The Arsenal Defense And The Jay-Z Phenomenon


What does Arsenal’s defense have to do with currently famous American rap-artist, Jay-Z? Apart from him once posing in an Arsenal strip, there is a tangential connection. Stick around and I will share with you.

I will start with the premise that there is a large body of Arsenal fans hanging fiercely to the belief there is a defensive calamity lurking around the corner which, sooner or later, is somehow going to derail any title ambitions the club may have.  To dispel the notion this may be an exaggeration on my part I recommend you take a gander at the post-Ludogrets Arsecast by the Sage of Dublin to get a feel for this type of thinking.  By way of a disclaimer, my focus on The Sage is not with any malicious intent. To the contrary, he merits this attention because he has the biggest online presence among Arsenal fans, and in my opinion reflects and projects the feelings of a large swathe of the fanbase.

Back to my main point, it struck me from the podcast, despite the many encomiums and panegyrics devoted to Mesut Özil’s fantastic last minute goal, there was an underlay of discontent with the team’s defending. Ludogrets it seemed had no business scoring on the mighty Arsenal even though they had a bevy of slick, speedy, tricky Brazilians as attackers. This led him and his guest to conclude it will soon become a “pressing concern.”

This thinking is entirely consistent  with a recurring meme over recent years; Arsenal may play beautiful attacking football but can’t defend if their life depended on it.  The mainstream media have been happy to play up this fear with pundits and journos quick to jump on any defensive error to send Arsenal fans into panic. Not to be left aside, the majority of Arsenal bloggers, podcasters and tweeters whoring for RTs quickly fall into lockstep. Apparently memories of Squillaci and Senderos conceding some stupid goal or being bullied by an attacker have blinded many to actual facts and data.

Surely, if the meme was true, then Arsenal must have one of the leakiest defenses in the premier league, at least over the past 10 years, which by the way coincides with Wenger’s anni mirabiles since the glory years at Highbury.

Year Lge Pos GA GA Rank
2015/16 2 36 4
2014/15 3 36 3
2013/14 4 41 4
2012/13 4 37 2
2011/12 3 49 8
2010/11 4 43 4
2009/10 3 41 5
2008/09 4 37 5
2007/08 4 31 4
2005/06 4 37 3
Mean 4 39 4
Median 4 37 4

The table surely speaks for itself, doesn’t it? On average we performed like a 4th place team; ranking 4th in Goals Against whether on a Mean or Median basis, despite the calamities of Squillacci or Senderos.  Also apparent is that since 2012/12, when the defense was clearly a leaky sieve, conceding 49 goals, there has been a steady and gradual improvement to a stable 36 GA in the past two seasons, i.e. at 0.95 per game. At the beginning of 2010/11 the club signed a clearly promising but inexperienced defender in Laurent Koscielny and later paired him with the veteran international Per Mertesacker in 2011/12, and despite a rocky start they have over time been able to forge a reliable partnership which is at the root of the current improvement.

If Arsenal has been a 4th ranked team in goals conceded, how important is improving the GA in advancing up the tables and winning the title? As usual we rely on the data rather than being driven by panic when multiple goals are conceded, as was the case after Liverpool scored four goals to beat Arsenal in this season’s opener.

Year Lge Pos GA GA Rank
2015/16 Leicester 36 3
2014/15 Chelsea 32 1
2013/14 Man City 37 2
2012/13 Man Utd 43 4
2011/12 Man City 29 1
2010/11 Man Utd 37 3
2009/10 Chelsea 32 2
2008/09 Man Utd 24 1
2007/08 Man Utd 22 1
2005/06 Chelsea 22 1
Mean 31 2
Median 32 2

As usual the data maybe silent but speaks loudly. It reveals a definite trend where league winning teams are conceding more and more goals. From a low of 22 GA in 2005/06 and 07/08, for Chelsea and United respectively, the GA dramatically increased by 45% last year when Leicester conceded 36 goals. Man United was profligate in conceding 43 goals in 2012/13 but won the League relying on Van Persie to outscore the opposition (more on the importance of goal-scoring). The Mean and Media data clearly demonstrating  that teams who win the title tend to average 2nd in the GA ranking with 31 or 32 goals respectively. In other words title winning teams do not need the best defense in the league. (How loud should I shout that?)

