Statistical Comparison Of Top Teams In England

The international break provides a good opportunity for statistical analysis of various teams and players. To begin with, let’s look at macro level stats for the big six. I will also try to cover some player comparisons before Premiership resumes on the 19th.

The following table lists passing, dueling, and other figures for Arsenal, City, United, Chelsea, Liverpool, and Tottenham. These are Opta stats taken from Epl Index (@eplindex), an excellent resource for those who like number crunching.

Click on the image to view a larger version

 

I wanted to look at Arsenal’s performance as compared to other teams and with the Gunners’ efforts from the last campaign. And since City have shown a remarkable improvement over their Champions League spot clinching run from last season, it was interesting to look at their numbers over the two periods. For the other teams only this year’s figures are included.

In the table, Passing Accuracy includes goal-kicks, throws, etc. OPP Accuracy is the passing success from open play passes. Other rows are fairly self-explanatory.

Arsenal Year-on-Year

It is very interesting to note that the Gunners have uncannily similar numbers over the two seasons despite wholesale changes in personnel. Passes per min are very similar. Duelling success rates are virtually identical as are interception figures.

One significant difference is in the Minutes per Aerial Duel or 50-50. In 2011-12 Arsenal are getting into an aerial battle once every 4.78 minutes. This is almost a 40 percent increase over the previous year’s 3.43 min figure.

Part of this could be down to opponents launching fewer long balls into the Arsenal half. Since this was perceived as one of the Arsenal’s main weaknesses, it would be surprising if opponents suddenly stopped using this tactic. Some might think that Arsenal’s defenders are doing better against the long balls, and that might be the case to an extent, but overall the Aerial success figure of 50 percent is not very different from previous season’s 49. These numbers provide an opportunity to pause and reflect.

The bigger difference, in my opinion, has come from a change in Arsenal’s tactical approach and the shape of the team. In many of my match reports I have noted that the midfield is playing a lot deeper – a point also mentioned by Arshavin in his recent interview (translation needed) – and this has led to better support for the back four. Opponents no longer find it easy to punt the ball into the Arsenal half and push numbers forward as Arsenal are doing much better on the second ball.

Gael Clichy was one of the most troubled left backs in the league and was often a target of such long balls last season. But after his move to City, their left side hasn’t suddenly become vulnerable. If Clichy was as bad as some Arsenal fans proclaimed, why is Mancini’s defence so solid? It again reiterates the point that it isn’t so much about individuals being good or poor but more about the system exposing certain players or zones on the pitch. Since City are well organized defensively, teams have not been able to target Clichy.

Another noticeable difference is in the minutes per loss of possession figure. With two direct attacking players on the wings, Arsenal are losing possession more frequently. Ramsey too has contributed to this decline which results from a bad touch or a player over-running the ball in an attempted dribble.

Nevertheless, Arsenal have improved the Passes/Min value – which means they are making more passes per game – despite losing a tika-taka player on the wing for a more error-prone but higher impact player. I had looked at these stats in the context of the Sunderland game and Arteta’s performance stood out. Β It seems Wenger has given the Spaniard a tactical role of linking play all over the pitch. Subsequently, this seemed to be corroborated by @whoscored,

Mikel Arteta: The only PL player in the top 10 for Average Passes per game from the top 5 leagues, with 76.7 passes per game

I will try to look at further details when I get to player analysis and I will return to the Arsenal numbers again later in the post when comparing all teams. Next, let’s look at City’s effort over the two seasons as there are interesting insights to be gained.

Man City Year-on-Year

Most Gooners and City fans will agree that Arsenal were by far the better side when the teams met last season. However, Mancini has taken his team up by a couple of levels this time around and the results are not down to luck.

The Blues from Manchester have improved on almost all parameters. They are passing more and with higher accuracy, their crossing is better, their tackling and duelling success rates have gone up, chance conversion is superior, and so on.

Last year Mancini created a strong defence but his side didn’t move the ball around as effectively as the Gunners. Their jump from 5.49 passes per min to 6.11 has brought them almost on par with Wenger’s side. More importantly, they have overtaken the Gunners in the three Min/duel ratios (ground 50-50, aerial 50-50, and tackles). That means they are engaged in individual tussles less frequently.

In some games we hear commentators describe the play with words like, “They’re chasing shadows here” or “They just can’t get near the ball”. When taken alongside a high passing rate, the decrease in duel frequencies seems to imply that opponents are finding it hard to get near City (and therefore there are fewer individual battles) and provide a more tangible expression to what may sound like platitudes from the commentators. It then forces opponents out of shape and leads to more chances for the team on the ball which correlates well with City’s extraordinary chance creation and goal tally thus far.

