Monday 24 November 2014

Arsenal 1 Manchester United 2. Confirmation bias is alive and well.

Arsenal 1 Manchester United 2



In his book about Pep Guardiola’s first season at Bayern Munich, Marti Perarnau recounts asking Guardiola what the best performance was from his Barcelona side in the time he managed them between 2008 and 2011. Guardiola ponders for a while before answering.

You might think he would choose one of his two Champions League final wins. Perhaps the 5-0 win against Real Madrid in the Clasico. But instead, Guardiola names the first half of Barcelona’s 2-2 draw at the Emirates in 2010 as the best he ever saw his team play, a half which finished 0-0.

Here we have the man acclaimed as perhaps modern football’s greatest coach, thinker and tactician naming a half which finished 0-0 as the best ever performance from a team which won everything and dominated European football for three seasons. If this isn’t evidence that sometimes performances don’t translate into results - but you can still appreciate it as a good performance - I’m not sure what is.

Which is why I find criticism of Arsene Wenger based on Saturday’s Arsenal-Manchester United match utterly misconceived. Honestly, it’s reached a point where people are so keen to draw an artificial binary distinction in the Arsenal fanbase between “AKBs” and “#Wengerout” that every single negative event has to be spun in such a way that if you want the manager to leave, that negative event must be his fault. Presumably people are worried that it’s intellectually incoherent to argue that you want the manager to leave and then acknowledge that some of the problems with this team are not his fault.

As it happens, I’m for Wenger staying but that doesn’t mean I defend everything he does. But to watch the United game and say defeat was the manager's fault just proves that confirmation bias is alive and well amongst many Arsenal fans. Given that the team was set up perfectly to take advantage of United’s weak defence, didn’t have its goal threatened until a highly unlucky own goal, and should have been out of sight before half-time, I find it hard to blame the manager. I do wonder whether people who are criticising the manager for this result, would do so if you conducted a thought experiment where one team played in blue, one in yellow, and they didn’t know the wider context of the two teams. I’d hazard a guess they wouldn’t.

Personally, I’d blame the loss on the poor conversion rates of the forward players. But note that while Santi Cazorla, Danny Welbeck and Alex Oxlade-Chamberlain are going through barren patches in front of goal, only Chamberlain has a habit of pinging hopeful shots from distance and even his conversion rates are not that bad. In other words, in a normal season, you would write off Saturday’s match as a statistical anomaly and move on - even Arsenal don’t often lose games in which the opposition only had one shot on target.

Unfortunately, it’s not a normal season. I think most people acknowledge that a team can be unlucky in one game, but few would accept that as an excuse for prolonged poor form. Alas, they’re wrong.

I think that’s what happening to Arsenal at the moment, and the numbers bear it out. This Grantland piece details how Arsenal had the best underlying numbers in both defence and attack before this weekend’s match - a match in which Arsenal had nine shots on target to the opposition’s two.

This sort of thing has happened before. The reason Liverpool’s title run last season was actually reasonably predictable was that they were scoring far fewer goals than would be expected per match, based on the chances created in 2012-13. Over time, things regress to the mean.


A similar story is currently available at Newcastle, where a team which had been on an unlucky run starts to turn it around. And the ongoing bad run of form is similarly in evidence at Borussia Dortmund, who blew a two goal lead this weekend and continue to sit in the relegation zone of the Bundesliga, despite dominating match after match.

And this is the issue for me with claims of it’s “same old, same old” for Arsenal. Sure, Arsenal dominated the match against Manchester United and conceded a soft goal on the counter-attack. But claiming this was like the Arsenal of, say, 2009-10 is simply untrue. Then, Arsenal let teams counter-attack repeatedly, didn’t have significantly more shots on goal than the opposition against big teams, and had a defensive midfielder in Denilson who was dribbled past over and over again. When Arsenal lost 3-1 at home to Man United in January 2010, it was difficult to claim that was unrepresentative of the match.

More importantly, that was almost five years ago. In the meantime, Arsenal won a few games against big teams but never really dominated the game. That changed in an enormous way on Saturday. Last season, against a poor United team, there was very little pressure on them in either game. This time around, Arsenal created a hatful of chances and should killed the game by half-time. That it remained 0-0 was a result of extremely poor chance conversion but in the long-run, given these players aren’t shooting from sub-optimal positions, there’s little reason to expect Arsenal’s poor results to continue.

