary Sagna, who has plenty of space between him and Rosenior.
However, Gibbs does not seem to see Ramsey as a passing option before Quinn gets in position to eliminate that passing lane. Gibbs then clips a pass to Per Mertesacker, who chests it to Ramsey.
Ramsey does not cleanly take the ball. When he finally gets it under control and faces Hull’s goal, he finds Fryatt in front of him. He has Sagna as a passing option, sideways to the right, and Arteta, sideways to the left. While there exists a large amount of space behind Fryatt, Arsenal do not have a player in that space. As I stated in the Chelsea-Arsenal preview, Arsenal tend to have a desire for verticality in midfield. That way, when the ball moves from midfielder to midfielder, the play moves up the pitch with greater speed. However, even if Cazorla came deep sooner than he does, Livermore has no problem following Cazorla. So, instead of waiting to see if Cazorla will drop into that space and get free from Livermore, Ramsey plays the ball to Arteta, who quickly shifts the play back to Gibbs.
With Gibbs receiving the ball with his back to goal, Elmohamady advances to close him down. Gibbs does not strike anyone as a player who would receive the ball with his back to goal, make the proper turn, and go past zealous defender. Therefore, putting pressure on him forces him to move back to his goal and increases the probability of a backwards pass, this time to Koscielny.
Koscielny plays it to Mertesacker, who then plays it to Podolski in a central midfield position. Livermore follows him and Podolski looks to get rid of the ball as quickly as possible, making a pass to Ramsey. Ramsey receives the ball with his back to goal, and Huddlestone looks to close him down. This forces him to move back to his own goal and then play a pass back to Arteta, who touches it on to Koscielny. Even though there existed a large amount of space behind Huddlestone, and even if an Arsenal player moved into that space, Ramsey has little chance of finding him in that advantageous position, unless he tried a risky backheel pass.
Again, Arsenal attack down the left, starting with a Koscielny pass to Mesut Ozil. Ozil finds himself with plenty of space because he drops in from the James Chester’s (the right-sided center back) zone. While Chester wants to close down an Arsenal player between the lines, he refrains from going beyond the midfield. He leaves Ozil to Jake Livermore, to whom Ozil represents a “new” piece of information, which he must integrate with the rest of his knowledge about the current situation, to come up with a decision on how to defend his zone at that moment.
Gibbs does not present much of an option for Ozil, given the proximity of Elmohamady to the Arsenal left-back. A pass into Giroud would call for the Frenchman to potentially face a 1-on-3. Therefore, Ozil opts for a pass into the interior to Podolski. If Podolski receives the ball smoothly, he has the chance to drive at Hull’s back line. This threat to the back line may give Giroud a chance to make a run on goal, breaking away from Hull defenders, who may be focusing too much on the ball. Instead, Podolski takes a poor first touch, and Livermore puts hoofs the loose ball towards Arsenal’s goal.
I will not continue to describe the play after this, but from the moment Livermore hoofs the ball, Arsenal spend about 1:10 with all of the possession and attempt only one pass through midfield line. The pass goes to Giroud, who does not successfully receive the ball due to the pressure applied by Davies. Hull regain possession; Ozil commits a foul; the referee awards a free kick to Hull, thus ending that flow of play.
Given the success that they have had with these particular tactics against Arsenal, Hull may attempt to clog the center of the pitch, have plenty of resources allocated to the back line to incentivize the aggression of their midfielders in closing down spaces and winning the ball, who have cover for the times such their aggressive approach fails. Therefore, Arsenal may want to create more space for their midfielders and some confusion in Hull’ back line through the use of their attacking front.
Depth from the Forward Line
The way Arsenal solved the problem of Hull’s defensive set-up primarily came down to the addition of Yaya Sanogo. Sanogo gave Arsenal another player who worked hard off the ball, especially compared to Podolski, and his positioning high up the pitch helped to better occupy the Hull City center backs. This reduced Hull City’s center backs’ ability to aggressively close down the space in front of them and forced them to take up deeper, more variable positions, as the combined work of Giroud and Sanogo pulled apart and caused confusion in the Hull back line.
Going back to analogy that the excess center backs served a loss-covering function (note 1), the introduction of Sanogo limited Hull’s ability to cover the losses of unsuccessful risks taken by Hull’s more advanced defenders. This may have helped to reduce the aggressiveness of Hull City’s defending, along with general fatigue, as they knew they did not have someone behind them, free to bail them out. Combine this with better movement², and the move to a kind of a 4-2-2-2, and Arsenal took control of the match.
While Arsenal will not have the duo of Giroud and Sanogo to push back Hull City’s back line, they can use a trio of Alexis Sanchez, Danny Welbeck, and Alex Oxlade-Chamberlain. Arsenal could look to push this trio high up the pitch, especially if Hull City opt for a back three (probably comprised of Davies, Chester, and Michael Dawson). Against a back three, playing three players in the highest line of attack could force Hull’s wingbacks to drop into deeper positions, lowering their threat on the counter and creating open space to the sides of the midfield line. Playing this high attacking front could help with some of the spacing issues Arsenal seem to have, which may worse without Mesut Ozil.