It remains to be seen when Mancini can keep his team performing at this level and sustain their defensive solidity but thus far they have performed at a level even Wenger and Ferguson would envy. Of course, there is also an argument related to the financial approach of the clubs, but in a purely football related discussion that should be left out.

Comparisons across clubs

Arsenal pass the ball more frequently than any other club – even with what some fans call the worst team ever – but Chelsea, with a very positive approach under AVB, and City, with significant improvements over last season aided by major acquisitions, are very close to the Gunners.

There isn’t a big gap between the sides when it comes to success ratios in various duels. City have a fair lead in tackle success but others are all thereabouts. Ground duel ratios are eerily similar with a slight edge for Liverpool and Tottenham. Aerially, Arsenal and United (!) have some catching up to do.

Liverpool and Tottenham are engaged in Aerial duels most frequently which could be a result of additional individual tussles in the opposition half from long balls played towards their strikers.

Surprisingly, United are lagging in the minutes per chance created metric where City are in a league of their own, as discussed earlier. However, both Manchester sides have a fairly high conversion percentage which has put them on top of the scoring charts. Liverpool are really struggling on that front. Chelsea, Arsenal, and Tottenham will have to do better.

A key figure for Arsenal is ‘Dribbled Past per Game’. The Gunners are beaten more often than other teams with only Tottenham in the same ball park. For a team that wants to play a high line regularly, this is a critical number. If you think back to Koscielny’s own goal against Blackburn, Song was beaten rather easily by Olsson. There are times when one gets the feeling that the Gunners are not watching the ball closely but are beaten by the player’s fakes/movement. An improvement here will make Arsenal substantially stronger defensively.

Crossing is another area where Wenger needs to get more from his players. While it is inherently a low percentage tactic, given that the teams are making similar number of crosses per game, poor crossing accuracy – probably a result of insufficient tactical thought and practice – from the Gunners makes it easier to defend against them and negates the impact of their ball movement to an extent.

Final Thoughts

As I have said in the past, it’s hard to form meaningful decisive opinions from limited statistics. But there are definitely enough pointers in the numbers above. Liverpool just have to improve their conversion rates and whichever side from Chelsea, Arsenal, and Spurs betters their own number will likely challenge the top two.

City’s results are not a fluke and they will run away with the title if this level of quality can be sustained. It’s a big if though, so others have to keep plugging away.

AVB deserves credit for his bold, attacking approach. Chelsea are not too far from the top two and can do much better. As Wenger says, it’s in the details. They need to cut out some of the errors at the back and improve in the final third.

Arsenal are moving the ball a lot but they need to be more effective. The duelling frequency should drop and the opponents will have to be stretched more. It’s a tactical challenge for Arsene with Gervinho and Walcott on the flanks. More on this later. Le Boss also has to get more value from the crosses and improve the dribbled past figure.

All in all, City are the team to catch but the others aren’t too far apart. There is certainly food for thought for Arsenal fans who don’t support the manager and some players.

With so many interesting numbers I am sure to have missed out on something, and many points are debatable, so am looking forward to your thoughts.

P.S.

No disrespect to Newcastle who have exceeded expectations thus far, but I will be surprised if they are in the top 6 at the end of the season. However, their numbers are not bad at all and consistency, if sustained, can surely see them running the top teams close. Optimistic fans cannot be faulted for believing they can better one or more of the top clubs.

 

Compiled by DesiGunner using Opta stats from EPL Index


26 Responses to Statistical Comparison Of Top Teams In England

  1. HexyDre says:

    Excellent and delightful read as usual, I commend you for the time taken to put this together. Always a pleasure to read your analysis apart from pre- and post-match reports. Hope to see more of this during the dreary non-football days ahead.
    Cheers

    PS: Sorry I have no insightful observation to add to this piece, just appreciating πŸ™‚

  2. SongBillong says:

    I agree with HexyDre – great work.

    Not sure about Newcastle. I assumed Pardew would be sacked by Christmas but it’s looking good, as you said. I reckon 7th place will be where they finish – still a very good season for them.

    • desigunner says:

      To be frank, I didn’t rate Newcastle’s chances either but Pardew has done really well. He had some good games with West Ham as well in the past but hasn’t shown enough consistency. Let’s see how this season goes for him. Certainly seems better than the likes of Phil Brown and others of that ilk.

