# What’s Ailing Arsenal and Can They Finish 4th? – A Statistical Analysis

If you ask, I say maybe there is a little bit chance we don’t make top 4 this season.

One of the things that statistics are best used for in sports is cutting through the crap the press spews at fans every day. The drive for fresh stories and angles produces an awful lot of noise, much of which doesn’t actually make any sense when you look at the facts behind the matter instead of the opinions.

Today I want to take a look at Arsenal across a number of metrics used to evaluate how teams have performed and where they are likely to finish at the end of the season.  In the real world, Arsenal currently sit 6th in the league table, seven points behind Spurs with a game in hand, and three points behind Everton. This type of lead isn’t insurmountable at this point in the season, but there are strong indications that this Arsenal team just isn’t the same as the previous models.

Pythagorean

 Team P GF GA GD ActualPts Pythag Exp Manchester City 23 45 19 26 51 47.50 Chelsea 23 47 22 25 45 45.77 Manchester United 23 57 30 27 56 43.90 Arsenal 22 41 26 15 34 38.25 Liverpool 23 40 28 12 34 38.06 Tottenham 23 40 28 12 41 38.06 Swansea 23 34 27 7 33 35.47 Everton 22 35 26 9 37 35.19 West Bromwich Albion 23 33 32 1 34 31.60

According to Martin Eastwood’s adapted Pythagorean Theorem, the top half of the table should look like this, with Arsenal, Liverpool, and Tottenham essentially in a three-way tie for the fourth and final Champions League spot. Arsenal’s Pythagorean expectation is buoyed by a number of blow-out wins coupled with close losses, explaining why this method of analysis probably makes them look fractionally better than they have actually performed thus far. (The inverse is true of Aston Villa, who have suffered a few enormous defeats, thus seeing their expectation rooted to the bottom of the league table for this methodology.)

Total Shots Ratio (TSR)
This methodology is widely considered the most effective way of evaluating how teams are performing right now and where they are likely to end up at the end of the season. It basically looks at the combination of offensive and defensive outputs for each team to create a new metric that examines how dominant that team is. The one weakness it has, is that this version doesn’t really account for teams that are creating more effective chances and shots on goal than normal.

 TSR_Rank Team Shots pg Shots OT pg SoTPar Delta Shots conceded pg 1 Manchester City 18.2 6.4 5.915 0.485 9.3 2 Tottenham 18.4 6.4 5.98 0.42 9.7 3 Liverpool 19.3 5.6 6.2725 -0.6725 11.5 4 Everton 17.5 5.9 5.6875 0.2125 11.7 5 Chelsea 15.8 5.3 5.135 0.165 11.5 6 Arsenal 14.7 5.1 4.7775 0.3225 10.8 7 Manchester United 15.3 5.8 4.9725 0.8275 13.2 8 Newcastle United 14.4 5 4.68 0.32 14 9 Swansea 13.6 4.7 4.42 0.28 14.8
[Apologies that the tables are ugly – I am just barely competent at posting in WordPress.]

As you can see, Arsenal rank sixth in this methodology, narrowly behind Chelsea, but a good bit ahead of Manchester… United? League leaders Manchester United?!? Both the Pythagorean and TSR methods suggest that Manchester United have been unbelievably lucky so far this season compared to what the statistical models would suggest. Part of this is on the back of a very talented strike force that is creating .8 Shots on Target extra per game, part of it is on the back of some incredible conversion rates from corners, and part of it is due to the amazing talents of Sir Alex Ferguson.

Anyway, with regard to Arsenal, they are ranked sixth in TSR largely on the back of solid defensive numbers.  If you look at both ShotsPG and Shots on Target numbers, they would actually sit one spot worse in 7th position, generating 4.5 Shots Per Game less than Liverpool.  This is highly unusual for an Arsene Wenger team, whose squads haven’t finished lower than second in TSR in any period I have data for (which goes back to 2010 at the moment).  What’s interesting is that the Shots Conceded numbers are almost constant for Arsenal (2010: 10.2, 2011: 10.6, 2012: 10.5, 2013: 10.8). It’s the offensive numbers that show a marked declining trend (2010: 17.4, 2011: 17.2, 2012: 16.8, 2013: 14.7).

