Arsenal Should Win The Title Says The Data


How often do we hear the old cliché “football is all about emotion”? While this may be used to explain the wild swings in opinions and attitudes of pre and even pubescent supporters of Arsenal Football Club, it can in no way excuse such conduct by the mainstream media as well as other opinion-leaders (bloggers, podcasters, etc) who pound the Arsenal beat in both print and electronic media. It took me a while in adult life to understand that exploiting emotions has been a longstanding practice of politicians and their stenographers in the media, who see a gullible, ignorant public as vital in maintaining their power and dominion over us. How else could they implement and sustain policies that are alien to the interests of the majority.

I apply the same paradigm to Arsenal-football as I do to current affairs. Four months ago, only four games into the season, after eking out a 0:1 victory over Newcastle, the mainstream-media went into hyperdrive with doom and gloom.

Alan Shearer, now a Match-of-the-Day supremo, opined:
“Not hard to judge Arsenal. They need a top class center forward.”

Danny Mills, Shearer’s frequent sideman on the Beeb, was equally prophetic:
“Arsenal are going to struggle this season.”

Arash Heckmat, who covered the game for the Mirror, had a gloomy narrative:
“There are still big question marks over Arsenal….”

Admittedly the Gunners did not have the best of start to the season. After 4 games they had scored three (3) and conceded three (3). Arsene admitted before the Newcastle game that performances had been “very average.”

In comparison Manchester City was on fire. Four games, four victories, scoring 10 goals (2.5 goals per game) and zero (0) goals conceded. Manchester City’s official twitter account bragged:
“City break club record with tenth consecutive league victory!”

Similarly, the media was blowing City’s trumpet. With the exception of DeBruyne, they had by that time done everything Wenger was later made infamous for apparently neglecting, i.e. making a massive spend on outfield players:
Sterling – £49M
Mangala – £32M
Delph – £8M
Otamendi – £33M
DeBruyne – £54.5M

That, girls and boys, adds up to a cool £176.5 million to supplement an already expensive squad containing the likes of Aguero, Silva, Toure and Kompany to name a few. Given such massive spending by our main title rival, on what is generally accepted are quality players, how can the media now say with any credibility that the quality of the Premier League is less than normal? Is it any coincidence that such a thesis gets traction because Arsenal are now top of the table due to the manager standing by the quality of his squad and refusing to spend wildly? Similarly, we currently have media whore and club parasite Piers Morgan flogging the meme that the FA Cup is no longer a big trophy conveniently after Arsenal has won it twice in succession and showing a strong commitment to winning it three in a row.

But I digress. Going back to late August-early September the MSM had established a narrative that big spending City was well on the way to smashing all before them and Arsenal was going to struggle. Some of us amateur bloggers, including yours truly, refused to drink the kool-ade. I said in my September blog:
“As with City and goal scoring, it is questionable whether this level can be sustained over the season given it is relatively the same midfield over the past 2-3 years.”

Since then they have gone from 2.5 to 1.95 Goals  per game compared to a 3-year average of 2.2. Most importantly they are leaking goals, from a statistical impossible 0 GA per game to 1.05 which is above their 3-yr average of 0.96. Talk about being brought back to earth with a thud.

As for Arsenal, after reviewing the statistics, I tried to convince my readers that far from dying, Arsenal was very much alive and based on immediate past history would soon get the results that would put the club in title-winning contention. The proof is clearly in the pudding; four months later we are now top of the table.

After 20 games, relying solely on descriptive statistics, I have come to the conclusion that the club is doing well but is still under-performing in some key areas. The figures suggest it is reasonable to expect improved performances in the second half of the season but not as dramatic as between games 4 and 20. I will use the following series of graphs to make my point. Note that the information used is publicly available at Squawka.com and I have compiled the data and color coded the information as follows:
Red: 3-yr Average
Amber: 4-week Average (1st 4-weeks of 2015-16 EPL season)
Green : 20-week Average (1st 20-weeks of 2015-16 EPL season)

Graphic #1: Goals For Per Game

Goals For

It was not long ago, for weeks after Newcastle, there was virtual panic inside and outside the Arsenalsphere because the club had stumbled out of the box with a mere 0.75 Goals Scored per game. Two competing narratives were “Giroud could never lead Arsenal to a title” and “Walcott could never be a central striker.” Even though the club had a three-year average of 1.85 goals per game with the mostly same combo,  instead of concluding that the scoring rate  was an extreme aberration, MSM-bloggers-podcasters in general instead chose to incite fans into demanding a new world-class striker. Four months later, with Giroud (10-goals) leading the way, AFC is scoring at an average of 1.7 goals per game and seemingly destined to meet our usual standard.

Graphic #2: Chances per Game

20 Chances per game

This is one metric that was simply off the charts at the end of August; AFC was creating an average of 16.25 chances per game compared to the prior three-year average of 11.71. The team has now regressed closer to the mean but clearly at a higher rate than previous years. Seems it is the Özil effect.

Graphic #3: Conversion Rate

20 Conversion rate

This picture speaks more than a thousand words. Historically AFC converts chances at a near 16% clip and after Newcastle the MSM and its echo-chamber went into emotional overdrive as the team stumbled at 5%. It begs the question how could the media and their lemmings project that a top club like Arsenal would sustain such a low rate. Sixteen games later they are converting at 14%, just about 2% off the 3-year standard. As for that lamp-post Giroud, he is knocking them in at 20.9%. In comparison, Aguero, almost everybody’s PL benchmark, when fit manages 24.1% and Lewandowski, arguably Europe’s best striker, is at 20.3%.