  3. drvics says:

    Hi Desi,

    Great work again. I guess they are right when they say ‘Numbers do not lie’. City’s improvement throughout the board is evidence enough for their tag as ‘Title favorites’. I am very impressed especially with the Min/chance created numbers for them. If they keep that up and with the finishing quality that they have, they will definitely blow their 2010-11 season goal tally (60 goals) out of the water.

    For Arsenal a couple of numbers stand out. The number of interceptions/game is obviously higher to other teams as we play a higher line. But if your logic of our opponents opting for fewer high balls is correct, then that would mean that our opponents now have to look to play ‘through balls’ and I would think that is more easier to defend and it would also leave our side less vulnerable to counter attacks.
    The other interesting stat is the crosses one. Though we are crossing less frequently and the accuracy is a bit lower than last season, I cant help but think that our crosses are a lot more dangerous this season. This is primarily been due to the team trying to go for ‘low’ crosses rather than aerial ones, which we almost always lost out on. Low crosses increase the chances of own goals and rebounds and with Walcott and especially Gervinho getting closer to the goal line and ‘cutting back’ most crosses, In last few, crosses were more in hope than anything else and with only a single striker fighting for the cross with 2 or 3 defenders, we invariably lost out. I think we have improved a great deal on that front this season.

    • desigunner says:

      I think Arsenal have also adjusted the high line to an extent as the players have dropped back in some games which were tight. For instance, against Chelsea, after the first five odd minutes the players did drop back. While much was made of the high lines of the two sides, the first six goals in that game came with the respective defence just outside or actually inside their penalty box.

      I feel Arsenal have improved in their decision making in terms of high line vs dropping deep. But yes, there should be fewer counter attacks if Arsenal can cut the through balls more effectively.

      I also agree that Arsenal are crossing into more dangerous areas but as with most passes, it’s not just about the delivery but also about having a player, or players, in position to get on the end of the ball. That’s where I feel the tactical approach of the team needs tweaking. Walcott and Gervinho have the pace to attack the back post. One of the midfielders can arrive on the edge of the box, and so on. It needs a little more functional and predetermined approach than what Arsenal currently have.

  4. Busayor says:

    Like dis!ur analyses are spot on

  5. CB says:

    Great stuff, much appreciated.

    Not a criticism or request for more work, but would be good to see the change in Arsenal’s figures over time ie the bad start to the season and the recent pick-up. If we are matching sides on the average of these it implies the pick-up period has been very good. For example our goal difference has gone from terrible to positive, so the last few games must be very positive with hopefully room for improvement still as the team gels.

    • desigunner says:

      I will try to look at the numbers but I think the United game skewed matters greatly. Against Liverpool the team wasn’t bad, considering some key players were missing. Against Blackburn it was more about annoying mistakes and overall quality was not bad. Against Spurs too the football quality was quite decent but the second goal was a freakish one. So I don’t think the stats will differ greatly. At least I don’t know which stats will vary or cover the subtle changes in team shape, tactics. But I will try.

      Overall my feeling is that Arsene likes a particular style and sticks to it. Obviously passing the ball, movement, and pushing forward is important. After that he probably leaves it to the players to find the answers on the pitch based on extensive training of playing in tight areas. But some errors can lead to extraordinarily poor results with such an approach, even if the players get 80-90 % of things right.

  6. CB says:

    drvics – point well made, though is a low cross/cut-back counted as a pass in the stats?

    • drvics says:

      @CB: If the cut back are considered passes, it could partially explain the lower crosses/match stat. But even if it isnt, I still believe we have been more productive with out wide play this season. Also remember that Gerv doesnt mind cutting inside from the left.

      • desigunner says:

        I believe cut backs are considered as passes. There seems to be a cut off of some sort. It must be there in @orbinho’s timeline but I don’t have the time to check it.

        Arsenal have been a lot more productive from wide areas no doubt. But is that enough? Considering the attacks through the centre have reduced due to different responsibilities on the midfielders and a relative lack of understanding thus far, wasn’t a lot more expected from wide areas?

        In other words, Gervinho might get 10 more assists than Nasri but if it comes at the cost of 12-15 assists down the middle, would it be considered an improvement? Just random numbers to highlight a thought.

      • gunner_expat says:

        That’s an interesting point regarding the nasri/gervinho assits debate.
        the way i see it,
        1. there’s absolutely no harm in adding a genuine weapon on the wings, it can only free up space in the center
        2. and since we no longer have the master assist maker in the center we need to change the style in any case.

