Why Good Teams Should Be Terrified of Players With Bad Passing Stats

giroud_double_denimThose of you that follow me on Twitter may have seen me mention my general confusion at Arsene’s Wenger’s “big man” policy. For years, Wenger has purchased (or in Bendtner’s case, pushed the development of) tall forwards, while being generally happy to buy midgets for placement in the rest of the team.

This is odd for a lot of reasons.

The first reason is that building a speedy, zoomy, variable attack and then plonking a tall, slow guy in the center of it is strange. Granted, they are probably not that slow for tall guys, but compared to Walcott and Gervinho, Chamakh, Bendt, and Giroud are noticeably not fast.  It takes attacking build-up play that can be really difficult to mark, and suddenly simplifies it dramatically for the defense. Why? Van Persie moves like a cat, but he’s only six feet tall. Adebayor was the one tall forward that worked for Arsenal, but he has a fairly unique skill set in that he’s tall, fast, has a good first touch, and is totally unplayable when he cares. Which is about 10% of the time. If Wenger was going for the Adebayor ideal, none of these other guys come close to matching up.

The second thing that bugs me are the percentages.  People always say “you want your forwards involved in build-up play.” This is a general truism, but it makes sense. You want all of your attacking players involved in build-up play because it moves the defense around, and makes your attack less predictable.

But what if your forward isn’t very good at passing the ball?

On good teams, forwards make a minimum of 20 passes a game, and most are closer to 25 passes per game, almost all of which are in the opponent’s final third.  Nearly every single errant pass kills an attack dead.

This is what the chart looks like for errant passes at the different passing success percentages.














A 65% passer kills five more attacks per game with a bad pass than an 85% passer. That’s a significant number, especially if you want this player to regularly be involved. Obviously there are plenty of other ways to look at loss of possession stats, but passing percentage is pretty important. This is especially true if most of your outfielders complete 85% or more like at Arsenal.

Olivier Giroud only completed 64% of his passes last season for Arsenal.  With that low of a percentage, he can’t be that involved in the build-up play, because assuming they don’t find a way to simply give him better passing options, he bleeds possession away at a massive delta with every pass compared to the rest of the team. For reference, Van Persie was 79% last season with Arsenal, and someone like Adem Ljajic was 88% with Fiorentina (which is insane for a forward).

Now take it one step further. Assume your team plays an intricate passing system that needs every attacker intimately involved like Barcelona. How many passes per game does Messi make?


In the league, despite the fact that teams likely try and mark him more closely than any other player in the world, Messi completed 85.2% of his passes, averaging 55 a game. Barcelona can’t even think about having a player like Giroud in their team, because they could never involve him.

At one point, Arsenal were considered Barcelona North. They had enormous possession, great passing stats, and a gorgeous attacking style that was the envy of most teams outside of Spain. Then the talent bleed began, something changed in Wenger’s plans, and guys like Adebayor became Chamakh, and Van Persie became Giroud. On the surface, this seems okay, but looking at the percentages, it’s just plain strange. Maybe Arsene sees something big forwards can exploit more than someone smaller, but man, the cost is enormous.

General Arsenal confusion aside, the point is this: All good teams should fear putting bad passers in their lineup, because all good teams are inevitably playing the percentages, whether they realize it or not. Or to put it another way…

If you’re a good team who wants to buy a big forward, he needs to come equipped with silk boots.


Post Script –I don’t mean to pick on Giroud here, he just happens to be the most obvious example. I actually think big guys like him serve an excellent purpose overall as later game subs.  Their height, muscles, and general athleticism seems to come in really handy against defenders who are already tired, especially if they have been chasing speedy forwards around for 60-75 minutes already.


26 responses to “Why Good Teams Should Be Terrified of Players With Bad Passing Stats

  1. Maybe Giroud doesn’t want to pass the ball for sake of it, he always look for final ball. Which is also results of some his wonderful assists and many turnovers. Plus Giroud gives a Physical balance to our team in set piece,both defense and attack. If he can shoot better and convert half chances then he can bag 20+ goals.

  2. It depends on your role in the team & wat u’re asked to do. If it’s just to be a finisher like Van Nistelrooy then u can’t expect good passing stats

    • I don’t expect them to pass like Arteta, but Gomez was 77.2% last year 78.8% this year. Higuain had a bad season for him at 74.8% this year in La Liga, but 83.6 in the CL, and 85% last year. The type of balls your receive will affect some passing, but 64% is awwwwwful.

      • Like I said, it depends on what the striker is asked to do. Van Nistelrooy was the best striker in 01/02 but he’s passing completion was worse than Giroud’s at between 60% – 65%. And I’m sure many managers would take hime with those goals & poor passing stats. Also if u think about it, a striker is the one player who’s “supposed” to give away possession coz he’s at the apex of the team, with no one infront of him.

      • I disagree with 3ent. Ruud was an excellent goalscorer, but United were a worse footballing side with him in it. In 5 years at United he won only one title.