[Note: The red highlight on Liverpool identifies the source of one of their problems – they just haven’t been very good at shooting the ball at the goal, especially compared to the rest of the teams listed there.]

ISG Coefficient
A team’s ISG coefficient is a way of analysing how they performed in the same fixtures last year to get an idea of how easy or hard the schedule has been thus far in the season. Obviously if you have already faced both Manchester teams, Chelsea, Spurs, and Liverpool at home in the first half of the year, you would have a markedly lower probability of winning those same fixtures away in the second half.

Anyway, after their loss to Chelsea at the Bridge (where they won 5-3 last season), Arsenal’s ISG Coefficient is now -15! This means that compared to the same fixtures as last season, Arsenal are 15 points worse off. This is the second worst number in the league, just edging out relegation candidates Newcastle in terms of poor performances in the same fixtures as last season.

Another reason to believe talent levels are on the wane is Arsenal’s performance splits against the Top5 and Bottom5 teams. As my colleague Ben Pugsley has pointed out, Arsenal have taken only 6/24 against the top tier, while taking 17/24 against the bottom clubs.

The easy conclusion from looking at all of these metrics is to say Arsenal are considerably worse than they were last season. As for that whole finishing 4th thing, someone should be offering you really exceptional odds if you were looking to bet on them actually making the Champions League next season. Between their terrible ISG coefficient and the reduced offensive production, I find it hard to believe that even Arsene Wenger can pull this one off, especially if they don’t make multiple major signings before January is out.

Individual Errors
A couple of weeks ago on Twitter, ESPN commentator Janusz Michallik said that out of all of Arsenal’s centre backs, they still don’t have a really good one in the bunch. This was about the time that Per Mertesacker was being bossed by Michu in the FA Cup tie at Swansea, and before Laurent Koscielny rugby tackled Edin Dzeko in minute 9 against Manchester City. The fact of the matter is, regardless of their individual strengths, Arsenal’s defense has been incredibly prone to individual errors that cost the team goals, and they have been this way for years.  This is something that needs to be fixed if they ever want to contend for a title again.

The Loss of RVP

24M£? You guys totally miss me, don’t you?

This past summer Arsenal sold the Premier League’s best striker to Manchester United. They also sold midfield cog Alex Song to Barcelona (and regardless of what you think of Song, both of those players are on track to win the league with their new teams). Obviously these sales were a big deal, but Arsenal have been a selling club for years. Arguably, the loss of Cesc Fabregas and Samir Nasri in the same season should have had at least as big an impact, right? Well, yes and no.

The fact of the matter was, RVP had been in England for a long time when Cesc and Nasri were sold, and was ready to make immediate contributions to the team, provided he could stay healthy.  Everyone knew RVP was an enormous talent that just needed his health to cooperate in order to become an outstanding player. Meanwhile, Arsenal had reasonable replacements waiting in midfield and on the wing for the roles that Cesc and Nasri played. They also bought a replacement in midfield in Arteta, who probably wasn’t as good as Cesc, but was still an excellent talent that was already adapted to the league. [This isn’t meant to disparage foreign signings. Far from it, since they are typically where you find the most value. It’s just meant to highlight the fact most players take time to adapt to new leagues.]

The same can’t be said of Arsenal’s squad when they sold Robin or their recent reinforcements. I like Giroud and Podolski, but players (especially from Ligue 1) typically take time to settle in England. More importantly, whatever their skills, neither of those guys is remotely near van Persie’s level of talent. I guess what I am trying to say is, there is a significant difference in replacing a player’s position and replacing their production. Arsenal did the former with RVP but failed dramatically to do the latter.

The REAL Issue

Graphic from SportingIntelligence.com – Click to Enlarge

The real issue with the club is something that has been going on since they moved from Highbury to The Emirates. Arsenal have not only failed to keep up with other teams contending for the Top 4 in transfer spending, they have failed to keep up with literally everybody.

Net.

Positive.

Transfer Spending.

While Manchester City and Chelsea were spending over 500M£, and Spurs, Liverpool, and Manchester United were spending around 200M\$, Arsenal made a profit on transfers for a decade.

