Arsecast
I had a moment to speak with Andrew from Arseblog on the Arsecast yesterday. We talked about Arsenal’s end of season run in and the problem of Arsenal’s away form.
I’m not 100% comfortable with my performance on the cast. I don’t have any excuse other than to say that 1) I haven’t recorded in a while and 2) when I gathered my thoughts for the recording, they were written down in long form, like I blog, and that’s not conducive to a recorded conversation. So I was a bit flustered at times but Andrew made the best of it on the cutting room floor.
We also covered a lot of ground and that can be somewhat of a problem. I think when it comes to stats pods, we may need to do just a deeper dive on one topic. But we didn’t. So, in the interest of clarification I have put together the main talking points I tried to bring up:
10 – Arsenal have earned more points this season in away games than last season – 10 more points when we compare like for like games.
9 – Nine of those 10 points, however, come from just three games – Cardiff, Newcastle, and Fulham. You can figure out why that’s a problem: two of those teams weren’t in the League last season and Newcastle barely escaped relegation.
0 – Arsenal don’t play any relegation teams away from home for the rest of the season.
6 – The other big problem is that we won those games this season with goals from outside the 18 yard box – ok, maybe not Fulham, but certainly the other two. And Arsenal have scored 6 goals from outside the 18 yard box in away games this season. As you know – long time reader – goals outside the 18 yard box are a low percentage shot < 3%.
This is the part where I got lost in the weeds.
Ok, so, one thing that makes stats guys nervous is just overall shots (both taken and allowed). I know that you CAN win football matches 1-0 and never take a single shot (Arsenal have done) but that’s not really a strategy you want to pursue.
Last year Arsenal took a lot of shots in away games 13.2 per game – which is more than this year, 10.6. BUT.. Arsenal took a higher percentage of shots last year from outside the 18 yard box (47%) than they have so far this year (36%).
Last year Arsenal were taking a lot of what I would consider bad shots.
And, they weren’t scoring them. Arsenal scored just 1/117 shots outside the 18 in away games last season.
This season, Arsenal have already scored 6 goals from outside the 18 yard box in away games and another 5 goals outside the box in home games and the problem is that these goals are not superfluous.
Arsenal scored 11 total away goals this season here they are:
Wolves (home) – scored 1 outside, earned a draw, 1 point
Tottenham (home) – scored 2 outside the 18, 3 points
Man U (home) – won the game, 3 points
Bournemouth (home) – just 1 of 5 goals scored
Chelsea (away) – scored as 1 of 2 goals but we still lost
Cardiff (away) – won 3-2, 3 points
Newcastle (away) – goal outside the box decided the game, 3 points
Fulham (away) – just one of 5 goals scored that day
Cry Pal (away) – one of the 2 goals that day, 1 point
Southam (away) – one of the 2 goals that day, we still lost
By my count, low percentage goals from outside the 18 yard box have accounted for 14 points this season. I’m not saying Arsenal shouldn’t take these shots, that the goals don’t count, or that the wins and draws don’t matter because they are low percentage shots, I’m just saying that there has been a big element of luck so far this season for Arsenal.
Just to give you a data point – in away games, Arsenal are converting 11% of our shots outside the 18 yard box. That’s an unsustainably high percentage. In other words, luck.
Those goals outside the 18 also explain why Arsenal have such a high overall finishing rate: last season’s shot conversion rate in away games was just over 7% for Arsenal. 10% is normal. This season we are converting 16% of our shots. Arsenal shot more last year but scored less because they were taking bad shots.
33 – Another reason why Arsenal struggled last season was that they converted big chances at a lower rate than normal. Arsenal took a lot of shots outside the 18 yard box but they still created 1.7 big chances per game. That’s a decent number for away games and is virtually the same last year and this.
What’s different is that Arsenal converted just 33% of the big chances last season and this season they are converting 44%. 44% is basically League average. And if we look at last season again, just a few more goals off those shots would have made a big difference in the final points total: Arsenal failed to convert big chances against Watford (0/2), Stoke (0/2), Chelsea (0/2), Tottenham (0/1), and Man U (0/1) last season.
5 – Another major difference is in the Arsenal defence. For people who are only interested in what the table says, the killer stat is that Arsenal have just 5 away losses this year and over comparable games last year Arsenal had 9 losses.
However, what has me a bit antsy is that In terms of stats like shots allowed, Arsenal have actually allowed more shots this year than last (13.7 v. 12.5) but… and this is important, we are allowing fewer big chances per game this year than last. Not a lot fewer, 1.8 last year and 1.6 this year, but some fewer! The opponents are also converting at basically the exact same rate, 11.4%, both last year and this year.
1.5 – In terms of expected goals, because Arsenal are conceding so many shots in away games and conceding basically the same number of big chances, the expected goals allowed are virtually identical, 1.54 per game last season, 1.5 per game this season.
You really can’t say a team is playing good defense if they are expected to give up 1.5 goals per game in away matches.
The final five
Arsenal’s opponents in the final 5 away games are among the tougher teams for Arsenal to face: Everton, Watford, Wolves, Leicester, and Burnley. Arsenal only took 7 points last season from those matches. And Arsenal conceded three penalties in those comparable matches last season: against Watford, Leicester, and West Brom (which I’m comparing to Wolves).
7 – The good news is that I think 7 points from those final five away games is actually enough for Arsenal to finish top 4.
75 – Arsenal are on 60 points. According to the model at fivethirtyeight, top 4 will probably take 75 points. They currently predict Tottenham will finish 3rd with 75, Arsenal 4th with 75, Chelsea 5th with 72, and Man U 6th with 72.
So.. if Arsenal take 7 points from these 5 away games and 9 from the remaining 3 home games, that’s 16 points. That’s what’s needed to finish in and around top four this season.
Aubameyang
Another topic we touched on in the podcast is Aubameyang’s finishing.
We know from watching him this year that he’s missed a lot of “big chances”. But it’s worse that you might think. He’s actually got the lowest big chance finishing percentage of any of the top finishers I’ve studied this year.
I looked at all of the top scorers from each of the five big European leagues and Aubameyang has the worst big chance finishing percentage of all of them (28%):
Aubameyang first caught my attention because he missed a ton of good chances but he was actually not that bad back in the day. We don’t have the final stats for this season but this looks like it’s part of a decline in Auba’s finishing over the years:
| Season | Big Chances Scored | Big chances missed | BC Finishing |
| 2018/19 | 7 | 18 | 28% |
| 2017/18 | 15 | 25 | 38% |
| 2016/17 | 23 | 25 | 48% |
| 2015/16 | 22 | 20 | 52% |
| 2014/15 | 13 | 14 | 48% |
Where Aubameyang is different this season is in his “small chance” finishing. Back in his Bundesliga days he had seasons where he only scored one or two “non-big” chances. This season he already has 7. That actually leads all other strikers in the League and his small chance conversion is 15%, which is on par with Mane and Sterling.
I don’t personally care that much whether he scores 30 small chances, 30 big chances, or a mix of both. The only reason I point this out is because people get frustrated that he’s missing big chances.
As an aside, one reason why people think Mo Salah is struggling is because he is missing a LOT of those “small chances” this season. He’s scoring 40% of his big chances (which is low but not horrible) but he’s also only scoring 5% of his small chances (4/72). If it weren’t for Mane’s ridiculously high finishing numbers, Liverpool would be out of the title race.
Leno v. Cech
And finally, Andrew and I also touched on Leno’s goalkeeping stats. From the data, Leno’s been overall better than Cech was last season.
Cech 2015/16 = 48% BC saves and 64% SiP saves
Cech 2017/18 = 34% BC saves and 43% SiP saves
Leno 2018/19 = 54% BC saves and 57% SiP Saves
“BC” refers to Big Chances, not the comic strip, and “SiP” stands for Shots in Prime which counts all shots taken inside the 18 yard box from the penalty spot to the goal.
Cech was clearly a keeper in decline. His stats went from world class to just under average in his three years at the club. When Arsenal purchased Leno I was privately skeptical – his saves data looked decidedly average for the last three years – but he has clearly done well in an Arsenal shirt earning praise from a number of places. That BC saves number is especially good. That’s “David de Gea” level.
But his record home/away is extremely weird:
Leno home 2018/19 = 78% BC saves and 76% SiP saves
Leno away 2018/19 = 13% BC saves and 33% SiP saves
Cech home 2017/18 = 46% BC saves and 52% SiP saves
Cech away 2017/18 = 29% BC saves and 48% SiP saves
There was a slight accounting error in the data I gave Arseblog but this home/away split is still shocking:
Leno has saved 14/18 Big Chances at home this season
Leno has saved 16/21 Shots in Prime at home this season
Leno has saved 2/15 Big Chances away this season
Leno has saved 7/21 Shots in Prime away this season
There are still 5 away games left in the season. Based on the number of Big Chances Arsenal have allowed per game so far this season and last season, they will probably concede 9 big chances and the opponents will get maybe 7 on target. If he saves 5 of those 7 (imaginary) big chances on target he will better Cech’s away record by 1%.
These are very small numbers of course and that is totally doable. He has saved 1 big chance in each of the last two away games. But this away/home bifurcation feels unusual, like there is something else that’s the problem at Arsenal.
As I said on the podcast, and like I have said for a few years now about Arsenal’s home v. away form (where Arsenal are the 2nd best statistical team in the League at home but a mid-table team away), I think this is mental. Not the British mental, the American mental: where people have a mental block against something. For years under Wenger, Arsenal were a fantastic away team. Better away than at home some years. So, why are we suddenly so weak away?