      • Somewhat agree with 3ent below; taking it at a slightly different angle: Arsenal’s ‘problem’ is often that we stroke it around in the ‘safe zone’ in front of the bus we’ve allow our opponents to park, by our not being incisive enough.
        “Have a [bloody] shot!” comes to mind often enough.
        For that reason, one could argue that having a system where your key attacker is a big man who does various jobs (hold up play, head ons, handful for defenders allowing pacey wingers in, etc), one of his jobs is to take responsibility to try to create something from nothing, to have a pop, etc.. He’s the #10 after all.

        My problem with these kinds of stats is that – unless contextuality is really seriously scrutinised – they aren’t necessarily illuminating, and can be misleading. One thing I’d love to see (and potentially utilised by the team!) is an analysis of when having a speculative shot is better than continuing to stroke it around.

        My thinking is this: say stroking it around in the safe zone has an expected pass completion probability of 95% per pass, and an ‘assist’ probability of 3%, 50% of which will be scored. Numbers probably aren’t perfect but you get the gist, right?
        And speculative play has an expected pass completion probability of 55% per pass, an ‘assist’ probability of 25%, 75% of which will be scored, AND a direct goal probability of 10%.

        After how many ‘safe’ passes, which offer a cumulatively low probability of leading to a goal, is there a >=50% chance that someone will have shanked an easy pass and turned over posession? And therefore: after how many passes does it make sense for the team to ‘go for broke’ and have a silly shot from range, speculatively chip one to the edge of the 6 yard box, etc.? The chance of turning it over is higher than if you carried on stroking it around, but the reality is that stroking it around would probably lead to it getting turned over eventually anyway, and – especially if you have a shot – you’re turning it over in a controlled way (avoids a break) AND have a low but tangible chance of getting a goal from nothing.

        Interested to hear people’s thoughts on this half-formed idea. It’s been rattling around in my head for a few years!

  3. I have very simple explanation. Messi, for example, receives the ball through short passes and so at this very moment he has nearby teammates. Giroud, on the other hand, is used for long balls and at the moment of the reception, he is all alone upfront, waiting for the fast wingers to emerge.
    I dont know why Wenger is doing this. He is also using Sagna in the same way.

    • confused about the Sagna comment, not sure if i’m being dumb here, but…. how is Wenger using our right back to receive long balls, alone & up front, waiting for the wingers?

  4. Pingback: Transfer hay fever | Arseblog … an Arsenal blog

  5. Pingback: Transfer hay fever - Unofficial Network

  6. I’d say that Sagna received 2-4 long balls (in the form of goal kicks) from SZCZ each game in the 2nd half of the season. He would then attempt to flick these on to Theo or occasionally into the middle of the pitch to the midfield. Worked fairly well and helped draw some of the opposing team’s pressure on our backfield away.

    • By the way, that full body acid wash denim suit that Giroud has on is even more terrifying than his 65% passing.

      • maybe he cold wear it in matches to do the classic “scaring the life out of defenders”, but from a style angle.

    • Ah, fair point, and yeah, good spot. Makes sense to a certain extent since he’s good in the air, and tends to be quite forward since lots of the attacks build up from his side, so he’s already near the right position for that.

  7. I think one notable difference since the ‘talent drain’ is how many shots we concede during a game. As a result we have more goal kicks and therefore more long punts down the middle. I think Wenger’s realization post adebayor was the number of seconds you save building an attack with a long punt up and that’s probably hurt Giroud’s passing stats.

    This isn’t to defend him, 64% is poor form for any team, but I do think it would explain a great deal. The biggest compromise Wenger was making when he used Gervinho, Walcott or (disappointingly) even Podolski was that he lost that forward pivot who could actually do something with a long pass from the back.

    As a result those conversions happen at a place where the opposition can launch another attack and push us back again. Giroud’s failure to be exceptional on the ground and in the air are probably a greater indictment than his failure to score simple chances.

    That being said, his defense of his own season is possibly worth reiterating. Even though he was a below average passer, an average header of the ball (outside of the box) and a half decent finisher he did end with 16 goals.

    In more ways than one, we simply cannot afford to miss out on a striker as finely balanced as Higuain.

  8. Pingback: Alvaro Negredo - Unofficial Network

  9. Pingback: Holy Shit, Football Is A Game Where Every Team Plays the Percentages | Mixed kNuts

  10. Pingback: The Numbers Game → The Arsenal Column

  11. Pingback: BENTEKEWATCH: Christian Benteke Looking Likely To Stay With Aston Villa - Unofficial Network

  12. Pingback: Things We Think We Know About Football – July 2013 | Mixed kNuts

  13. Great work again.
    Isn’t his passing success rate excluding headers actually 77%.
    It’s his availability as an option for long balls and crosses which become headers for Giroud that hurt his stats so much. Success rate for headers is always low for almost anyone.
    He clearly has touch and movement and his linkup play “seemed” to improve as he built confidence.
    It will be interesting to see how he improves in his secong EPL season.

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 )

Google+ photo

You are commenting using your Google+ 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 )


Connecting to %s