Amidst all the nonsense cliches which are peddled about football, by far the single biggest myth is that the league table ‘never lies’. I’m more confident than ever that the League table is a terrible way of accurately ranking the genuine quality of teams.
For all the media tells us that the ‘hallmark of champions’ is playing badly and winning, this doesn’t tend to be true. If you play shit regularly, on the whole you won’t win – but 38 games is often nowhere near enough to remove those statistical anomalies. I don’t think anybody who watched Swansea in 2014-15 and then this season would think they’re a noticeably worse team now, yet they came 8th last season and look like finishing around 15th this year. The difference is that last year they were incredibly lucky and won a whole series of matches while being comprehensively outshot (and from good positions too) – take the game at the Emirates last May as a prime example of this. This year, their conversion rates have regressed to the mean, and their League position is much more indicative of the team’s underlying quality.
I know a lot of fans think concepts like Expected Goals are nonsense but I would say two things in their defence. First, I have actually watched the matches – it’s not as if I believe that Arsenal are better than results suggests simply because of some small boxes on a picture on Twitter. Watch back the matches against West Brom away and Southampton, Swansea and Palace at home and if you really think that on the basis of chances Arsenal only deserved to take two points from those four games you’re watching a different sport from me. There are other examples I could add to this and there are almost no games (perhaps Newcastle at home) which can be placed in the ‘undeserved win’ camp to balance it out.
The additional reason I’m a fan of statistical concepts is they remove all the pre-existing biases fans have. It’s natural to remember the colossal fuck-ups and not remember the many times your team actually did hold on to a lead. And it’s also natural to remember things which happened recently more vividly. The outcome of this is that when your team has blown a lead recently (like at West Ham), and hasn’t been getting very good results, people come to view this as much more likely to continue than it actually is – and so despite Yannick Bolasie’s equaliser for Crystal Palace undoubtedly being a very low percentage shot and one any decent goalkeeper should save, all these other irrelevant factors come into play in judging what happened. So fans saw a Palace equaliser as ‘inevitable’, despite them having one shot on target in the preceding 80 minutes.
So I projected the league table based on xG this season. You're gonna be shocked but Arsenal has underperformed. pic.twitter.com/jE9caeyJGW— Michael Caley (@MC_of_A) April 21, 2016
And the fact is, almost all statistical models (based on chance creation/negation) out there have Arsenal as one of the two best teams in the Premier League this season. Dreadful conversion rates (and from players who have no precedent for this) have meant Arsenal’s performances not getting the results you would expect them to. It easily goes forgotten amidst a poor run of form that between August and November, Arsenal were by far the best team in the League – while there were a couple of horror shows in the Champions League, this was a team which went 3-0 up against Manchester United in twenty minutes, went to Leicester and hammered them 5-2, and produced one of the single most dominant performances of the season when they had 29 shots against Stoke. And after the Dinamo Zagreb and Olympiacos calamities, there was the small matter of deservedly beating the second-best team in Europe at the Emirates.
If you think I’m biased that Arsenal were playing really well, go to a neutral judge – Arsenal went off favourites to win at Stamford Bridge with every bookie, and this was well before the extent of Chelsea’s slump was apparent. It was based on Arsenal’s incredible chance creation stats.
So my first defence of the manager is that he has effectively just fallen victim to random statistical variance. While obviously I cannot prove a counterfactual – and that’s the beauty of football, that 38 games isn’t a perfect sample size and there is a far greater element of randomness than in other sports – I think that had the team performed to par with chance conversion, none of the current hysteria would be ongoing. And if you think that saying the performance matters more than the result is the thinking of some crazed Wenger acolyte, you might like to know that two of the most highly respected managers in Europe, Pep Guardiola and Mauricio Pochettino, have said exactly the same thing.
Goalkeepers haven’t suddenly played completely differently against Arsenal, the team’s style hasn’t rapidly evolved, the players underperforming chance conversion aren’t getting old and losing their skills and nor are they new players who we lack data on – all of which overwhelmingly suggests to me this will regress to the mean with Arsenal scoring more goals going forward.
Tim Stillman suggests in his column this week that the slowing of Arsenal’s build-up play means Expected Goals isn’t accurate (and explains chances being missed) but if you read Michael Caley’s explanation of his methodology, he outlines how speed of attack is a vital part of his calculations. In addition, it’s notable over the last few weeks that as Arsenal’s build-up has become more laborious, the Expected Goals numbers have dropped – much of the under-performance of Expected Goals came when the team was right at the top of the League in the first half of the season. The point was that a much larger buffer should have been created to allow for some of the dross witnessed in recent weeks.
So for me, certainly, you could change the manager – and the team would probably achieve better results with a sensible appointment. But that would likely not be an outcome of the new manager’s appointment, but simple regression.
Clearly though, despite these low conversion rates, Arsenal could have been better. The difference between me and a lot of other people is I’m not convinced the role of manager is actually anywhere near as important as is generally suggested. Ultimately, a lot comes down to trusting the players. There’s a story Aidy Boothroyd, the former Watford manager tells, of how one of the best pieces of advice he was ever given came from Wenger. The Frenchman told him he couldn’t possibly get to each player at the end of a match and ask them to evaluate their performance, so players themselves have to be good coaches and understand where they’ve gone wrong.
If you look at Arsenal’s struggles this season, lots stems from absolute idiocy on the part of the players. Olivier Giroud diving in on a yellow card away in Zagreb. Gabriel reacting to Diego Costa’s nonsense. Per Mertesacker diving in on Costa 45 yards from goal, fully aware he was the last man. Francis Coquelin making slide tackles while on a yellow card while Spurs were going nowhere. Each of these came at the cost of a red card, none of them require hindsight to suggest they were terrible decisions by the players, and none of them were typical of the players, to indicate negligence on the part of the manager for picking them. It’s difficult to explain why each player made their chosen decision, but it requires some serious logical gymnastics to suggest that it was the manager’s fault.
And even to the extent the manager has made mistakes, this puts him in a bracket alongside every other manager currently operating in world football. Current flavour of the month Pochettino rested his best player for a crucial Europa League game away in Dortmund so he’d be fresher to face this season’s Aston Villa, the worst team seen in the Premier League for some time. Luis Enrique, the man who won the treble last season, proved unable to actually manage Lionel Messi, Luis Suarez and Neymar’s workload, with the outcome that Barcelona were short of fitness at the crucial part of the season and blew the Champions League and possibly La Liga. Pep Guardiola’s team has been similarly beset by injury, at least partly a result of a lack of rotation and his very physically taxing style of play.
I’m the first to say that bringing back Alexis Sanchez back after just a few days of pre-season was unhelpful and likely contributed to his iffy form, if not the only reason behind it. But even in that case, it’s not unjustifiable or negligent as some people put it – a team which was struggling for a result is likely improved by having one of its best players present. While it doesn’t require hindsight to suggest bringing him back early was going to have negative repercussions, it was never the plan but for going 2-0 down at the Emirates to West Ham. I’m not sure he’s entirely sure who his first-choice right midfielder is but not being certain of your first-choice starting line-up is hardly the greatest weakness in a manager. Similarly, while I would have liked to see another midfielder purchased last summer, I don’t regard that as the ‘great failure’ some like to paint it as.
Michael Keshani lists some other criticisms in this thorough piece - in particular with regard to squad building, team selection, and transfers.
In terms of squad building, the main criticism is the failure to buy a player to replace Francis Coquelin in the starting line-up before this season started. While I’m far from his biggest fan it’s notable that the team has achieved excellent results with him and Cazorla playing next to each other in central midfield and were top of the League through doing so.
The additional point though is that to the extent he should have been replaced with a better player, there’s only so many people you can reasonably put on the wage bill to play one position. With Coquelin, Mathieu Flamini and Mikel Arteta all already in the squad, it was difficult to justify signing somebody in the summer. Any sophisticated analysis of how Pep Guardiola will do at Manchester City acknowledges that although he is a good manager, it’s unlikely he’ll be able to change the squad to entirely meet his desires in one transfer window, because players have existing contracts and are often difficult to move on. In much the same way, even to the extent the argument can be made that it was necessary to replace Coquelin before this season – one which is dubious at best – that does not mean it is something which was possible to do.
Critically, the exact same people who are so unequivocally critical of Coquelin suggested Daley Blind as a potential DM for Arsenal – a man who has endured many of the same problems with distribution that Coquelin is slated for. It’s almost like buying good defensive midfielders is really, really difficult.
In terms of team selection, the main criticism seems to be playing Gabriel and Laurent Koscielny together. Given, teams kept on coming to play Arsenal and sitting deep and looking to hit the counter-attack it made sense to play a pacier defensive pairing which could play higher up the pitch, allow the whole team to take up a higher average position and hopefully force more chances. You can dissect at length whether this was a particularly successful tactic, but having seen Mertesacker sent off against Chelsea while trying this and then him ending up caught out of position, it’s hardly a bizarre tactical decision to implement.
Even with regard to transfers, I don’t share the view that there is no coherent plan to build a good squad. People who work in player evaluation have suggested that Arsenal are probably the best in the Premier League when it comes to transfers. While the churlish might suggest it doesn’t take a genius to work out Mesut Ozil, Sanchez or Petr Cech are particularly good at football, the repeated failures of other teams in the transfer market while buying established talent suggests Arsenal’s track record deserves credit. Matthew Benham, the owner of Brentford and FC Midtjylland and a pioneer of using stats to complement the naked eye acknowledges that “no matter how much research you do, you get a couple of duff ones” underscoring how well Arsenal have done looking at things in the round.
