May 14, 2015
Disclaimer: within this article I use various stats. I am most definitely not a statistician, and I apologise in advance if any statisticians read this and cringe at their use within my piece. I hope that I balance their use with descriptive sections about the way that I see our play (rather than just stating the way that the stats see our play). Please feel free to leave a comment to explain any flaws you see in my logic – it’s always good to learn).
So, here goes.
There have been many excellent articles recently about Spurs’ horrendous defensive performances in recent weeks. The ever-reliable @brettrainbow nailed the defensive performance against Stoke City and, before that, had written well about our defensive midfield problems.
With such loose defensive play, you might think that Spurs should expect to see plenty of joy at the other end of the pitch. However, as James Yorke pointed out in his article this week, ‘Tottenham take a high percentage of their shots from range and struggle to create opportunities inside the box.’
It seems on the surface that we don’t have many players in our squad capable of picking the lock of the opposition – of exploiting gaps and seeing the next move before the opposition defence.
Key passes can be seen as a slightly crude measure of creativity, since they are defined as a pass leading to an effort on goal. When you have a player like Gareth Bale, for example, that can just be a square pass on halfway! Across a season, however, the data is useful, and a comparison of key passes per 90 minutes (KP/90) across the other top seven teams is interesting – note that I only include players who have played more than 500 minutes.
The first thing to note is that the top three clubs each have two players with more KP/90 in the Premier League than any of our players. Manchester City have four with better KP/90!
Most of the players listed – certainly in the top threes across the board – play as the forward or in the band behind the forward in their respective teams’ set-up. The notable players that don’t are Steven Gerrard, Jordan Henderson, Aaron Ramsey and Santi Cazorla, who have mostly played in the central midfield zone. Cesc Fabregas has moved between the number ten role and playing as one of a double-pivot.
It’s no surprise that Spurs’ top three have been the three most regular incumbents of the band of ‘3’ in our 4-2-3-1 in the league this season. That could mean that the players most suited are getting the game-time, or that a consistent run in the team leads to a better return.
Whilst Spurs’ top three lead the chance creation, the assist output is slightly different. Lamela has more assists than any other Spurs player in the league (6), Nacer Chadli is second (5), but Christian Eriksen (2) is behind Danny Rose, Harry Kane and Ryan Mason, and level with Andros Townsend, Aaron Lennon and Nabil Bentaleb. This is despite taking a lot of our set pieces (20 of his 79 key passes come from corners or free kicks). There are actually 76 players with more Premier League assists this season than Eriksen. James Tomkins (3) has more assists than Christian Eriksen! Staggering.
Perhaps Eriksen is just unfortunate – perhaps players have just not scored from the chances that he’s created. Of players who have made more than five appearances, Eriksen is 21st in the Premier League in terms of chance creation per 90 minutes (James Tomkins is 280th; have some of that, Tomkins!). That’s not elite level, but it is passable. He’s been our best at creating chances.
Just to linger on Eriksen a little longer as he is our chief chance creator – he has made 79 key passes, of which 15 were corners and five free kicks. But only three were through balls. That’s a record of 0.1 through balls per 90 minutes – joint 40th in the Premier League for through balls per 90 minutes for those who have made more than five appearances. In fact, those with the best record for making through balls for us in the Premier League are Lennon (0.3 per 90 albeit in only 276 minutes, a tiny sample size), Paulinho (0.2), and Roberto Soldado and ÉtienneCapoue (both 0.1). Eriksen makes through balls at a near identical rate to Bentaleb, Mason and Erik Lamela.
Eriksen has managed 10 goals – as has Chadli – and other than Kane (20), they are the only Spurs players with more than two Premier League goals. That’s astonishing in itself. In terms of minutes per goal or assist so far in the Premier League for us, our top five are:
1. Kane, 96.4 minutes
2. Chadli, 148.5 minutes
3. Townsend, 185.8 minutes
4. Eriksen, 248.5 minutes
5. Lamela, 264.9 minutes
This doesn’t make good reading for Eriksen, who has more opportunities to both create and score as a regular set piece taker. It also shows how Chadli justifies his inclusion, despite some poor overall performances.
It is well worth noting that Eriksen has famously covered more ground than any other players in the Premier League this season. This has left him fatigued, heavy-legged and well below his best in the final third of the season. Early in the season there were many articles questioning how he is adapting to Pochettino’s high-press style, and there was a period in the middle of the season where he not only seemed to be adapting, but truly flourishing. Hopefully after a pre-season break he can come back refreshed, and with a stroke of luck he will have some direct competition next season so that he does not have to play so many matches.
So why are we struggling to create chances? Is it that we have a forward that doesn’t read passes? Do we lack other bodies making run into the box? Is the system failing us? Or do our players just not have the vision? Is ‘all of the above’ a cop-out?
Harry Kane’s movement is good. He excels at drifting into the channels, and coming deep to collect the ball and linking play. He finds space well in the box, and we have seen him be in the clichéd ‘right place at the right time’ on many occasions. But he is not a player that regularly runs in behind defences, or who makes regular runs off the shoulder off a centre-back; he doesn’t have the pace to make that worthwhile. So we do legitimately lack a through ball option at times.
In recent weeks we have lacked support in the penalty area. Ryan Mason has notably made more forward runs – missing a glorious chance against Manchester City, for example. But where Chadli looked good at making runs off the shoulder of the defence or to the back post earlier in the season, he’s failing to do this with any regularity. Eriksen himself does most of his work outside the box, and Lamela does not seem overly keen on getting beyond Kane either.