And looking at Bernd Leno’s saves numbers, clearly he has the reflexes and positioning to make great saves – that’s what I read into high big chance saves numbers at home – so why is he missing them away?
Dunno.
I’m confused. Why aren’t those states?
It was mysterioso Tim…QAnon Tim, if you will 🙂
I apologize, Tim. That was a poor attempt at humor.
It’s all right dude. You’ve been huffing chemtrails again.
Ha. Indeed.
I thought your guest role was fine. Blogs is a more-than-decent editor, though.
Ok, so, the joke (in my mind) goes like this: DC isn’t a state and it should be and let’s face it, LA should really be its own state as well. So, I switched the places that should be states and well.. the rest is comedy history!
I still don’t get it.
Great work mate. Fantastic blog at your devastating best Genius..the stats are totally under your control…..
Thank you.
The problem with Auba is that he tends to be too casual when presented with easy chances. He tends to finish in a blasé manner that results in pulled/scuffed/underpowered shots.
The 1st time I noticed this was in the UCL when he played for BVB vs Benfica in the 1st leg last 16 tie 1st leg 2 years ago (BVB lost 1-0). He missed 3 presentable chances and a pen and was subbed by Tuchel in the 62nd min only to then score a hattrick in the 2nd leg at home when the pressure was on.
All this shows me its a mentality issue. If he took the easy chances seriously and didn’t score them in his head before he even makes contact he would have a better finishing rate.
You mean the guy who celebrates goals with a superhero mask with 20minutes to go, while the game and tie are far from done and dusted, might have some issues?
It is a distant possibility.
Wakanda Forever…
Wakanda Forever.
Leno has been huge. The delta over Cech is noticeable. He is so fast to the ground. When I saw Cech in EL games against Rennes, I was shocked at how slowly he got down to make a save. I also believe there was some stat about Cech making a huge number of errors leading to goals last year – and Leno has improved that significantly this year.
I think there are some very subtle but important advatanges a keeper has at home vs. on the road regarding positioning. When you leave your line, you see faimliar markers that help you establish position. maybe it’s where ad boards break, maybe it’s the color of a sign or a column – but you know if you’re 2 ft to the right or left, or a yard further out. Those tiny differences really impact your ability to make a save. You can understand opponent positions and angles just that fraction better. That’s what I remember from my keeper days. Nothing you could 100% point to, but just a comfort/familiarity that you get at home.
I suspect it’s exacerbated in Leno’s first season in the PL. After a few laps around the league, I bet the home/away dispartiy decreases. Does Kepa show the same home/away difference in big chances saved this season?
You make a very valid point
That’s a really interesting insight – thanks!
And I know neither you nor Tim said this, but I’d be surprised its all down to the keeper and rather partly what you’ve suggested, partly how defence is set up and how much pressure is on the shooter and also where the shots are coming from (eg if unpressured the shooter can shoot from a more-Prime place)
Really insightful comment.
PS – Heard your Arseblog appearance and it sounded pretty good to me. That points you make here came across clearly on the pod. Maybe editing helped, but I was surprised you didnt think it came across well.
Perhaps he’s just self-conscious? [Speaking only for myself but] I tend to replay recent conversations in my head, trying to pinpoint the moment/moments where I revealed my innate idiocy.
Sorry, this is meant as a reply to LAGUNNER’s PS.
Very likely. Hearing recordings of one’s own voice is disturbing, often leading to “I can’t possibly sound like that” responses. But it’s easy to be hypercritical after the fact. Sounding engaging and articulate in a recorded setting is really Hard!
Very true. I did a language test just a week ago, and it included a speaking section where I had to speak on a subject handed to me on the fly. I probably did fine, but the Hindsight Self-Loathing Machine in my head thinks maybe I went off the rails for a minute or three.
To Tim, sounds to me like you did fine on the podcast, really. Editing or not.
I was thinking about doing a weekly stats podcast. I wonder if you all would be interested in that?
I would be interested in any podcast you do Tim
I thought I was the only one who
mentally replays conversations, determined to find something that I said stupid .
More common than you might think!
Firstly, I thought you did a great job on the Arsecast! Secondly, has Leno’s away numbers improved at all since late January-ish when the team has improved (a bit) defensively? 🤷♂️ Basically from the time Kos has started to be back in form, have we seen an uptick in Leno’s numbers/confidence?
why tf are were playing on Monday when we have 2 rescheduled games in April, and about 8 overall?
Must be nice to be Man City. Get your game off early after the long interlull. United play in the 3pm kickoff, Chelsea, Spurs and Liverpool all Sunday. Not saying it’s deliberate, mind. But TV directed scheduling always seem to have it in for us.