Without the selflessness of Ozil, Arsenal do run the risk of Wilshere, Cazorla, Sanchez, and Welbeck congregating in the left half-space, in front of the opponent’s back line (this happened anyway against Chelsea. Look at the Player Positions figure at the bottom of the Whoscored match report). The proximity of Arsenal attackers in this cluster does allow for quicker interplay and a better ability to win the ball if possession is lost. Also, with a player on the right-side playing in an advanced position, looking to make diagonal runs in behind, these relatively deep clusters of Arsenal attackers on the left could work to suck the opponent’s defenders ball-side. This would increase the effectiveness of a diagonal ball from left-to-right, looking to pick out the diagonal run of an Oxlade-Chamberlain or a Theo Walcott. However, the trade-off of this clustering is pretty clear. By shrinking the attacking space, they make themselves easier to defend. Therefore, they make it easier for the opponent to win the ball.
The depth created by Arsenal’s attacking front could prove essential in helping them maintain possession and giving their midfielders more space to operate. And it is not like the attacking front has to chain themselves to their advanced positions. By having three players in Sanchez, Welbeck, and Oxlade-Chamberlain who can also drop into midfield, Arsenal can also create quite a bit of uncertainty for defenders.
Think of uncertainty as an implicit tax placed on individuals. Say I want to make an investment, but I perceive that there will be a significant amount of variability in the interest rate, regulations, taxes, etc. over the timeframe of the investment. More importantly, I do not have a feel for the pattern of the variability over that time. So, I know there may be changes in these factors, but I do not have a clue about the number of changes, the time any particular change would occur, each one’s magnitude and direction, etc. That excessive uncertainty makes the difficult task of making a good investment harder. Maybe one kind of investment looks good now, but in a few years, changes in those factors could turn my investment into a poor one. Maybe investments that appear as poor options, under current conditions, wind up being very good investments down the line, due to changes in these factors. This excessive uncertainty not only makes it hard to make good investments, but it also makes it harder to plan, increasing the difficulty of coordinating with the investments of others (or even my own investments over time).
In a football match, where defending essentially comes down to the coordinated actions (investments) of eleven individuals, creating uncertainty helps create goals³. If Arsenal played a static two up top, with another player looking to link-up in midfield, then that predictable set-up makes it easier for a defense to plan. Hull City can determine a way of making investments that will have relatively predictable payoffs. This certainty allows them to plan (both centrally and with respect to the individual decision makers) on what to do as a team, helping them to coordinate the actions of all eleven players, making them more effective as a defensive unit.
If Arsenal have a front three where any one of the three can operate as the linker with the midfield or in the most advanced line, then Arsenal become much more difficult to defend, as the increased uncertainty they impose on the opponent increases the difficulty of making the right investments for each of the defenders. If the individual defenders cannot make the right investments consistently, nor can they predict what their teammates will do around them (how they will invest), the wrong decisions will be made more often and that crucial sense of trust among the defenders might evaporate because they cannot properly predict each other’s actions. If I have no idea as to how my teammate will react, then how can I trust him to do his job? Do I have to do his job for him? Who will take over my responsibilities if I go help him? Ultimately, the ability to coordinate the actions of the defenders, so to work as a unit, becomes more difficult, and the risk of a bunch of individuals defending in an uncoordinated manner significantly increases. This makes it much easier for the attacking team to have their way with their opponents.
I have seen a few tweets and pieces (here is one by Michael Cox) highlighting Arsenal’s propensity for dribbling. Now, the quantity of dribbles may have to do with a lack of cohesion, reducing the value of the pass-and-move game that Arsenal tend to base their game around. However, if Hull employ a similar strategy to the one they used in last season’s FA Cup Final, then the individualistic skill that Arsenal have showcased could prove valuable. In that match, sometimes an Arsenal player received the ball with his back to goal with a Hull City player closing them down. Behind that Hull City player, there existed a large amount of space. If an Arsenal player had made the proper turn and gotten past that initial defender, then the whole pitch would open up for them. They could dribble towards the back line, have plenty of time and space to play the through ball, and could occupy the attention of Hull’s defenders, helping their teammates create separation from their defenders.
Given the potential gains from dribbling in this game (yes, there are downsides, like turnovers that results in counter-attacks that lead to goals.), Jack Wilshere could have himself quite the match. According to Whoscored.com, Wilshere has made 3.9 successful dribbles/90 minutes, completing close to 50% of his dribbles. Along with Cazorla, in Arsenal’s 4-1-4-1, Wilshere could severely punish Hull City’s aggressive style of defending. With a front line creating depth by pushing against Hull City’s back line, Wilshere could find himself dribbling into quite a bit of space, with plenty of freedom to pick whatever option he sees as the best one to open up the Hull defense.