Creative vision comes from confidence, and it could be argued that our teams’ confidence has been low for much of the season (for various reasons). But we have only played 15 through balls all season – Eriksen with 3, Lamela 2, Kane 2, Bentaleb, 2, Mason, 2. Chadli has not played a single through ball this season. To not have completed one through ball in well over 2000 minutes of football suggests a systematic problem.
Mauricio Pochettino wants us to win the ball as early as possible, but it is pretty clear that we are then are not doing enough to spring opposition defences. Our movement off the shoulder is insufficient, and we lack players who will look for runners – these issues could work in tandem and create a self-fulfilling prophecy, whereby our attacking players don’t make runs in behind the opposition defence because they don’t expect the ball to come to them, and players don’t attempt through balls because our forwards are not making runs – or are making insufficient runs.
Kane has been double and even triple-marked recently, and so one might have expected Spurs’ other attacking talents to profit from this. In theory, Kane being occupied should mean that there is more space for others to exploit, and should have led to an increase of goals and assists – or at least of chances and shots. If I had the time I’d love to delve deeper and investigate whether this is the case. On the surface we seem toothless with even less creativity present than previously.
There’s no simple answer, but there are things we could do better.
In my opinion, Erik Lamela has been one of the few positives in the last few games, but if we are to persist with this 4-2-3-1 (*sigh*) it would be a good idea to experiment with him as the ‘number 10’. He has shown signs of having creative vision, and also has the intensity in the press which we’ve really lacked at times. Lamela is second to Eriksen for non-set piece KP/90 (ignoring Lennon, because the sample size is way too small).
Eriksen should probably play wide on the left or the remaining games (if at all). When Eriksen has played on the left previously there has been uproar on social media and cries for him to be moved into the centre. And yet in the centre he seems to frequently be crowded out by opposition defensive midfielders, whereas on the left he finds pockets of space to work in.
In a recent article about Alex Pritchard for The Fighting Cock blog, Joshua Olsson argued that Eriksen has actually been more productive from the left:
Eriksen has started 10 times in the Premier League on the left this season, and has accumulated 4 goals and 1 assist in those appearances (he has also played 4 times on the right and scored 1 goal).
By contrast, in his 22 appearances in a central role, he has scored 5 goals and made 1 assist. These numbers alone would suggest that Eriksen is more of an attacking threat from the left, where he is able to find space, get time on the ball, and come inside and shoot on his right foot.
Again, the sample size is small, but it is an interesting point to keep an eye on. Creative players are being pushed out to the wings more and more as teams play 4-2-3-1, often with two dedicated holding players. David Silva thrives in this role for Manchester City, for example, and Eden Hazard is much the same for Chelsea. Perhaps teams’ most creative players will play as an ‘inverted winger’ by default now, rather than as a number 10.
Chadli is a strange footballer. He scores goals at a good rate, his non-set piece KP/90 is nearly identical to Lamela’s, and yet his work rate is substandard and so often he flatters to deceive in his play. We need to get more from him but, equally, perhaps he will always be a player that doesn’t do a great deal across a match, but who will pop up with a vital goal – much like Dempsey did a few years back. There is value in these types of players, and if we can encourage him to play on the shoulder and make more runs to receive through balls, he could easily replicate what Jay Rodriguez was achieving at Southampton.
Interestingly, Mousa Dembélé comes fifth in terms of non-set piece KP/90. That came as a surprise to me, but it makes sense, since over half of his appearances have been as an attacking midfielder, or number ten. His rate is just 0.15 off Eriksen’s – perhaps he is more of a viable alternative than I had previously given him credit for. That said, if this is an area we are looking to improve in, we should be looking for a player with a rate better than Eriksen’s, and thus it would absolutely make sense to upgrade on Dembélé as our back-up trequartista.
It won’t surprise you to hear that I think for our remaining matches we should give match time to youngsters. Harry Winks or Josh Onomah could play in midfield or as a number ten, and both are creative, in different ways. Winks is an excellent possession player who also has an eye for a key pass. Onomah’s creativeness comes from running with the ball and drawing players to him – a little more like Dembélé.
And for next season, we have two youngsters who could make a difference coming in from promising loans spells.
Alex Pritchard made 114 key passes in 3779 minutes in the Championship this season. He makes KP/90 at a better rate (2.7) than any of our players (again, bar Lennon and his small sample size). He also makes non-set piece KP/90 (2.1) at a better rate than any of our players. It’s the Championship, and we need to be aware that this might not translate directly to the Premier League, but it will certainly be interesting to see. He won’t help with through balls, though – he makes those passes at a rate less than all of our players that have attempted a through ball – he only played two across the whole season.
We will also have the option of using Dele Alli, who has a phenomenal goal and assist rate in League One (16 goals and 9 assists in 3399 minutes – that’s a goal or assist every 136.0 minutes from central midfield!). We have no idea how he will adjust to the Premier League, but it’ll be fascinating to see.
There’s a lot of work to do at both ends of the pitch, and my hope is that it ‘clicking’ at one end will have the result of making it click at the other too. Hopefully our new Head of Recruitment and Analysis, Paul Mitchell, will be using data like this (but better, much better!) to help advise Pochettino on new players, on existing players, and the system in general.