Surely then there is no basis for early-season sensationalism about a poor defense when the club merely needs to reduce the GA by 4 goals, year-on-year, to hit the average GA of the last 10 title -winning teams. By signing Xhordan Mustafi, a defender,  for the second highest transfer fee ever in his tenure at the club, i.e.  £35 million, Wenger was making, to use that hackneyed cliche, a statement of intent. I already argued  in a prior blog, with the use of historical data, that this was the most important signing of the last transfer window.

As  already alluded, the data indicates that to win the league a club must consistently outscore the opposition and Win the vast majority of games. Draws will not cut it. This is strikingly obvious based on the massive point incentive for winning games; a 3:1:0 ratio for Win:Draw:Loss.

This is confirmed by an analysis of the data from the league winners of the past 20 years, which  provided the following statistics.

Mean 79 32
Median 77 34
Std Dev 10 8
Mean Absolute Dev 8 6
% Std Dev  13% 24%
% Mean Absolute Dev 10% 19%

Without being too technical, the data is saying 87% of the time, an average of 79 goals will win a club the title. In comparison, there is only 76% probability that the average 32 Goals Against will guarantee a title. (Just a reminder that anything less than a 90-95% probability is useless as a reliably predictive statistic.) The data is providing confirmation that a club with ambitions of winning the premier league must emphasize goal-scoring while being at best #2 in defending. There have clearly been exceptions over the years, for example Chelsea under Mourinho who have given defending as much priority as goal-scoring.

I am fairly confident that despite the data and statistics, the majority of fans will cling grimly to their fear that Wenger is stuck in his “dated” Highbury philosophy and will give insufficient attention to defending and keeping clean sheets. Mind you Cech was golden glove winner in 2015-16 and in 2013-14 Wojciech Szczęsny shared the same prize with Cech.

I can only ascribe this to Confirmation Bias. According to Psychology Today:

(This) “occurs from the direct influence of desire on beliefs. When people would like a certain idea/concept to be true, they end up believing it to be true. They are motivated by wishful thinking. This error leads the individual to stop gathering information when the evidence gathered so far confirms the views (prejudices) one would like to be true.

“Once we have formed a view, we embrace information that confirms that view while ignoring, or rejecting, information that casts doubt on it. Confirmation bias suggests that we don’t perceive circumstances objectively. We pick out those bits of data that make us feel good because they confirm our prejudices.”

This I am afraid is more widespread than most of us at Positively Arsenal would care to admit.  This leads me to the only commentary I will make on these pages concerning the US presidential campaign, now coming to a close. I have observed many of my friends join lockstep with the Democrat candidate who made the infamous statement:

“To just be grossly generalistic, you can put half of Trump supporters into what I call the basket of deplorables,”

“Right? Racist, sexist, homophobic, xenophobic, Islamaphobic, you name it.”

Yet the thousands of Wikileak dumps indicate the deplorable meme could be fairly applied to the originator as well as Mr Trump. To put it mildly, it is easily a case of a candidate “seeing the mote that is in thy brother’s eye, but considerest not the beam that is in thine own eye”.

If you believe the Wikileaks are somehow tainted, even though their authenticity have never been questioned, then I commend to you last Friday’s get-out-the-vote concert in Philadelphia featuring Ms Clinton. In the run-up to her speech, the sometime gooner and rapping phenomenon Jay-Z left nothing to the imagination.

The rapper’s repertoire included his hit ‘F**kWithMeYouKnowIGotIt.’

He also performed a song called ‘Jigga My N***a,’ including a line that declared: ‘[I’m] Jay-Z, motherf***er!’

As he took the stage, a PA announcer blared: ‘You’re tuned into the motherf**king greatest!’

‘Ladies is pimps too, go and brush your shoulders off. N***a is crazy, baby, don’t forget that boy told you. Get that dirt off your shoulders.’

I don’t f**k with you. You little stupid-a** b***h, I ain’t f**kin’ with you.’

