Sorting out the BITs from Emery’s defense

45 comments

1. “If we are so defensive, why are we still leaking so many goals?” Makes me think about your suggestion earlier in the season talking about how possession is the best form of defense. Increased BIT actions mean we are chasing the ball. Seems like Emery moving us away from being possession controlling team is responsible for the increased BIT numbers. For me this can still improve us defensively provided we have the right players across out playing XI.

   1. This was my thought as well – if you are in possession of the ball you are called upon less for defensive actions. Klopp’s teams going back to Dortmund have stressed control of space vs. control of the ball and until this year when they’ve adopted a higher possession strategy they typically were in the 40 percentile range.

      I think our stats get skewed by the chaos we still see at times – for example it be neat if there were a metric to compare give-aways to defensive actions i.e. Xhaka or Mustafi inadvertently pass the ball to the wrong shirt and then Torreira has to rescue the movement, Leno can’t play it out from the back because there’s no options and has to boot it long and our forwards then have to harry the defenders on the other team.

      A true Barca/Cryuff/Bielsa high press that keeps the ball in the other team’s end of the pitch is still missing.

      On Ozil – I’ve come around to the conspiracy theory that he’s being intentionally dropped to save money, his appearances bonus gets forfeited. The fact that this club is pinching pennies now makes me wonder if it’s time to become a neutral. It’s not like Ozil’s numbers are unrecoverable – probably just a modicum of motivation would bring his defensive stats up.

      1. As I have said many times in the past, possession stats and defensive stats have a very weak correlation (almost none). Many teams with low possession numbers also have low defensive stats. Pep’s Barcelona had the best possession numbers of all time and also had very high defensive numbers. It’s more about style of play than possession.

2. Thanks for sharing as always Tim.

   The player most similar to Ozil in the league is: D.Silva?

   And the comparison between the two is helpful, especially from the 2014 WC!

   Emery praised Guardiola’s football.
   And Guardiola would have Ozil in his team if he had him in his squad. Ahead of D.Silva!

   Klopp likes his Goetze’s, Coutinho’s (more goals with him) and would surely want such a player in his current squad. Would he have started Lallana over Ozil? You all know the answer to that!

   Perhaps the worst year of injuries in Ozil’s career have been a significant factor in the coaches thinking.

   Or he’s less suitable for this job then the likes of Pellegrini – quite a record over the last twenty years in Europe. How do i feel about a coach who likes to sign players like Cazorla and Monreal? Young Pep will do well to match or better his record at Moneybags City this season (needs a CL SF for the latter and a title for the former)

3. So we’re being busy but not productive. Maybe Arsenal is my spirit club. Lol.

   Ozil’s numbers are low this season? Gee, I wonder if he’s motivated to give his best for the new coach. Ramsey was also excluded for large parts of the season and his numbers seem fine.

   I think Jack is right in that it’s mostly a club decision to try and force Ozil out. Although I think you’re wrong. All Ozil has to do is be patient enough to pull a Pogba on Mourinho. The sort of football we’re playing will only convince him he could do that.

   Where I think Jack is not right is that it’s to save on bonus money. I think Emery just doesn’t want to play with AMs. As far as I can figure he wants to attack from wide on both flanks, and use his CMs as screens to cover for FBs bombing forward. Without Bellerin we can’t do this. And with Ramsey and Ozil we have players who want to join in the attack as their primary function.

   At least, I THINK that’s what Emery wants to do. If it is, I hope he knows what he’s doing because we just lost a great HG talent for free, and are devaluing our star player’s value, all while playing some pretty terrible football.

   1. Do you honestly believe Arsenal will let Emery go after one season when he is on track to easily better Wenger’s points total and at least match his league position, just because the football isn’t exciting enough?

      And then hire who, Arteta?

      Not happening.

      Emery will get his contractual two years and Ozil will be sold in the Summer.
      And by sold I mean given away with Arsenal paying part of his salary.

      1. It doesn’t matter what I believe. It only matters what Ozil believes and wants. Arsenal cannot sell Ozil if he doesn’t want to be sold. And while it’s true that the club can make his life difficult, if Ozil has friends in the dressing room it could make the club’s life difficult as well.

         Play well without him and the players are happy, Ozil has less clout. I’m not sure the players will be happy about dropping Ozil and playing the way they’re playing.

         Part of the reason I think Emery will want a large scale retooling of the squad.

      2. Also Jack might be onto something there.
         I don’t know any details of Ozil’s contract but if there’s a appearance / minutes played clause in it , then how big a stretch would it be for a club that e- mails employees to cut back on toilet paper usage ,to keep his minutes down for the sake of money.

         Alexis Sanchez reportedly gets £75 k extra every time he sets foot on the pitch for United and if Ozil has the same” incentive” ,this might explain things a bit.

      3. I believe the key here is making the CL. If we do that the players are happy at progress, the fans are happier, and Emery, and it must be said Raul, stay in the club’s good graces.

         The board didn’t plan for Raul to take sole charge last year. If there’s a war between Emery (Raul’s man) and the dressing room, without visible signs of improvement, there might be a price to pay.

         Ozil doesn’t seem like much of a troublemaker, but there must be a point where either he, or the club, give in and decide to go along. I don’t know what that point is or what Ozil’s preference is. I just feel this whole thing reflects poorly on our club and on a coach who came in knowing what the deal with Ozil (and Ramsey) was. I doubt he said they don’t fit his style before.

         Actually come to think of it, this ‘economy’ phase didn’t exist before Raul took over. We were happy with Mesut and paying Aaron 200k. Maybe that was his way of winning Kroenke over. That he’ll reduce our spending and still get us CL. Monchi will be the final move in that regard.

   2. if arsenal are supporting emery on his ozil stance, ozil can’t do what pogba did for mourinho. the club won’t care as they’re on board.

      1. Yeah, but that’s assuming their stance is absolute and doesn’t have a breaking point. If Ozil wants to stay regardless, there is nothing the club can do to move him on.

         I have a theory that the lack of permanent signings without even an obligation to buy is because the board are assessing Raul’s pitch for the top job where he would reduce costs and get us CL. If that doesn’t look like it could happen, I’m not sure they’d be happy having wrecked the value of Ramsey and Ozil.

         Maybe Emery isn’t playing two of his captains under directives from the top, and in return he gets more time to play his own style of football and get some players loyal to him. Rumours linked us to Nzonzi, Banega, and now Rabiot, and of course Suarez.

         If we don’t get CL, I wouldn’t expect the club to spend too much in the window. Cross laughably said 45m, but Ornstein said up to 100m including wages? Doesn’t seem too far apart and is what I expect. It’s all about getting to the CL.

4. Grateful for the insights this provides.

   Would it be possible to normalize BIT by multiplying by a team’s nonpossession % so that defensive actions per 90 discounteted if the team had more opportunities to play defense? Sorry that I’m not more of a statistician / maths / Opta guy. Love the blog, but have trouble keeping up with it!

   1. This is something Statsbomb did and I wrote a pretty comprehensive debunking of. They persist with it, even though it’s still completely without merit.

5. Our overall stats compared to the other top 6 teams are quite unexpected considering our reputation for poor defending. Less possession which means more defensive phases of play partly explains it as we are 5th on the possession table ahead of only Man U.

   Looking at the Action Zones help explain it further. Despite being 5th on the the possession table we rate a lowly 17th on the amount of play (29%) that goes on our defensive third. City 1st with 19%, Liv 2nd with 22%, Chelsea 3rd with 24%, Man U 4th with 26%, Tot 6th with 27%.

   We dominate possession @ 5th but do not dominate territory hence @ 17th. It shows that most of that our possession is in our defensive territory and that we loose possession quickly when we eventually venture upfield hence the high defensive numbers of our front players.

   The big question is whether our inability to keep possession up field is a lack of the requisite skill to do so upfield or that Emery’s tactics is geared to quick transitions which goes with high turn overs (high risk: high dividend). Tim what about further stats analysis along these lines?

   1. Important to remember that possession is just passing. It’s nothing to do with tackles (which are mostly about opposition dribbles and the team’s set up to attack players with the ball), even blocked passes and interceptions aren’t dependent on opposition passing. If I look at the fit for possession and INT and Blocked passes with possession, it’s just .46. That’s not bad but means that more than 50% of the reason why teams are getting INT and blocks is because of something else (the team’s setup, physical ability, etc.). If I add in tackles, that number drops to just .36. That means the majority of defensive stats have to do with something other than possession.

      No matter how I adjust for possession there are always teams like Bournemouth, (any Mourinho side), and (any Allardyce side), which have very low proactive defensive stats. Which makes me think that possession adjustments are highly problematic.

      1. Does the number of times possession change have a correlation with BIT numbers?

6. “then averaging among the players, Arsenal have the most active 10 in the League.”
   OMG!That really makes the defense look worse than it is.
   “Player Blocks Interceptions Tackles Total”
   I’d love to see recoveries too. Sometimes, recoveries are just as valuable as interceptions. Whoscored.com doesn’t show them, while the Arsenal website does.

   1. Recoveries are really a non-stat. Sorry, hate to break it to you. They are meaningless.

      1. Disagree.
         You clearly don’t rate recoveries, but there’s a reason why managers use that stat. Recoveries are at the heart of Klopp’s gegenpressing philosophy. For instance, in this interesting feature (https://www.wired.co.uk/article/premier-league-stats-football-analytics-prozone-gegenpressing-tiki-taka), an accountant in the 1950s noticed that “50 percent of goals came from balls recovered within 30 metres of the goal line, the last third of the pitch”. There was barely any data analysis at that time. That accountant was a stat guru before they even existed.
         Your answer is all the more troubling since tackles won and interceptions are also ball recoveries by definition. Winning a tackle, making an interception or recovering the ball are all about winning possession. I like recoveries as a stat because it tells a lot about a player’s work-rate, and also his positional play to a certain extent. I guess people value interceptions more because it requires to read the game well.
         If that can change your mind, here’s a scientific study about the importance of recoveries in football: https://www.researchgate.net/publication/260226858_Ball_recovery_patterns_as_a_performance_indicator_in_elite_soccer

         1. Thanks for this article , so can you identify the 5 most important actions we get right in our play at the Arsenal?

7. I’m wondering how good this stat is per 90. Seems like you could have sample size issues. I’d have not included Welbeck at all. He only played in 8 matches, and only started 1.
   And it’s hard to say if this is comparable between cases where a player started vs. where they came in halfway through the second half and whether or not we were trailing or winning when they came in.

8. Heard a story once that during WWII, the Royal Air Force was analyzing fighter jets to try to make them more battle-proof. They analyzed all of the planes and noted where they had been shot, so that they could fortify/strengthen them in those spots. Until someone pointed out that those were the planes that had made it back. When they get shot in other places, they don’t return! I think there’s a little bit of that here. Tackles, interceptions and blocks are showing where we are succeeding defensively. Not where we fail. If we made fewer, our goals against would be even worse! Defense is such a team effort, and our inability to track runs, failure to execute offside traps, and passing mistakes in our own end are the issues that happen over and over and wind up costing us. That was compounded Sunday by our inability to possess the ball much. In the first half, we had the ball a lot more, and until that second goal, were looking like we belonged. The second half we just rolled over let them possess the entire time.

   So yes, it’s nice to block, intercept, and tackle, but doing more doesn’t indicate we’re playing better defense. Sometimes you block, intercept or tackle to cover for someone else being out of position, failing to track, etc.

   1. agreed.

   2. Great comment LA

9. Nice one Tim. Not many will admit to their mistake.

   I too believed Wenger was the reason for our poor defence and that if he got good enough coaches in his backroom staff, he could have achieved much more.

   So football is not that simple, like the top coaches make out to be – like Wenger (when he won the double in his first season).

   Hope the hierarchy have a clear plan (wishful thinking) and back the coach to get us into title challengers sooner rather later.

   Honestly I’m just hoping Emery (or Arteta or any other young coach) will be able to take the step up and become successful like Wenger, Klopp, etc did. I have no faith in the hierarchy!