That shows that at least in terms of the tangible, Arsenal have improved and there’s a lot to be optimistic about. Ah, you might say, but the mentality of this team is one of underdogs against bigger teams, and Wenger spins a narrative of how Arsenal are fighting an uphill battle. In other words, there are intangible, statistically unmeasurable factors at work which are impacting on Arsenal’s form.

I don’t doubt it’s possible for statistically unmeasurable things to have an affect. Certainly, once Arsenal start to win a few games, confidence will rise and momentum will return. But I find it hard to believe that this manager is a block to that occurring. I appreciate this is more opinion than fact-based, but given the way Arsenal attacked United on Saturday in a manner which I hadn’t seen for years, I reject the contention that Wenger instills a negative mentality in the players. Quite the opposite in fact. Consistently, even when Arsenal were not very good, Wenger would be in the press bigging the players up and saying they could win the League. If anything, it’s the case that he goes too far the other way.

In terms of this season, winning the League title is almost certainly a write-off. But it was only a small chance at the beginning of the season - and that was without losing our best defender for a prolonged period, best striker for a prolonged period, and best player for several months.

For me, the manager is at fault for not bringing in another defender. But equally, Chelsea have exactly the same number of first-team defenders (six). Go and check their website if you do not believe me. Neither Koscielny or Debuchy’s injuries are the result of the dreaded ‘over-training’ but rather in one case, a trauma injury, and in the other, an ongoing medical condition, for which there is limited medical treatment.

What people fail to recognise is that the issue with the defence at the moment is less one of personnel, and more one of a lack of playing time together. That’s when players get the understanding necessary to defend effectively. In time - even were Debuchy and Koscielny to remain injured - I would expect the occasional defensive error which is drifting in (quite literally, sometimes) to Arsenal’s game to lessen in frequency.

Thus far this season, while missing a series of key players, Arsenal have still dominated almost every match they’ve played. In time, the goals will come, and Arsenal will move up the League table.

Keep the faith.

Tuesday 11 November 2014

It's all about the money, money, money



Across Arsene Wenger’s 18 years in charge, there have been few consistent trends. One that does stand though is that Arsenal have pretty much always done better in seasons not preceded by an international tournament. The three seasons in which Arsenal have won the League under Wenger, he had a full pre-season. More recently, seasons in which Arsenal have challenged for the League title have tended to begin in odd years (07-08, 09-10, 13-14). The only real exception to this trend is 2011-12, and that’s fairly easily explained away by the disruption caused by the prolonged departures of Messrs Fabregas and Nasri.

Obviously, there’s a big difference between spotting a statistical trend and explaining it. As I am wont to say far too often, correlation is not causation. But in this instance, the sample size is large enough to argue with a reasonable degree of confidence that the cause of the trend is that Wenger teams do better when he has more time with them in pre-season. And that therefore, Wenger is - contrary to popular perception - good at coaching his teams.

The skill of great managers is to build a team which plays in a way that it becomes more than the sum of its parts. That’s particularly important for Arsenal: the club is simply never going to be as rich as Manchester City or Chelsea and so needs to overcome having weaker players by playing better as a team.

That’s something which is very difficult to work on during a season. I’ve really enjoyed reading ex-pro David Farrell’s blog in recent weeks, with his insider’s take on what it’s like to play as a professional footballer. This is his explanation of how training often works:

There is a general pattern most clubs will follow throughout the working week. Sunday/Monday will generally be recovery days. Light sessions of gentle exercise, five-a-sides, pool work and massage after the weekend game. Tuesday is the day when the main fitness work of the week is done. Aerobic work, strength and conditioning training, and more demanding possession and pressing sessions, alongside possible gym work in the afternoon. Wednesday is the traditional day off to allow for recovery from the exertions of the previous day. Thursday is usually when the technical work, practice matches and team shape is done, working towards the match and exploiting the opposition’s weaknesses, allied with some crossing and finishing and some small-sided games. Finally Friday, a light session, five-a-sides and possibly some set-plays. Some teams will also do some short, sharp sprint work to ready the legs and mind for the challenges of the match.