        I think with ramsey growing more confident and once jack is back, we’ll be able to be a bit more unpredictable and dynamic, and in the the long run, gervinho might even become a more prolific goal scorer.

  7. critic says:

    nice write up.

  8. Good read, I agree it’s difficult to make decisive statements at this stage from the numbers but it’s definitely worth a comparison.

    I would say there are a couple of key statistics not listed/talked about above worth looking at though. The first is shooting accuracy/shots on target, with SOT strongly correlated with success. Liverpool, for example, aren’t only failing to take chances, they’re not even hitting the target often enough, which is perhaps more worrying.

    I’d also look at passing accuracy in the final third; that is, making passes where it really matters. Teams who can pass in dangerous areas are always in business.

    But other than that, pretty much agree with all your thoughts!

    • desigunner says:

      Do you mean a ratio of goals to shots on target? The above table has shots on target/total shots and goals/total shots but goals/shots on target might also be an interesting variable.

      It would probably be same as the ratio of the last two rows. United would have a good ratio 54/22 – roughly a goal every 2.5 shots on target.

      Liverpool certainly have a problem with their shooting accuracy which at 39% is by far the lowest. They also have a conversion issue with 9 of those 39 going in i.e. One out of 4.5 shots on target leading to a goal.

      Passing accuracy in the final third should be interesting. Will add it next time. Number of final third entries could also add some value.

      Thanks for the 5 added minutes over here πŸ™‚

    • cupsui says:

      agree good point, also i think some assessment of the goal keepers would be handy. Saves per goal, mins per punch, successful clearance, accuracy of distribution,

      I haven’t been overly impressed by any of the big clubs goalies this season except perhaps Hart and Woj although i haven’t seen enough of city to comment too much. De Gea looks very nervy, which is understandable and indecisive, i thought czech was particularly bad against arsenal and coulda saved at least one of our goals, he doesn’t have the same presence pre head break. The big yank for spurs has been good too i must say it would be interesting to see his stats. and reina is good but from what i have seen a bit prone to his usual gaffs

      great article desi, this blog is streaking ahead in the gooner realm as it is one of few that gives football related commentary instead of off the field, tabloid spun, garbage that every other blog focusses on!
      Thanks Des

      • cupsui says:

        that was i agree with the passing accurarcy in the final third…could a good stat to see

      • desigunner says:

        Need to find the time for comparing the player stats based on positions. My first interest is in the wide players and midfield as this is related to the tactical changes that I feel teams have made, especially Arsenal. Will also try to squeeze in goalkeepers but I am not convinced saves, punches, etc are particularly insightful as it doesn’t cover quality of chances.

  9. TheTactician says:

    Hi Desi,

    Regular reader, but first comment here.

    I like the number crunching and you have made excellent sense out of all the stats available.

    2 points I felt could give a better insight:

    1) Duels / tackles won/lost figures – can they be looked at from which area of the ground? Like opponent’s half, final third etc. Dont know if this information is available.

    2) Cross-field balls / switching flanks – numbers and success percentage. I have seen City, United, Barca etc doing it very successfully, many a times catching the opposition defense on the wrong foot, mostly causing them to make mistakes. This is one area I feel we could improve with the skillful pacy wingers we have.

    Once again, great piece of work…

    Cheers πŸ™‚

    • desigunner says:

      Thanks mate.

      I too have been looking for stats on duels in various parts of the pitch. On a per game basis these are available in the Guardian Chalkboards but I haven’t found any that provide across the whole season or a few games.

      Cross-field balls are also interesting and it’s an issue that I have discussed in the past. But again there aren’t any specific stats on those to the best of my knowledge. Will keep looking.

      Cheers.

  10. Kushagra India says:

    Desi very nice read, waiting for ur thoughts on the England vs Spain game Parker was immense…

  11. […] made this comparison for two reasons. Firstly, I have observed, and noted in previous articles (here and here), that Arteta is compensating for the lack of technical contribution from the attacking […]

  12. […] so am putting the links here. Those interested in stats would find something interesting in this comparison of the top teams in England. And this is the updated article covering Arteta, Wilshere, and others. It now has numbers for […]

  13. […] won’t be easy though. As was discussed in this article, City have improved considerably over last season and have moved the ball as well as Arsenal do. […]

  14. […] the midfield in general (I’d touched upon this earlier in the season in different ways here, here, and […]

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s