Yes, they spent quite a bit on wages, but unless you are attached to La Masia, it is impossible to have this pattern of expenditures and expect to compete at a high level in perpetuity. It is also impossible for your current players to look at this pattern and expect them to stick around if they want to win trophies.

It’s clear at this point that Arsenal’s management have been hoping to scrape by and continue qualifying for the Champions League each season until their Pre-Emirates commercial deals came up for renewal, at which point they would be free to dramatically increase spending on transfers and wages. Incredibly, they have succeeded at doing just that. Until this year, when it looks likely Arsenal will finally be on the outside looking in at Champions League football.

However, all is not doom and gloom for Arsenal or their fans. In addition to their 70M£ transfer budget, they have an additional 30M£/yr from the new sponsorship deal with Emirates Airlines plus the additional Premier League TV revenue coming in. In one more year, they will likely have another 25M£/yr windfall from a new kit deal, either with current sponsor Nike or new suitor Adidas. That sort of cash can afford multiples of just about any player in the world.

What Do They Need?
Looking at the data above, it’s fairly clear that Arsenal have reached the bottom of their talent well on the offensive side of the ball. Santi Cazorla is a massive addition (and an absolute bargain) and pairs wonderfully with Jack Wilshere in the center of midfield or as a type of Juan-Mata floater on the left wing. Olivier Giroud has shown promise this year, but lacks RVP’s movement and genius in helping to create offensive chances, and his polish when it comes to putting them away. Podolski lacks the pace of Arsenal’s traditional wingers, and though he is an excellent finisher and interacts well in the passing game, he doesn’t seem to be able to beat opponents off the dribble.

In short, one of Arsenal’s summer purchases is world class (Cazorla), while the others are adequate, but don’t match the talent level of the players sold in recent years.

What Arsenal really need is as follows:

All Arsenal fans wanted for Christmas?

1)  An elite striker with good movement, who also passes well enough to bring teammates more into the game. Edinson Cavani has been the hot, mostly PR-inspired rumor for the last week, but he is exactly the type of player that Arsenal need to replace RVP’s production. On the other end of the budget spectrum, Michu would also have been an outstanding choice. What ever happened to Arsenal’s amazing scouting department?

2)      A quick left wing that can beat people with pace, off the dribble, passes well, and converts chances. Gervinho actually can do two of these things, but is so bad at the important ones (shooting, passing), that he makes the team worse when he is in the lineup. David Villa was this type of player in his prime, and Santi Cazorla might even be this type of player now. If Cazorla becomes the primary option on the left wing, then Arsenal need to replace his passing and vision in the center of the pitch instead. Talented wingers are typically cheaper and easier to find than playmaking midfielders, but either will work.

3)      One or even two authoritative centre backs who read the game well and don’t make stupid defensive mistakes with alarming regularity. At least one of these guys also needs to be fast. Because of Arsenal’s system, they also need to pass the ball well. Montpellier’s Yanga-Mbiwa was rumoured to interest Arsenal earlier in the year and has looked good when I have seen him. Someone like Swansea’s Chico would also be an excellent choice. Neither of these guys were particularly expensive before the season started.

4)      A deep-lying defensive midfielder that breaks up attacks and then passes well enough to immediately restart the attack at the other end. Wenger has been using Mikel Arteta for this, and he’s been excellent, but he’s also older and needs protection. He also needs a backup that is healthier than Abou Diaby, which at this point is just about anyone. This summer Yann M’Vila was rumoured to be the player that fit this bill, and if you look at his stats, M’Vila was incredible last year. Seriously, the dude was not only a physical beast, but also produced some of the best midfield numbers I’ve seen. That said, he’s apparently a total head case who has had run-ins with the law, and Arsenal backed off because unlike AC Milan, they usually avoid signing this type of player. They need someone, though, because their midfield depth is desperate.

Can these needs be filled with a budget that looks like it will be between 100 and 130 million pounds this summer in combined wages and transfer spending?

Absolutely.

Is Arsene Wenger going to finally step up and spend the money necessary to return Arsenal to being perennial league contenders?

Ahhh, now that is the real question.

–Ted Knutson