Note that this weekly schedule is entirely contingent on playing one game a week. At the point you play twice a week, you still need at least one day off, you still need some recovery time, you still need to do fitness work, but you have an extra day devoted to playing a football match. That leaves far less time for players to build the understanding needed to play together, particularly playing at the sort of intensity which is actually helpful in preparing for matches.

You can see how having a shorter pre-season for most of the squad, but particularly for key players like Mesut Ozil and Per Mertesacker, would actively diminish Arsenal’s play. Without pre-season, at a top club playing twice a week, it becomes much harder to build the understanding necessary to make a team tick. It’s one of the reasons why Santi Cazorla’s first season was so exceptionally impressive, in that he didn’t have a pre-season with his team-mates but was still consistently superb.

This lengthy goes at least some way to explain Arsenal’s travails this season. The next question is to what extent is the manager responsible. While he was never going to get a full pre-season after the World Cup, have his tactical choices exacerbated the problem?

To be honest, I don’t really buy in to the narrative of some wholesale system change this season. Insofar as the midfield three might be 2+1, rather than 1+2, yes. But I struggle to see why that should be be the biggest causal link behind Arsenal’s poorer performances year-on-year.

Probably the biggest tactical change has to been to favour more wing attacks, pushing the full-backs higher up the pitch, and asking the defensive midfield player to sit a little deeper. I think that’s a change based on personnel, but one that Wenger would philosophically want to make.

He’s an attacking coach by nature, whose philosophy of wing play is that if the opposition push players forward that leaves us space to exploit, and if they don’t, then there’s no harm in throwing players forward. And I’m not convinced those have been misguided tactics. With the personnel available, I think we were always going to ship a few goals this season. Chambers does not strike me as a full-back and Oxlade-Chamberlain is never going to be competent enough defensively to protect a full-back. So the attempted solution has been to outscore the opposition.

It hasn’t really worked, but in my eyes it hasn’t worked for two reasons. One, profligacy: the most curious aspect of the draw against Anderlecht was the over-abundance of missed chances. Against Leicester, 24 shots and just one goal. Against Tottenham, 16 shots and one goal. Against Hull, 25 shots and a scraped equaliser during injury time. Heck, even away at Chelsea we restricted them to just three shots on target - problem is, we haven’t taken our chances.

And that for me has been the biggest issue. I don’t think that’s a failure of the tactical system. I think it’s a failure of certain player’s finishing (especially Oxlade-Chamberlain) and just a bad run which is unlikely to hold in the long run. Similarly, I think Szczesny is a good goalkeeper and we will concede fewer goals as the season progresses.

The second reason for Arsenal’s problems has been a quirk of the fixture list. Arsenal have thus far dropped ten points from six games following Champions League matches, all of which have been in away games. In my season preview, I suggested Arsenal repeatedly having to play away after playing midweek would have a knock-on effect and it has. When you have a smaller squad than your rivals - at least in terms of depth of talent - it makes it far harder to rotate, and this means that your players are carrying minutes in the legs the opposition aren’t. Couple that with the fact it’s considerably harder to play away from home and it becomes especially difficult to win.

It’s one reason I expect a noticeable upturn in form over the next few weeks. After the Dortmund home game, there should hopefully be no competitive Champions League games until mid-February. This added energy, combined with the return of key players such as Theo Walcott (who is rarely guilty of the aforementioned profligacy) should give Arsenal the chance to kick on.

Ultimately, you might say that every season under Arsene Wenger is ‘same old, same old’. I don’t agree. I think in terms of personnel, Arsenal have improved sufficiently that they might well be able to beat a bigger team in the Champions League, something they have struggled with for several years.

But more importantly, even if you believe that the manager is tactically bereft, would another manager also be able to overcome the vast resource disparity between Arsenal and the two better teams in England? Because if not, you should be exceptionally wary of throwing away the benefits having Arsene Wenger as manager brings. The aspects in which Arsenal’s seasons are similar is that they do not win the League: that’s not the fault of the manager.

Keep the faith.