Graphic #4: Shot Accuracy

20 Shot Accuracy

The shot accuracy story is similar to conversion rate already discussed. Interestingly there is a perfect equivalence with Giroud’s shot accuracy at 50%, as measured by Squawka, compared to the club’s. Is there a statistician willing to do a correlation coefficient (“r”)? That is beyond the scope of this blog but it should make an interesting line of inquiry for those who need confirmation of Giroud’s current importance to the club.

Graphic #5: Goals Against Per Game

20 Games AgainstFinally, graphic #5 demonstrates that while AFC was unable to sustain a miserly 0.75 Goals Against after four games, the club is defending even better than it did in prior years, only conceding  at 0.9 goals per game compared to 1 gpg. In a game of small margins, if that rate is sustained until May, the club would concede only 34 goals over the season. That would best any of the recent season’s efforts, practically Chelsea-esque who in the past three seasons conceded an average of 33.

From the data above, with no further injuries and some key players returning from the sick-bay, it is clear that Arsenal has the capacity to improve in all the key metrics that indicate success. The only exception I would make are those crazy Chance Creation numbers. Goal-scoring, in particular, I am convinced that with Alexis(soon returning to fitness), Giroud, Theo and Ramsey meeting their average numbers plus an improving Campbell, I can easily see us getting back to that 1.85 average. That means 36 more goals in 18 matches or 2 goals per match. Given our ability to hold teams to an average of one goal per game, such goal-scoring and defending is title winning form.

In the same breadth it must be recognized that that City has the capacity to sustain a high performance level and win the title. Those attacking players they signed are not chopped liver. It is well chronicled that their defending is the problem. Mangala and Otamendi must be an example of the biggest waste of £65 million by a football club, and an indictment of the mantra that a club must pay over the odds to improve their league position. Setting aside their profligacy, the data below clearly shows if MCFC can improve their goal scoring and defending even marginally they could be a real force.

MCFC 3-Yr Avg 4 Games 20 Games
GF per Game 2.20 2.50 1.95
Chances Per Game 12.40 11.75 13
Conversion Rate 16% 17% 15%
Shot Accuracy 48% 60% 52%
GA per Game 0.96 0 1.05

But my focus is Arsenal, not City. I am concerned that the media narrative is settling for the sentiment that Arsenal will win the title. I must emphasize that in the prediction business, media sentiment is a contrary-indicator. Their standard procedure is to assume that the most recent results are a predictor of future performance. To the contrary, any serious student of the social sciences  would have learnt that historical consistency is the only reliable indicator of future success. No wonder Arsene Wenger emphasizes consistency for himself and his teams. With clearly inferior squads during the barren years his teams were among the top 3 or 4, never less. In the past three years he has been able to acquire better players, and as evident by 2015 statistics (AFC being the team that earned most points in the calendar year), we can predict with a reasonable degree of certainty that with better quality players, Arsene’s team will either meet or exceed their 3-year average.

As I concluded four months ago: “Unlike the media, which thrives on emotion, in the silent statistical world, there are no headlines. There are no narratives. No excuses. No hope and no despair. Just data.” And that data suggests Arsenal has done well and will do better.

May cannot come too soon for me to see how right or wrong I am.

Comment navigation

Newer Comments →

66 comments on “Arsenal Should Win The Title Says The Data

  1. Trying to imagine a situation where Dean would give us that penalty. Failing so far.

    Liked by 2 people

  2. Jon Torol scores for Birmingham COYG


  3. AFCB lead 1-0 at half-time – Benik playing OK. Had a header over the bar when under pressure and trod on the ball when trying to get round the keeper one-on-one.
    Boruc has made two really good saves.


  4. An exceptionally sub standard football match at SJP – truly 4/10 from both sides

    Pleasure to watch really

    Liked by 1 person

  5. Corp Jenks on at Bmth.


  6. Payet scores from free-kick – bloody obvious. Keeper far too far left.


  7. Afobe subbed.


  8. 1-2 Hammers – eduardo would be horrified at Bmth’s defending.
    But WHU have been on top second half – deserved lead.


  9. Alexandre Song Billong, wearing goggles, on for Payet.


  10. Atkinson denies Bmth a pen for obvious handball. The man is consistently awful.

    Liked by 1 person

  11. WHU now 3-1 up – I think Boruc must have money on a Hammers win. Two free kicks and he was awful on both.


  12. 3-3 at SJP
    The quality of both sides football was decent Championship quality – I can’t see either being relegated but the Mancs will be very very lucky to qualify for CL playing this kick and rush stuff.


  13. Just listening to a lad from Sheffield on the radio, just returned from the Wednesday v Bolton game. He said Bolton had this fantastic lad on the wing who was absolutely different class called Wellington Silva. Well another one for next year then.

    Liked by 1 person

  14. I suppose Wenger will be blamed for overlooking Silva’s obvious potential if he comes good.


  15. PGMOL rules must mean that penalties are handed out only when the narrative of a game requires one. And foolish me used to believe the ordinary laws of the game would be applied.


  16. A few years ago I remember Kos hauling down a City player from a corner: the penalty was given (did Szczesny save it?) but Kos was sent off. Was Dean the ref, and if so why no red card for Smalling last night? And good news to hear about Wellington Silva: it would be fun if he did a Campbell.


Comment navigation

Newer Comments →

Comments are closed.

%d bloggers like this: