Last season Arsenal played with a three headed goalkeeper: Szczospiniano. Together they conceded 36 goals but only three goals from shots outside the 18 yard box. Arsenal weren’t the very best at shutting down on shots from distance and allowed 43 shots on target from outside the box, but Szczospiniano saved 93% of the shots from distance that they faced. In terms of stopping shots from distance, that was the best record of any team in the Premier League.
This season, Arsenal have played with Petr Cech for 33 matches and Ospina for the other 4. Arsenal have conceded more shots from outside the 18 yard box this season than last (65 v. 43) but the saves percentage has plummeted from 93% to 83% as the Gunners have allowed 11 goals from distance in the Premier League this season. In terms of stopping shots from distance, that’s the second worst record of any team in the Premier League.
Kevin de Bruyne’s scuffed goal added one more to Cech’s total on the season and he has now allowed the most goals from outside the 18 yard box with 10. Cech also has the worst saves % of any keeper who has faced more than 20 shots on target from outside the 18 yard box with just 82%. Over the last two seasons, the League average has been 86% so, Cech is performing well below League average on shots from outside the 18 yard box.
Ospina allowed the other goal from distance in League play and he ended up saving 9/10 shots from that area. Little known fact, Ospina’s middle name is small sample size, David “small sample size” Ospina.
Cech has the worst record for saves outside the box but the best record for saves inside the box. Shots on target inside the 18 yard box are scored at about a 40% rate over the last two seasons. Cech has stopped 71% of the shots he’s faced in the 18 yard box – meaning that the opposition scored just 29% of the shots on target that they created inside Arsenal’s 18. That’s incredibly good. The best in the League! De Gea is second best at 69%, Schmeichel 66%, Lloris is slightly better than average with 63%, and Joe Hart is bog average with 58%.
This odd record is compounded by the fact that Arsenal have allowed 36 goals this season — the (second) most from the outside but the least from the inside. Using a rudimentary “expected goals” application if you take the shots Cech has faced and multiply them by the season averages (14% inside, 40% outside), we would expect to see 36.5 goals but Cech has only allowed 31 goals. In other words, Cech is well below average saving goals from distance, but also well above average from inside the box, and combined, he’s a slightly above average keeper this season.
I suspect that part of the problem with Arsenal’s goals from distance is systemic with Arsenal this season. The first problem is that the team is allowing more shots on target from distance than they have in seasons past. That “on target” bit is important as it indicates that overall shot quality from the opposition is higher than expected.
For Arsenal the shots on target trend for the last three years is mixed: 148 this year, 125 last, and 151 two years ago. I suspect many people will ask whether Coquelin’s inclusion has anything to do with this and the answer is maybe. Szczesny and Ospina’s reigns are almost exactly bifurcated along the Coquelin line. Szczesny allowed 21 goals, 19 from inside the 18 yard box. His save percentage from outside the box was great, but his save percentage from inside the box was just 25/44, 57%. Meanwhile Ospina allowed just 10 goals inside the 18 yard box and faced 6 fewer shots on target saving 28/38, 74%. Oddly, both keepers faced exactly the same number of shots from outside the 18 yard box, 18 with Ospina allowing 1 and Szczesny allowing 2.
All shots from distance score at about a 3% rate. So, in theory, you want the opposition shooting from outside the box. However, shots on target from distance score at a 14% rate. That’s a substantial increase in chances. In which case, your defense needs to close down on those shots and especially right in front of the box where the player has the best angles.
This gets to another reason why Cech might be letting in more goals from distance this season compared to Szczespina. This season, we have all seen Arsenal beaten from distance with shots that weren’t properly closed down: Bolasie, de Bruyne, Routledge, Herrera, Martina, and Joel Ward (Palace) come to mind.
But just as equally, Arsenal have been unlucky with some of these drives from distance: Yaya Toure’s chip was unstoppable, as was Harry Kane’s curler from a tight angle, Ross Barkley’s shot was deflected off Gabriel, and Firmino’s shot into the top right corner couldn’t have been stopped if Cech had those Betelgeuse comedy long arms attached to his real arms.
Cech has been beaten more often from distance than anyone would like and on some of the goals conceded he’s looked quite weak. But every keeper is going to have a few mistakes. More to the point, players are going to jump out of the way of shots (all defenders do it), not close down, and generally make it look bad from shots from distance. I suspect, and it’s something I’m going to look at this summer, that shots from outside the 18 yard box (from open play) are basically all flukes — there is too much random variation in distance shots to make them something to blame on the keeper.
What I’d much rather have is Cech’s 71% saves from inside the 18 yard box. This is a much more important metric because once the opposition gets that type of penetration and gets the shot on target, their chances of scoring jump from 14% to 40%.
The only two questions is whether Arsenal can get the defense together to defend the distance shots better and whether Cech can repeat his incredible saves percent from shots in the 18 yard box again next season.
@7amkickoff
Sources: exhaustively sourced from Whoscored.com and Squawka. This information is seriously not readily available.
Allowing 50% more shots from outside the box than the season before. Nuff said.
Yes, but less than the season prior. So, not quite enough said.
I know keepers make mistakes, and Cech has saved us many points this season, but conceding these goals at his near post is getting quite frustrating to watch. One a season is ok, but it’s happened three times now. I’m sure both him and the gk coaches realise this and will work on it.
As a goalkeeper i can tell you that in many of the cases Cech has let one in at the near post its the defenders job to block it allowing the keeper to only cover around 70% of the goal.. That is what Chelsea were so good at while defending with especially that cunt Terry, they know how to block, Arsenal defenders do not. So i say dont blame it all on the keeper, he expects his defenders to do their bit.
At last someone who understands goalkeepers can’t cover the whole goal eventhough it is outside the penalty area. Cech expected Gabriel will block De Bruyne’s shot. And also that ball went between the Gabriel’s legs and that is even harder for a goalkeeper.
I expect Cech to have a major impact on defenders this Summer. He’s an experienced keeper and he will either straighten the defenders out in front of him or they will have to go. He’s not going to let his career end this way.
But why would he have waited for the summer to do this?
goalie in 2 sports (hockey and footie). I totally agree with Swedishgunner, we look really wimpy as a defence when it comes to blocking shots. All of our CBs (especially Kos sadly) turn on every shot. As an amateur in many sports, those of us who have been hit by a baseball pitch or blocked a hockey puck sort of laugh when watching footie players turn away from a ball filled with air. I hate that you made a Chelsea comparison but that team excelled at this. Leicester were also very good at throwing themselves in the way of shots… Read more »
As a defender in footie, and being drilled many times on getting “big” and closing on shots near the penalty area…it still hurts mate. Don’t get me wrong, I agree that our backline can shy away at times, but I don’t know if it’s substantially more than any other defenders from other teams in the average game. Deflected goals are luck, like fumbles in American Football. As a natural reaction, which is very much trained, drilled, coached, instructed, out of a player and top professionals alike, when someone cracks a thunderbastard at you it generally hurts to take square in… Read more »
Top notch analysis! We need more protection of Petr from our midfielders and defenders around the edges of the box.
I wonder whether, as mentioned on the Arsecast, there is a specific issue with his ability to dive low to his weaker right side?
‘We need more protection of Petr from our midfielders’
Yup. My favorite arsenal team I’ve seen was 98. Top keeper, legendary back 4, but crucially, 2 large, aggressive midfielders ‘sitting’ in front of them when the other team had the ball.
It’d be nice to have some numbers from the time Chech was at Chelsea
Also waiting for the analysis on Granit Xhaka when the time is right.
It would be interesting to see who the 2 central midfielders are when those long shots do come in. While our back four has been largely unchanged this season, our midfield 2 have been rotated several times. I suspect that Ramsey/ El Neny would play higher up the pitch than would Coquelin/Carzorla, leaving the space between the midfield and the defence, for example.
That’s some serious number crunching. Cheers
Great work Tim. Thanks for providing those stats.
As a mathematician I understand how much time and effort all this takes; you are awesome 7am, thanks for all the hours you spend to enlighten us.
71% inside the 18 is seriously impressive. We can work with that. Now just get rid of Koscielny’s new Grenadier Guard blocking technique.. (I googled that…”what are those guys with the hats who stand outside buckingham palace”) I wonder what the Gent refers to them as
7amkickoff–100% fucking excellent
Athose would be beefeaters
Beefeaters guard the Tower of London, don’t they?
Thanks for the hard work of sourcing the information and parsing them for us lesser mortals, Tim!
Yours is truly a unique and much valued voice in the Goonerverse 🙂
Cheers 7amkickoff
Impressive work
Another stat that shouldn’t be ignored in this debate is when a defender attempts to block a shot, standing still with both arms behind his back is 100% ineffective and is sure to make him look like a completely useless fucktard.
Thanks for that.
How does the saves from outside the box number compare to his at time at Chelsea? i.e. did Cech always have that weakness, but it was masked by the fact Chelsea’s defense was so organised it hardly allowed any shots on target, from distance?
Off topic but no news on Welbeck’s injury ? Why is it taking three days for a scan ? I am eagger to hear the news that he will be OK for EURO 2016.
Ummmmmmm 🙁
Give me defenders who block by attacking the shot – irrespective of the possible deflection into the goal – rather than the “try to twist round to get my back in the way and hands out the way” (also known as the “don’t hurt my face” approach). Kos and Gabby are experts at the latter.
Watch those teams who get results at Emirates and you’ll find that bravery brings points. The exception strangely being Leicester who do the brave bit but we still beat.
Good article, but the ending is kind of bizarre: “What I’d much rather have is Cech’s 71% saves from inside the 18 yard box. This is a much more important metric because once the opposition gets that type of penetration and gets the shot on target, their chances of scoring jump from 14% to 40%.” Surely save metrics inside vs. outside the box are *equally* important. A goal is a goal, regardless of where it’s scored from, right? And it doesn’t make much sense to treat the opposition’s chances of scoring from shots on target as being independent of the… Read more »
Great stuff 7am! Whether or not it’s a 1:1 comparison that goals are of equal importance is dependent upon the number of on-goal attempts. Most attempts at goal that are on frame come from inside the box, so if you save a higher percentage of those it more than offsets the odd extra goals over the same period of time from distance. With Cech’s numbers above, if he improved from distance from 18% to average 14%, it would be a goals difference of -2. But if he went from his 71% from inside the box to an average 60%, that… Read more »
*In the first sentence above “goals” should read “metrics”.
Time to sack my editor.
A keeper with a higher save percentage on goals that are more likely to be scored points towards more points saved.
Hes making the tougher saves and letting the ‘easier’ ones in for various reasons, which could be Our defenders being to keen to spread their legs.
Well… Never liked him
Great. This is why it takes an extra two days to get the By The Numbers thread.
The weirdest goal was opening day, with Zarate beating Cech at the near post.
If you run a Fisher’s exact test on the saves from distance comparing 15-16 with 14-15, where there is the biggest difference, the p-value is 0.23, indicating a 23% chance that the difference is random. That means the difference between last year and this year is not statistically significant. You might call Ospina “small sample size” but I would wager that none of the differences shown here would be statistically significant, because all of the sample sizes are too small. That is not to say there is no difference, but the number of observations is too small to be certain.… Read more »
I agree with you. Thanks for the suggestion.
Thanks for the analysis, a great read. I am sure that there is a definite case to make that we need to close down opponents more effectively in the outfield; we do seem to invite teams onto us, allowing them to get within strike range. However, I wonder how much the stats are caused by Cech also being ever so slightly slower than he used to be- the weekend’s hand definitely brought to sharp contrast the fact that Cech couldn’t quite get down to shots that were just out of his reach. I can’t help thinking that a younger Cech… Read more »
I think our defenders need to grow a bigger set of cojones and get better at blocking shots. Take the two City goals on Sunday as examples. On the first goal, Koscielny is just standing there watching with his hands behind his back while letting Aguero shoot. Why not try to narrow the angle instead of just staying static? For the second goal, not only is not Gabriel not close enough to DeBruyne, but he also gets turned like a gyroscope. How can you be that far away and still get turned like that? His last minute tackles are fun… Read more »
We can’t count Cech’s 2014/15 stats because they are too small. Cech’s numbers at Chelsea in 2013/14: he conceded 1 goal from distance and made 34 saves, that means he saved 97% from distance two years ago. He also allowed 22 goals inside the 18 yard box and made 46 saves, a 68% stopping rate inside the 18. Cech’s numbers at Chelsea in 2012/13 (that is as far back as I can go with this shots from distance stat). He conceded 6 and saved 46. That means he faced 52 (similar to Arsenal) and saved 88%. That season he also… Read more »
Yes as you mentioned some of the shots he conceded were pretty much impossible to save.
Also, I apologize for the verb tense fail in the final paragraph. My alarm was ringing to tell me I had to take the little one to the doctor so I didn’t proof.
Fascinating stuff, thanks Tim! Slightly hilarious to see Mannone and Fabianski up there as two of the best long-shot stoppers in the league. One suggestion – I’d be fascinated to know how other conditions affect the data. For instance, the weather. How much is the keeper’s ability affected when the sun is in their eyes? Does the angle of the sun relative to the keeper’s horizontal sight make any difference? What about when it’s raining? Maybe if it’s not too much trouble you could go over each match from this season, taking into account which way each stadium faces, the… Read more »
I think we have a problem with Cech. For me he appears visibly slow. The shots that he has allowed in have not been down to his positioning (which is excellent and explains why he is so good at shots inside the box and dealing with crosses) it is down to slow reflexes and movement. I think teams are going to exploit this next season.
If his reflexes were slow, he’d have let in more shots from inside the box because he gets less time to react.
Maybe it’s his anticipation or reading of the game outside the box, or may be he’s just shortsighted☺. I don’t think he’s the only keeper with this problem (hence Spuds conceding second goal to Saints ?) but it is Cech’s and Arsenal’s defensive weakness this season.
So Ospina had 74% inside and 95% outside last season. That’s ridiculously impressive. Why this guy remains so underrated is beyond me
Cech is great, Ospina is pretty bloody good too. Both have made errors, but who hasn’t this season & stats can show a lot, but Cech brings more to the side than just his saves.
Why our team can defend really well as a unit when completely written off, but can’t defend at all when we expect them too. That’s the main thing Arsene needs to worry about surely?
Where was all this serious number crunching when Ospina was being crucified for letting goals in? I know you guys have a favorite keeper it’s too obvious to see. However, it’d be only fair to measure all keepers with the same measuring stick and without the name calling on this blog…I’m yet to see Cech being called names but yet Ospina is “small sample size” “can’t kick far enough” “guy with no neck” “too short” “not tall enough”. By this I’m in no way defending or favoring one keeper over the other. Just want us to get behind our goalkeepers… Read more »
Small sample size relates to the fact that he didn’t play enough games for the sample of data to be large enough to consider… No one was calling him any names.