10. Wenger IS the reason why our defense is in shambles. Poor defending is in our DNA now. He bought one error prone defender after another in Squilaci, Santos, Eboue, Silvestre, Senderos, Mustafi… need I go on? The list is quite damning. For every Koscielny, he bought 3 Silvestres. To make things worse he didn’t organize the defense so they could at least play together at a unit, and this permeated throughout the whole team which made us tactically naive adding to the individual mistakes in the process. He got lucky with BFG at a time when things looked really bad cos that he was experienced and despite his lack of speed, him and Koz made a good pairing. Emery is, at least, TRYING to organize his defense but he can’t catch a break and is stuck with our master of disaster, Mustafi, because all our good defenders are injured. We have looked pretty good in set pieces (which tells me they are working on organization) and many of the goals we have conceded have come down to individual mistakes – which will take time to fix. Papa and Holding were just starting to show signs of a good CB pairing before Holding went down. I commented on your By The Numbers piece on ArsenalReview right after Holding went down that Holding’s injury was a worrying sign because he was our least error prone defender (not sure if the numbers bear that out, but at least he passed the eye test for the most part and didn’t make as much calamitous errors). The MOST important factor in having a good defense is having a constant back four. Everyone knows this. Injuries have pretty much made this an impossible task for Emery. I don’t envy him and I think fans need to be patient here. But here comes every Arsenal pundit on the internet talking about how Emery doesn’t play with an identity or how the defense has been just as bad as Wenger’s Arsenal conveniently forgetting all the injury problems or that our two most creative players are 1) leaving and 2) a diva. And this whole notion that Ozil somehow deserves to start because he is our most creative player is pure nonsense. He plays when he wants and he simply doesn’t do enough offensively to justify a place in the starting XI when he isn’t willing to help out in defense either. That’s not just under Emery – he was pretty poor last year for the most part as well. I, for one, am glad Emery is putting him in his place and will be happy to see him gone.

11. If we are so defensive, why are we still leaking so many goals? If we are playing more defensive, why are we still so poor away from home?

    ===

    Because stats don’t tell the whole story? 😀

    I feel sorry for Emery. Inherited Mustafi and a crocked Koscielny, a Derelicht, an ageing Monreal, an injured Mavropanos, and then loses our best CB for a month, and our best and only RB (through which he wants the attack to be directed) for nine months). Zero stability in our back line this season.

12. The first thing to establish, in my mind, is whether BIT or any part of BIT is associated with outcomes we care about, like winning football matches or expected goals against, or big chances allowed. If Arsenal are leading the table in BIT (as we can see from Tim’s numbers) but are trailing far behind in all other accepted markers of defensive efficiency (as they certainly are), then does BIT really matter? And if BIT can’t tell us which teams defend better, then what DOES it tell us?

    It’s an open question whether team BIT correlates with harder work at all. It seems to based on anecdotal observation, for example Torreira clearly hustles all game and has a really high BIT. But in statistical terms, in order to establish that, one would need to be able to show an extremely strong association between BIT and some gold standard value for hard work on the football pitch, which I do not think exists at this point. Even if it were proven to be such a surrogate marker for hard work, it would be difficult to make a compelling case that a player with 2.0 BIT is working significantly less hard than a player with 2.8 BIT. I mean, how important is that delta of 0.8? And how much is it saying about their average position on the field and what they are asked to do by their coach vs. about their work ethic? I don’t think these are questions that simple numbers can reliably explain. Still, this represents a step in that direction of more rationally understanding defensive actions and I thank Tim for gathering the information and presenting it to us.

13. There sure is a lot of sympathy for a processional manager that took a job on the premise that he had extensive knowledge of the squad and what he wanted from it. Blame Wenger, blame the players, blame the owners….my biggest grouse with Emery is the football is crap to watch. I really cant blame Kroenke for that can I?

    1. Professional sorry 🤦🏻‍♂️

    2. Yeah this. We have a better squad than last season with an elite attack, and instead of playing to our strengths we’re moaning about our weaknesses (which had been addressed to an extent).

       I mean if your style of play needs all these new players, maybe don’t say you’ve got plans for all these guys, will play the youth, and prefer to win 5-4 than eke out a 1-0. No sympathy for this change in tack.