Racist, sexist, misogynist …. you name it. Yet Jay-Z is not pilloried as a “deplorable”. Isn’t this a classic example of what Psychology Today describes of us becoming prisoners of our assumptions of what Ms Clinton and the Democrat party represents and simply rejecting any data to the contrary?

Give this inevitable intersection between football and politics I urge my readers to honestly reflect to what degree we may be afflicted by old precepts and assumptions that are not rooted in reality. (I almost said data.) And may this blog remain as independent-minded as ever.

50 comments on “The Arsenal Defense And The Jay-Z Phenomenon

  1. Great article ..very enlightening.
    One of the reasons for the increasing GA trend could be that a lot of mid table teams have had decent attacking players over the last few years (Michu, Mane, Bojan, Darren Bent…).
    I’m not a statistician ..but one concern which I have with aggregating GA in isolation is maybe a loss of context in certain situations. For instance, I feel it also matters “when” the goal(s) have been conceded. Lets say we had won 1-0 against spurs and won 4-2 against Sunderland …it leaves us with the same number of goals conceded ..but we earn more points. Similarly if we win 5-2 or 6-2 , the 2 goals conceded have less of an impact as opposed to a 2-2 draw. Can/Should this be factored in somehow?

    Liked by 5 people

  2. I am still not sure what Jay-Z has to do with Arsenal ?

    Unless I have missed something Jay-Z is not standing for any “office”, nor is the Arsenal defence likely to be subject to the popular vote any time soon.

    Never say never though.

    Liked by 3 people

  3. Confirmation bias Andrew, never forget.

    Liked by 3 people

  4. well the signs on our defense this season is that it has improved, kos and mustafi have not conceded a league goal from open play, and the second goal v ludograts was the one goal scored from open play since the partnership formed.
    I can only assume that the speed at which we conceded the second goal so soon after the first goal v ludograts, was the cause of concern for many, as it brought back a failing we had last season, – once we conceded one goal we were likely to concede a second. And often within a short period of play. In fact once we conceded a goal we were more likely to concede a second goal or more than keep it at one.
    We lost one game last season 1-0, and had three 1-1 draws. we won 8 games where we conceded one goal, we conceded more than one goal 17 times last season. In those 17 games where we conceded more than one goal we won just the one of them, 2-5 away at LCFC. We kept 23 clean sheets, winning 18 of those.
    We have played 17 games this season, losing none so far when we conceded one goal, we have won 2 games conceding more than one goal, and we have lost one game conceding more than one goal. 8 clean sheets, winning six of those.

    Liked by 2 people

  5. I have the time to make a comment this week Shotts!

    Brilliant, comprehensive analysis. Thank you.

    I get the feeling that, vaguely inspired by Zimpul, that this series is building up to a analysis of the L column which ultimately is more in important then GA or GF or GD, although they are and always have been useful measures.

    That’s why the gleeful goading of the gullible and the depressed by the pro trolls is a sad sight after a draw in November in a league hich finishes in May. Confirmation bias might be too gentle a term here! What would an anthropologist make it all?

    Your ellections will affect all of us. The interview between Pilger and Assange will get little or zero coverage upon our etherwaves. It didn’t happen as Pilger would say. Even though the interview contains an incontrevertible nugget of information which for many here and across the pond would be the most shocking bit of data that they will have exposed to in the last fifteen years. And it all does tie back to football, eventually, the World Cups and the wonderful Fufa etc.

    Liked by 2 people

  6. Wow, Shotta, that was a tour de force and clearly shows your grip on the subject.

    As someone who relies on statistical analysis in a working environment, I occasionally hear a rejoinder when the other party does not like what the stats evidence points to, and tend to quote Mark Twain that there are three kinds of lies: Lies, damned lies, and statistics.

    I have heard it many times, and it always makes me grin as it generally means the other party is grasping at straws. Of course the problem with statistical data when used to prove something, is that it relies on careful analysis and interpretation, and to that end it also enables other dissenting voices to insist their interpretation is more meaningful, even where that is patently untrue.