My impression is also that Wenger retains his ability to spot a good player – witness him signing a 19 year-old right-back from Southampton who had twenty professional appearances and converting him to an excellent centre back who might well play the position for Arsenal for ten years, or his ongoing commitment to signing young players like Gedion Zelalem and Jeff Reine-Adelaide. Arsenal’s transfer policy of combining signing established world-class players with buying up talented youngsters has helped to foster an exceptionally talented squad and to ignore that simply because Cech was the only outfield signing last summer is absurd. Certainly, to only credit the manager for this is simplistic and many of those involved in transfers would continue to work for the club should there be a change of manager. But given the control almost everybody acknowledges Wenger has at Arsenal, a reasonable chunk of the credit belongs to him.
In his preface to Michael Calvin’s excellent book Living on the Volcano, Wenger outlines four things he sees as fundamental to being a successful football manager: the passion an individual has for the sport; an eye for talent; man management; and the ability to evolve and adapt.
It’s pretty obvious that if Wenger no longer had the passion, he would not seem so enraged by how this season has panned out and would probably just walk away. Similarly, the discussion of transfers above shows he still has a good eye for talent. Two questions remain: is he still a good man manager? And can he still adapt to the modern game?
The man management question is particularly important. Wenger’s strength as a coach has long stemmed from his ability to improve individual players on the training ground and his capacity to motivate them. For what it’s worth, I regard the talk of him losing the dressing room as complete rubbish. The evidence seems to be that Ramsey and Nacho Monreal argued a bit on the pitch against Sunderland last week. The problem is, this sort of thing is always judged to feed pre-conceived ideas. I guarantee that if the team were winning, this sort of on-pitch argument would be shown as a manifestation of the team’s determination, will to win and strong mentality. People said exactly this about Coquelin shouting at his team-mates to focus early in the season; the same is probably true now.
More pertinently, several players have shown distinct improvement this season – in particular, both full-backs, Alex Iwobi and Danny Welbeck. While I remain a little dubious about how coachable certain players are (especially Jack Wilshere), Wenger’s coaching and ability to motivate the squad remains strong. What’s been pretty notable to me is how little any of the players have said against the club or manager, despite the remarkably poor results over the last couple of months. It’s fair to say they acknowledge that the responsibility is largely theirs.
Finally, in terms of evolving to the modern game, it’s really interesting to me just how far Wenger has gone. He is by far the most outspoken user of analytics amongst Premier League managers and was certainly part of the process of the club purchasing StatDNA. It’s difficult to know exactly what Arsenal are doing from a technological point of view as for obvious reasons it’s kept very secret, but in terms of adapting the team to the modern day, Wenger remains a pioneer. Even with tactics, the decision to stick with the 4-2-3-1 stems from it working quite well, rather than his ability to innovative tactically disappearing – there are also quite large drawbacks from constantly playing a high press (for example), much those these tend to go unacknowledged.
Ultimately, my view is Wenger remains an excellent manager, who continues to make good signings and improves the players he works with. He has long been tarred with the brush of ‘not doing tactics’. While clearly nonsense, his focus has always been on giving the players a certain freedom to play within the system and for the most part, I think that’s been quite effective this season – with the caveat that multiple injuries to crucial players have somewhat broken the system.
Nobody lacks perspective quite like Arsenal fans. Redaction, the club’s official supporter group, writes in the biggest Arsenal fanzine this week of how “there really seems to be no light at the end of the tunnel” for a team which continually finishes in the top four of the Premier League and reaches the knockout stages of the Champions League. To almost all outsiders, the reaction is one of bemusement or annoyance. Arsenal operate on the fourth biggest budget in the League and it’s precisely because the club is run as a model business that they can get as close to the teams on bigger budgets as they do.
This season has almost certainly been a random statistical anomaly. Spurs are certainly much improved under Pochettino but Leicester’s run stems from incredibly high conversion rates at one end of the pitch, and opposition conversion rates so low they are historically unprecedented across the course of an entire season. Most people buy into the nonsense that the League table does not lie, but the reality is that it does and while richer teams like Chelsea and Manchester United have endured terrible runs, Arsenal have done pretty well this season and have been better than Leicester.
It’s more than a little frustrating that given how badly the richer clubs have done Arsenal haven’t taken advantage, but I remain utterly unconvinced this was even to a medium-sized extent down to managerial errors. More broadly, whilst it’s not particularly romantic, that the team with the fourth biggest budget continues to finish third or fourth in the League while suggesting more might be possible (and picking up a couple of cups on the way) is actually a sign of the club being managed well, not badly. It might seem a little boring but given fans have witnessed so many other teams change manager and regress, the upward trajectory visible at Arsenal is one I’d rather stick with. Wenger remains an excellent manager and I hope he stays for next season.