Set-Pieces and Counter-Attacks
One of the reasons that Hull City may look to adopt an aggressive and defensive approach in their own half, but not look to press high up the pitch, so to win the ball back closer to Arsenal’s goal, is because they have a willingness to rely on set-pieces and counter-attacks initiated from deep positions. In the 2014 FA Cup Final, both of their goals came from a passage of play that started with a set-piece, and they would have scored a third off a set-piece had Kieran Gibbs not made a headed clearance to prevent that goal. With the likes of Curtis Davies, Michael Dawson, Abel Hernandez, Nikica Jelavic, and Mohamed Diame all above 6’1”, Hull City have a clear advantage in the air that they may look to exploit as a part of a low-risk attacking strategy. The same goes with counter-attacks. By countering from deep, Hull City would try to exploit the defensive problems of Arsenal’s attacking shape, combined with an inconsistent commitment to counter-pressing,4 by looking to hit them on the counter, particularly with their wingbacks bombing forward to deliver crosses into the box or long balls up to the attacking front.
There are ways to combat these things. One could try to counter-press to prevent the opponent from exploiting them on the counter. Along with that pressing, one could also add a win-the-ball-or-take-the-man approach where if the ball is not won, at least a foul gets conceded to stop the play. While Arsenal do not want to concede set-pieces, a set-piece 60 meters away from one’s goal should not pose much of a threat and gives a chance for the defense to get organized. There is always the option of having a fullback sit deeper, potentially sacrificing the team’s effectiveness in attack to reduce their susceptibility to getting ripped apart on the counter. With Nacho Monreal probably playing left center back, Arsenal probably need to put a special emphasis on either mitigating the damage from counters and/or limiting the number of counter-attacking opportunities that make it into their half.
And yet, maybe not switching off represents the solution that provides the most bang-for-the-buck. As Tim points out in his notes on the goals Arsenal have conceded, a lot of the goals Arsenal have conceded are a byproduct of lapses in concentration. A team is made up of eleven individual decision makers, who look to coordinate their actions to achieve the desired result (ideally). Therefore, not only do lapses in concentration increase the probability of conceding goals, they also represent another element of uncertainty, an element that is difficult to plan for, one that can potentially impede the eleven players’ ability to coordinate their actions. So, while it may seem odd to conclude a tactics piece with a call for greater focus, sometimes something as simple as not switching off can go a long way in improving results. And for a side that has only earned 10 points from their first seven matches and is still figuring out how to play with one another, they cannot afford to throw away points due to a lack of focus.
¹If I tell you that you can make an investment (going for a tackle or an interception in this case), and I will cover your losses if the investment goes bad (this is the role the free Hull City central defenders played), it stands to reason (an assumption that people are not risk-seeking or even risk-neutral) that you will be more willing, on average, to make the riskier investment (that tackle or the interception) rather than the perceived “safe” investment (holding your position). This is because the “guarantee” that my losses will be covered increases the expected value of my action (a full guarantee cuts out losses from the distribution of outcomes), but maybe more importantly, lowers the variance (the uncertainty) of the outcome. This idea that reducing the risk associated with an action ends up incentivizing the action is the concept that can be applied to things like driving and implicit guarantees to particular institutions that may have started back in 1984 with Continental Illinois. This can also be observed when losses (or gains that are not sufficiently large) do not matter to the decision-maker. Down 1-0, it makes sense to engage in riskier strategies on average, since losing 2-0 or 3-0 is pretty much the same as losing 1-0, but drawing or winning the match is a significantly better outcome.
²A lack of aggression from Hull City may have facilitated this as well, as Hull City’s lowered willingness to follow Arsenal’s midfielders may have led to Arsenal’s players perceiving that their movement had more value. In the first half, an Arsenal midfielder could work hard, move about the pitch, and have a Hull City player dog him the whole time. He may never break free and may not receive the positive feedback (time, space, and possession of the ball) needed to reinforce that behavior and educate him as to which movements and which spaces are the ones to use and exploit, respectively. After the 60th minute, that hard work and movement leads to more success in escaping Hull City’s defenders and finding + receiving the ball in open spaces. The rather evident success of their actions means that Arsenal players receive better and more feedback concerning their movement and the spaces they exploit. Not only do Arsenal play better as a unit with this increased movement, but that, combined with better feedback, allow Arsenal’s players to reveal more knowledge about how to best rip Hull apart. That is not to say that Arsenal’s players should not have been more active in the first half, but incentives and feedback loops probably matter. Getting rewarded, as long as the reward is rewarding, when doing the right thing, and not getting rewarded when doing the wrong thing probably leads to better patterns of behavior than not being rewarded, regardless of the action.
³You could have an individual matchup that is so much in your favor that you can exploit it over and over again with success unless there is a massive shift in defensive resources allocated to evening up that matchup. However, such advantages rarely exist.
4Against a side with players like Livermore, Diame (coming back from African Cup of Nations qualifying duty with Senegal, who played two matches, including one on Wednesday), and Huddlestone in midfield, along with Dawson, Davies, and Chester at the back, forcing Hull’s midfield + back three to make quick decisions and rely on their ability to control the ball may prove a rather profitable strategy. This game could represent another potential opportunity for Arsenal to employ a high pressing and counter-pressing game to great effect.