    Anyway, well done, and your interpretation is fine by me.

    Liked by 2 people

  7. Not surprisingly, the Sage of Dublin and his sidekick, Gunnerblog, have, in their post-Tottenham podcast proved me right. For a perfect example of confirmation bias check-in at the12 minutes mark where they discount 7 clean sheets and proclaim their fear that AFC is “suspect” defensively and not “solid enough”. This is separate and apart from the fact they have absolutely no awareness of 3.5 years of data that Santi Cazorla makes a decisive difference in winning games for Arsenal. Over 2 ppg when he plays, less than 2 when he doesn’t.

    Fear sells. Not facts and data. That is why the majority of punters get slaughtered in the stock market.

    Liked by 2 people

  8. Nice one Shotta – just like the “Arsenal always collapse in November” story being repeated ad nauseam in the press and by commentators despite a lack of evidence this hackneyed old chestnut about Arsene not doing defence gets trotted out time and again.
    Nice to see it dispelled.
    It may be that statistically November is the month we garner fewest points taken on average over our history but there will be a month like that for every single team won’t there? Bound to be unless every team gets the same points every month for ever of course. Strange we don’t hear which month is Liverpool, Chelsea, Burnley or West Brom’s weakest, Why only Arsenal? Can’t possibly be a negative media agenda aimed at putting pressure on the manager and players of course not that would be paranoia to think that.

    Liked by 5 people

  9. HenryB @ 11:50 – Thanks for your kind words. This is the sole reward we get from blogging. As for the “lies, damn lies and statictics” meme, when I read the manipulation of statistics by politicians and the mainstream media, I understand Twain perfectly. That is why I emphasize data. The raw data has no agenda and no headlines. Rather it is cold, hard and objective. It is for us to draw objective conclusions.

    Liked by 1 person

  10. More respect for Senderos required . He and Kolo Toure were the central pairing that took us to our only Champions League final and held out for 995 minutes ( yes 995 minutes) without conceding a goal in the competition – 27/09/05 to 17/05/06.

    Its all here for those whose taste has been wetted;

    Click to access 2285688_DOWNLOAD.pdf

    Liked by 2 people

  11. As soon as I read the first paragraph I new this was going to be a corker, well done Shotts.
    I think our experiences will affect our bias even though those experiences may not reflect the situation in the wider world.
    It is probably why I look at the refs mistakes whereas A5 will look for the positive decisions although I was never traumatised by a ref as a child.
    On a serious note confirmation bias is why Murdoch and his ilk have so much sway in the world and why if they said the queen was an alien the majority of folk would believe it.
    As far as the defensive issue is concerned we are always compared with the best and so fans believe that we are actually a disaster. If asked the average position we finished defensively over the last ten years the perception would be around mid table ( same as the perception % of immigrants is also widely inaccurate.
    Basically if you tell someone something often enough in the end they will believe you.
    Shotts for president, Steww for PM, George for minister in charge of fury creatures and A5 in charge of the department of Men in black.

    Liked by 3 people

  12. Shotta analysis is very detailed and emphatic, confirmation bias is real. But I thought this was positively Arsenal , I do always say football reflects life like no other sport. With it’s ebb and flow, unfairness, blatant cheating, crudeness,cruelness unforgiving unabating,loud,quiet,politics and some oh so brief moments that are otherworldly. So i guess you have made a tenuous link. My problem with Donald Trump is he will lie about anything and everything to get his way. He will lie in the morning, deny at lunchtime, then deny the denial,by tea-time then lie . He will try to be a dictator he his showing the traits already and small or minor dissension will be dealt with in the most severe of force. He has vowed to make it so,maybe he was joking maybe not. This is a man with paper thin skin, criticism will be met with force. I know there are some problems with Mrs Clinton ‘but Jay-Z is not one’, he is busy running his own corporation and doing quiet well from what I have seen, I wonder what the percentage of businesses that Jay-Z started that have gone into receivership/foreclosure, compared to Donald Trump?.

    Liked by 3 people

  13. Woah! Another beauty Shotts!

    Who is the Sage of Dublin?