       When Wenger came in he had to win over the old guard and work with them to build something special. Emery had the chance to do that with this squad, but he chose to throw a grenade to blow it up and apparently everyone’s convinced we needed to raze it to the ground anyway. I don’t buy it.

    3. The football has been crap to watch for 3 seasons now but sure why dont we blame Emery for the last two seasons as well. The only issue that’s been addressed is that we finally have a midfielder who can win the ball but the squad remains as unbalanced.

       1. How is it unbalanced? Except for lacking a competent backup RB..

          We added experience at CB. Improved at GK. Added the ball winner, and a silky skilled youngster at CM. Laca is now completely fit after his knee surgery last season. And He and Auba have been in rare form. Saving our bacon actually, because we are 12th in shots taken in the league. We’re not creating chances because the system isn’t working.

          And that’s fine too. I just doubt that the intention to play an attacking game is even there. It seems to me that Emery will always prefer to set up with caution and hope to win than open up the game. He does this even against small fry let alone not even trying against City.

          Actually, the problem isn’t that it isn’t working, but that I suspect it’s just about working to plan.

          1. Auba has not been an effective player out wide. He is great at running into good goal scoring positions but not a guy who creates plays. We actually don’t have anyone who plays well as a wide forward (Welbeck was the closest player to a WF we had) and I suspect given Emery’s tactics in using the wide spaces and full backs, this is a problematic area for him which is why he was trying to get an experienced wide player like Perisic on loan but had to settle for Suarez after that deal didn’t work out.

             The CM issues has been partially addressed, as I mentioned earlier,(though I think we need to replace one dimensional, error prone players like Xhaka and Elneny) and yes we added experience in CB but that’s an area where we have been hampered with injuries and haven’t been able to field a consistent back 4 all season. We have one full back on the right side (injured for the rest of the season) and none on the left side. Monreal has been a decent player in the past but he is too old now to be playing every week and Kolasinac was never a full back. For a system that emphasizes so much on the full backs overlapping, we are very short there. So yes, I think the squad is unbalanced, especially on the wide areas, which seems to be the bread and butter of Emery’s system.

             “Actually, the problem isn’t that it isn’t working, but that I suspect it’s just about working to plan”

             Come on Shard. Emery didn’t send out a team to eke out a 2-1 draw with Cardiff or lose meekly to 3-1 to City. Against Cardiff, we had a front three of Ozil, Laca and Auba. On paper, that’s our 3 best attacking players. And what did Ozil do? It’s not on Emery that 90% of the time, Ozil plays like a number 10 for Fulham rather than a number 10 for Arsenal. Against City, I thought we did okay for the first 42 minutes but we got undone by lapses in our concentration before finally just getting outclassed. And that’s just it – we simply didn’t have enough quality players on the pitch.

             I am not saying Emery is the guy to take us to the next level but I think given what he has to work with, he has done okay. I think he has steadied the ship and we should give him at least one transfer window to see if there are any real improvements. I just don’t understand the complaints about quality of football. People are quick to forget that the football has been dire for a long time. Emery hasn’t had any new attacking players to work with. If anything he has fewer, given Ozil’s refusal to work hard for the team on a consistent basis and Ramsey being non-committal. I would argue the home results vs. Spurs and Chelsea show that when the players are motivated and focused enough to carry out the manager’s vision, and we don’t have that many injury issues, we have seen flashes of Emery’s work behind the scenes.

             Wenger wanted his players to play jazz but in this hyper competitive world of football that we have now, that’s a pipe dream . You need structure and at least Emery has gotten his players to start reading sheet music. Maybe he isn’t brilliant enough to challenge the likes of Pep, Klopp or even Pochettino (though are you going to tell me you didn’t enjoy the 4-2 thrashing this year?) but there is some foundation of a structure in the team now which will make the transition to the next manager easier.

             1. It is Emery’s fault just like any other manager takes the blame if his players underperform. I hope you’re right though. I just don’t see it because Emery should be using the players we have better rather than trying the one tactic we lack players for (wide forwards)

                But if he does want to stick with it, then STICK WITH IT. I don’t get the constant switcheroos. Obviously the players aren’t going to be consistent if you do that. I had the dreadful thought today that Emery is Bruce Rioch to Arteta’s Wenger. The safe option.

14. if arsenal’s defensive numbers with their front ten are so awesome and they’re still leaking goals, it must mean that when opponents penetrate beyond arsenal’s midfield, the numbers must go to crap…….meaning the defense must be awful.

    i’ll readily admit that emery’s been dealt an awful hand with defensive injuries. credit to him because i haven’t heard him complain once about it. however, now is the time to show his pedigree as a manager. if your defense is garbage, why not focus on the strength of the team; the arsenal attack. even without the injuries, arsenal’s best players are in the attack. if your attack scares the piss out of everyone, teams will be afraid to throw numbers forward.

    i really feel bad for lacazette as he’s a pure goal scorer and is trying to get his goals while providing a bridge from midfield to attack. problem is mesut ozil is so much better at that than he is. it makes absolutely no sense to continue in this mold.

    i don’t know if i can continue to be an arsenal fan. it’s one thing to lose but quite another to play boring, bad soccer and lose. at least mourinho won. what has emery done other than desecrate the beauty of the game once played by our beloved team?

    1. once fans stop being entertained by arsenal football, they’ll stop paying the most money for that product. if that happens, suddenly emery and his strategy begins to effect kronke’s bottom line. then, he’ll likely be gone. we’ll see.

       1. So right Josh – we have become a dreadful watch, especially since the victory over Sp*rs. The Chelsea game was a positive blip, but even then we sat back and defended quite a bit. Winning and getting to the CL will aboslve many sins, but if this is an indication of Emeryball, it may not be so pleasing for a lot of us longer term. Soulless, defensive football will not fuel the pashun we have felt for the badge.

15. pogba doesn’t defend and look at the results of that teams since the manager has played him. he’s atp to their attack.

    1. I would love to see how Pogba’s BIT stacks up vs. Ozil. I have seen him defend, he just doesn’t do it the way Mourinho wanted him to i.e. sit deep and be more conservative with his forward movement. In fact, his natural style of play is the opposite. He is aggressive with this fwd movement and drops deep occasionally. In any case, he contributes far more with goals than Ozil does which makes him indispensable and I suspect if Ozil contributed more to our attack, Emery would think a bit more before benching him.

16. “Statistics such as line breaks and possession in the last third are important for City but would probably be irrelevant to a team with a different style: football analytics is a discipline in which the way a team plays dictates which statistics are significant. The challenge is to find out which. “Instead of looking at a list of 50 variables we want to find five, say, that really matter for our style of play,” says Pedro Marques, a match analyst at Manchester City. Marques and his colleagues are currently using data-analysis techniques, such as principal-component analysis, to home in on the match-related variables about winning.”
    *Culled from “the wired zone”*

    I think stats are 50% important, 25% tactics from the coach and 25% of the talent or football brain and application of the player.

    On sky I found that we have taken less shots than last season, but we seem to have a better point and goals accrual than the previous season.

    From the statistical approach at City mentioned in the article, how effective has our possession being?

    Emery stated he prefers a 5-4 win to a 1-0 win, how effective is he as a coach?
    Because stats show we have the highest defensive numbers of the top 6, while we have conceded the most goals, the strength of this argument is based on the notion in the article, that not conceding a goal is worth 2.5 points per game, while scoring a goal is approx 1 point on the average. And the article mentions Sam Allardyce using data to sustain Bolton until he left in 2007.

    How good are we really? Would better recruitment result in a chance to win the league with Emery in 2 years?

    1. Actually, let me answer that in a short post.

17. Sorry, the copied article is from the wired.co.uk, the zone idea came from the mention of prozone. 😂

    If Josh Kroenke is raving about Unai Emery and his similarities to the Rams’ coach Sean, perhaps it is best to analyze the coach’s failure at PSG in the Champions League against Real Madrid last season to establish a case for the ceiling of the man.

18. What are our individual errors stats within the defensive actions?

    1. We are still top of the League I think. It’s weird that errors aren’t published the way that they used to be.

Comments are closed.