  14. George at 1:38 pm: Shame on you! Sage of Dublin? I have given you so many clues.

    Liked by 2 people

  15. Correction above Shotts the quote “it never happened” was from our very own Harold Pinter in his last prominent and inspiring talk.

    “It never happened. Nothing ever happened. Even while it was happening it wasn’t happening. It didn’t matter. It was of no interest.”



  16. No idea why some posts from When We Were Boring and Rich’s yesterday ended up in the Pending filter. I try to keep an eye on what is waiting, best I can do so don’t be put off.

    Liked by 1 person

  17. A5 @ 2:24 pm: Even HenryB got caught in the filter. I had to release him. That’s very strange.


  18. Ladies and Gentlemen of a Positive Persuasion,

    Over the next 12 to 24 hours the US Presidential election will be completed and the 45th President elected, whoever he or she is.

    I am aware that the election has caused, perfectly understandably, considerable passion among many people in the US and elsewhere with regard to the outcome, and the personalities involved. We may be a haven of sanity in a crazed world, but no man is an island.

    IF Positive subscribers wish to discuss the outcome of the election rather than football and Arsenal over the next 24 hours I am sure I can rely on you to conduct those conversations in a respectful manner, in line with our invariable good practice.

    Liked by 4 people

  19. for fucks sake, if ever any more proof was needed that anything that is favorable for Arsenal is treated differently than for any other club here it is folks, the BPL have actually released a statement explaining why the Wimmer OG was allowed to stand. Has anyone ever seen the BPL release a statement to explain why any other goal in the history of the PL was ok. Even one goal will do.



  20. Alexis pulled out of training with Chile today, with a “minor muscle injury”, he is expected to still play in their two upcoming games


  21. just seen an update on the alexis injury, which says he is out of the game with Colombia with a calf injury, but is expected back for game v Uruguay


  22. A statement from the ANFP, Chile’s national football association, read:

    ‘The medical team of the Chilean national team can confirm Alexis Sanchez has a low grade muscle tear.

    ‘The player will stay in Santiago, under treatment and evaluation, ahead of Chile’s match against Uruguay.’


  23. That is bad luck for Sanchez, he loves his international games.

    Does anyone know when Perez is back ? After his injury against Reading it was said by the club he would be to to the end of October. Assuming that is right the Spaniard should be fit by the end of the international break.


  24. not sure when lucas is back but he has been given permission to spend the international break back home in spain, and arrangements have been made for him to receive treatment for his injury from the Deportivo medics.


  25. perez is back in January, same time as Danny.


  26. I thought Perez was out until December.


  27. Bellerin missed spain u21 training today, due to “a little discomfort”


  28. Bloody clubs been lieing to us again;



    Lucas featured twice in October, starting the 2-0 win over Reading in the EFL Cup fourth round and coming on as a substitute as we beat Ludogorets 6-0, laying on two assists for Mesut Ozil. He picked up an ankle injury in the win over Reading at Emirates Stadium, ruling him out for the rest of the month.


  29. End of the January window ?


  30. I don’t care – Sanchez is a beast – he will be back for United.


  31. That’s bad luck for Alexis indeed, but with luck Monsieur Giroud can strut hsi stuff in his absense. What news on the fitness of Sanogo should we need a plan C off the bench?

    Liked by 2 people

  32. Paddy Power ‏@paddypower 23m23 minutes ago
    Trump has basically just said on Fox that if he doesn’t win it’s unlikely he’ll accept the result… sounds like a Man Utd fan.


  33. I see trumps legal team already bringing court actions on voting, funny its in a mainly hispanic voting area, it seems the judge is throwing it out. trumps side trying to claim election is rigged etc.


  34. as Lucas prognosis was supposed to be 8 weeks I was expecting mid December at the earliest. Our game at manure is an early ko on saturday so as Alexis doesn’t normally get back until friday I thought him, Iwobi and Elneny would be benched anyway.
    If he has a small muscle tear then we may not see him at all. That would mean Ox on the left, Theo on the right Oli up front with Mezut behind and a center midfield of Santi and Le Coq with Xhaka and Aaron waiting to supplement the midfield at any time. If you consider what the restrictions might be thats still a brilliant squad.
    I think it would be interesting to hear kelly’s take on the election coming from a republican area.

    Liked by 3 people

  35. Spent most of the morning in my Panic Room cleaning and getting it into order – is it safe to come out yet though ?

    Liked by 1 person

  36. at the moment I’m more worried at the reports that Alexis is willing to play injured for chile v uruguay in a few days time

    Liked by 2 people

  37. He wont let a little thing like a torn muscle put him off eddy – he will be out there hurtling up and down, leaping clear of the Uruguyan psychos

    Liked by 2 people

  38. just seen that just like the brexit vote, the usa election result was down to how the over 45’s voted, the under 45’s were for remain in the uk, and for clinton in the usa.


  39. well reports suggest alexis has a torn calf muscle, and that is normally six weeks out, so if accurate, we won’t be seeing him again till new year, but even if it is a torn calf muscle, i would not put it past him and chile for him to play for them next week.

    Liked by 1 person

  40. Do you reckon these polls that purport to record how people voted are any more accurate than those earlier in the week that claimed to show how people intended to vote ? The pollsters called the last General Election in the UK badly wrong, the Brexit vote equally wrong, and now the US Presidential election wrong.

    What reason could we have to put any faith in anything thing they say ?

    If I was a pollster today I would STFU to be honest eddy and hope people have short memories ( which apparently they have)

    Liked by 2 people

  41. Ben Dinnery ‏@BenDinnery 1h1 hour ago
    Hector Bellerin [#AFC] is replaced by #SAFC full-back Javier Manquillo in the #ESP under-21 squad.


  42. anicoll,

    Do you really care what the stupid ‘predictors’ tell us? lol

    To give a percentage of how many people will vote/do something they have to ask the people concerned. The result depends on what they said, and if they were telling the truth to some anonymous person on the end of a random ‘phone call.

    Hardly a reliable system – as you pointed out.

    I have never been interested in such data — and if the public all got bored with such a poor forecasting system, they would instead try and interest us in how many times voters farted on the the way to the polling booth. Just as inaccurate and pointless but if it gets the TV people showing us all singing all dancing graphs on the screens and boosting their income they will be happy.

    Back to Footie reality: Jack seems to be talking himself out of Arsenal – and that would be a pity for him – probably more than for us, tho’ I would be sorry to see him leave for pastures new.

    Liked by 2 people

  43. Thank you Andy, another item for my UEFA file!

    Most games played – Arsenal at No.5 with 173, and that of course, leads to

    Most defeats – again No. 5 with Arsenal – 51, which of course, leads to

    Most goals conceded, again at No.5 Arsenal – 192.

    The Clintons are not very popular in middle America, The President does not lie, so!!


  44. A5, the word to bar is – Trump?

    Liked by 1 person

  45. Good evening Henry,

    As you say and we agree what the pollsters tell us has to be treated with extreme skepticism if not outright rejection on principle. In spite of the vast sums invested in the market research, quantitative and allegedly qualitative, the results are consistently wrong.

    What provoked my comment this morning was these same shysters appear to have NO SHAME.

    Within a few hours of calling the US election wrong they are pumping out new nonsense that apparently shows significant differences in voting patterns on grounds of gender, race, age etc. Claims are made in mainstream and social media on the basis of these alleged results. Men v women, old v young, white v black all neatly packaged in bar charts (again)

    Sorry – as I said – these pollsters have no credibility. It is there fucking bare faced cheek that aggravates me – not the subject itself.

    Liked by 1 person

  46. Having said that we have executive government forever seeking to ride the popular wave – to discover what the public “want” through focus groups, research and surveys.

    No reason any of that should work any better than predicting voting behaviour.


  47. anicoll,

    Not sure from your reply whether or not I have offended you. Certainly unintentional, if so, but I will take my leave just in case as I do not wish to fall out.


  48. Not at all Henry – I was trying to add a little more flesh to my comments on the negative effects of polling and pollsters – shameless bunch that they are.


  49. New post up


Comments are closed.

%d bloggers like this: