July 18, 2015

Football Intelligence – the game’s direction of travel

Some people are turned off by statistics in football. Frequently I get responses when I tweet basic stats saying ‘I use my eyes’ – or words along those lines – disregarding any form of data or numbers. That’s fine – we all appreciate football in different ways.

Well, I unashamedly like data. I don’t ‘do’ stats  particularly well; I don’t have a formal background in statistics and I’ve never been involved in that type of scholarly research, and so I leave it to the experts and mostly keep it simple, focusing on ‘events’ – tackles, passes, assists, shots, goals. But there’s a growing community dedicated to ‘advanced’ statistic and, at major football clubs, teams of statisticians are starting to use ‘football intelligence’ data.

I had been recommended this presentation by various people over the last couple of months, and today I finally got around to watching it. It is Damien Comolli – formerly Sporting Director at Spurs (2005 – 2008) – presenting a Performance.LAB Innovation Seminar about Squad Management.

I don’t want to spoil it, as it’s a must watch. But it’s fairly Spurs-centric, and Nabil Bentaleb is used as an example throughout. In addition, if this doesn’t make you want us to sign Christoph Kramer, nothing will.

At the same event, Paul Power gave a presentation about Game Intelligence. He states towards the beginning of the presentation that they used 260 million points of data to prepare for that presentation alone. That is remarkable in itself. But look at how they’re using it, and it will blow your mind.

It amused me somewhat that Comolli talks about how Luka Modrić – a player who doesn’t stand out when you look at event data alone – helped drive them to work on some of these models. To essentially prove, statistically, that he is the brilliant player that we know he is from simply using our eyes.

Where this will lead next is fascinating. This sort of data is already being used to recruit players, and to analyse existing players’ performance. Indeed at Spurs we have Paul Mitchell, recruited at the back end of 2014 from Southampton as Head of Recruitment and Analysis. He who is famous for his ‘black box’ and use of data. He uses data and visuals together, presumably in ways similar to what is seen in Comolli’s seminar.

The next step will be to drive coaching forward – to apply the numbers to coaching methods and to help players improve based on very specific data. To prepare players and teams for specific opponents by targeting their weaknesses and stopping them from using their strengths.

Of course you can’t just have the data. You need someone to analyse it, you need someone to present it, and you need to make judgements and have conversations – to remember that different coaches want different players to do different things. It’d be no good to give Mauricio Pochettino a player who is brilliant at controlling possession and tempo by taking lots of touches because he actually wants his players to move the ball quickly.

And so conversations need to be had, logic needs to be applied, and also personalities need to be analysed. But it’s fascinating nonetheless.

I expect some form of backlash in terms of how players will adapt to being analysed. For example, players all get their numbers back immediately after a match, and there have been stories about players playing lots of safe, square passes between one another late in games to drive their pass completion up. Advanced analytics and football intelligence will know this, and these sorts of safe passes will not allow the player to stand out – because only passes that matter will add value.

You can, of course, still continue watching football without digging any deeper – the beauty of the game need not change. But rest assured that beneath the surface is a team aiming to ensure that clubs get every inch of value from players in an ever-changing environment.

Tagged ,

July 18, 2015

25-man squad update – July 2015

Prior to the past few seasons I have written about how our 25-man squad is shaping up in line with the ‘home grown’ rule. I have been asked to do the same again this year.

To summarise the rule again, we are able to name a 25-man squad if eight of the players are ‘home grown’. We could name fewer than eight home grown players, but would need to also name fewer than 25 players in our squad – for example, if we only have seven home grown players, we can name a 24-man squad, 6/23, 5/22, etc. A home grown player is defined as follows:

… one who, irrespective of his nationality or age, has been registered with any club affiliated to the Football Association or the Welsh Football Association for a period, continuous or not, of three entire seasons or 36 months prior to his 21st birthday (or the end of the season during which he turns 21).

We do not need to name players who are under 21 on our squad list; for the 2015/16 campaign, players considered ‘under 21′ will have been born on or after 1st January 1994.

Since the beginning of last season we have lost one one ‘home grown’ player (Kyle Naughton) from our squad list, and can expect to lose at least one more (Aaron Lennon). We have added Kieran Trippier, who is home grown, but also two non-home grown players in Toby Alderweireld and Kevin Wimmer. Dele Alli falls into the under-21 category.

Also, since last season, Harry Kane has passed the age threshold and will need to be named in the squad, whereas last year he was simply included in our list of under-21 players.

As it stands, our ‘named’ 25-man squad would probably consist of the following (* = home grown player):

Hugo Lloris
Michel Vorm

Kyle Walker*
Danny Rose*
Kieran Trippier*
Ben Davies*
DeAndre Yedlin

Jan Vertonghen
Toby Alderweireld
Kevin Wimmer
Federico Fazio
Vlad Chiriches

Ryan Mason*
Mousa Dembélé
Benjamin Stambouli
Tom Carroll*

Christian Eriksen
Nacer Chadli
Erik Lamela
Andros Townsend*
Alex Pritchard *
Aaron Lennon*

Harry Kane*
Emmanuel Adebayor
Roberto Soldado

That would mean that the following would miss out:

Grant Hall*
Ryan Fredericks*

We are then able to select any players who were born after January 1994 without needing to register them. This means that any of the following (plus the other first and second year academy scholars) would be available for selection. NB: I have presented them in age order.

Eric Dier
Shaq Coulthirst
Kenny McEvoy
Nabil Bentaleb
Grant Ward (on loan at Rotherham)
Dominic Ball
Luke McGee
Milos Veljkovic
Harry Winks
Connor Ogilvie
Nathan Oduwa
Emmanuel Sonupe
Dele Alli
William Miller
Joe Pritchard
Harry Voss
Anton Walkes
Luke Amos
Anthony Georgiou
Cy Goddard
Kyle Walker-Peters
Joshua Onomah
Shayon Harrison
Cameron Carter-Vickers
Ismail Azzaoui

Essentially what this means is that if we want to add another player born after January 1994 to the squad, we have to remove another to make space. This should be no problem, as I would expect Soldado, Adebayor, Lennon, Stambouli, Chiriches, and possibly Dembélé/Fazio (depending on offers) to move on, opening up at least five spaces for new signings should we wish to make them.

Our ‘home grown’ quota is looking very healthy, with ten in the 25-man squad as it stands.

With so much young talent coming through, we look pretty comfortable in terms of home grown numbers for the coming years, although it’s worth noting that Eric Dier will not be able to be named as a home grown player when he passes the age threshold next season, as he does not meet the criteria due to his football upbringing in Portugal. I am fairly certain, though, that Nabil Bentaleb *will* be counted as home grown (depending on when he was formally registered).

Tagged ,

July 4, 2015

The Danger of Expectation

Let’s make one thing clear: what happened to Harry Kane last season was not just a ‘once in a lifetime’ experience for the player – it was a once in a lifetime experience for Tottenham Hotspur Football Club. It was perfection. Can it ever get better than that for a player in his first full season in the top flight? It’s so, so unlikely.

In the past few weeks I have received countless tweets asking who the next young players to break through are – and I respond by reeling off the same names that I’ve been talking up for months. There are articles written each week which list the most likely academy players to ‘follow in Kane’s footsteps’. But it struck me recently that I should probably stop reeling off these names and adding to the building hype – because how can it be helpful?

Essentially we are setting these players up to fail – how can they reach the standards that we are unconsciously creating for them? What would they have to do to impress: perform at a level over and above our already high expectations?

The truth is that part of the reason that Kane’s terrific season was so enjoyable was that – for many – it was utterly unexpected. Written off as a lumbering, awkward target man (mostly owing to the way he was used in early appearances, and an unfulfilling loan spell at Norwich City), he fell somewhere between a cult hero and figure of fun after launching a ball forward to waste time and spitting on himself when trotting back into his own half having done so. What happened after that was a rare and beautiful thing that led to a lot of words being eaten.

Nabil Bentaleb was chastised in his breakthrough season – many fans questioned ‘what he did’ before it became clear to all that what he did was, actually, rather remarkable for a player of that age and in that position. Ryan Mason has received widespread criticism for his defensive play despite last season being his first in the top flight. Inexperienced players make mistakes.

We have the most talented group of academy players that I have seen at Spurs, and there are many that *could* make the step up to Premier League player status. But there’s so much that can go wrong on that journey. There’s so much that can happen between now and ‘full England international’. Just getting to Jake Livermore level – a solid Premier League player (well, pre-incident) – is absolutely not to be sniffed at. Many, many academy players fall by the wayside and end up playing non-league football whilst finding a job outside the sport.

What would help is for fans to lower expectations, and just enjoy the glimpses of youth that we will hopefully get next season. Let’s bring them into an environment where they’re allowed to make mistakes without moans and groans inside the ground, and over-analysis on social media. Let’s not be so keen to be the person that called it right first that we make a firm decision on players after just a few appearances.

Let’s let our young players make mistakes and learn from them. Let’s support them through that process and accept that it’s the norm. Alas, they won’t all be Harry Kane.

Tagged ,

May 31, 2015

Wimmer – a Q&A with an FC Köln supporter

Like many other Spurs fans, I know very little about our new signing, Kevin Wimmer. To fill us in, my good friend Paul King spoke to UK 1.FC Köln.

Firstly, thank you for offering to share your thoughts and knowledge on Kevin Wimmer. What has been the general feeling from FC Köln fans during his time as a player at your club?

He came into the club as an unknown quantity from LASK Linz in Austria and struggled to become truly established initially. It wasn’t until new coach Peter Stoeger (also Austrian) arrived at the club in the Summer of 2013 that Wimmer made a true impact and since then he has been like a rock in the back four alongside Slovenian Dominic Maroh. He was a huge part of us gaining promotion in 2014 and consolidating our Bundesliga place last season. The club had the fifth best defensive record in the division which, for a promoted club, was excellent.

He is best known as a central defender, but has he played in other positions for FC Köln?

If I’m honest I can’t recall Kevin being played ‘out’ of position in the last couple of seasons. I would view him as a pure defender and as such would like to see him concentrate on that role. I do believe, however, that he played as a midfielder earlier in his career at youth level.

What are his main strengths?

He is well-built to be a central defender and his solid, no-nonsense approach makes him a pure player in this role. You won’t see Wimmer messing about with the football at the back. He also has decent pace which we all know for a player in the Premier League will be vitally important.

Does he have any weaknesses which he still needs to develop?

For me he isn’t the best footballer technically but, as I have said previously, he is a no-nonsense defender. Don’t get me wrong, I wouldn’t expect him to just ‘hoof’ the ball out at every opportunity, but you will see few risks taken.

How does he compare to better known central defenders in the Bundesliga, such as Jerome Boateng of Bayern Munich, Mats Hummels of Borussia Dortmund and Naldo of Wolfsburg? Do you feel that he has the potential to get to that sort of level?

I would say that of the names mentioned, Hummels would make the best comparison with Wimmer. I am confident that he will become a great player in years to come; I just hope he gets off to a good start in the Premier League as, from experience, it can be an unforgiving place to play football.

Are there other central defenders who Tottenham Hotspur fans may be familiar with who compare to Kevin Wimmer?

From established defenders it is very difficult to make comparisons at the moment. A similair-styled up and coming player is Kurt Zouma at Chelsea.

It is often said that it takes time for overseas players to adapt to the Premier League. From your knowledge of this league, do you think he is ready?

Yes, he is ready for action without doubt, but the Premier League by experience can be like a shark tank and the fans can be quite unforgiving early in a player’s career. That is my main concern for Wimmer moving to England, although I do believe he will perform very well.

Are FC Köln fans surprised that he has joined a club such as Tottenham? Would you feel he could have gone to a ‘bigger’ club, or is the level of Tottenham Hotspur well suited?

If he had moved to another club in Germany it would have been one of the top six clubs that would have shown an interest. I think Europa League level is about right. The majority of fans are pleased to see him progress and it shows the good work that has gone in at Cologne in the past few years that our players are attracting attention from clubs in the CL and EL.

Thanks very much to @effzehUK for the insight, and Paul for conducting this interview.

Tagged ,

May 14, 2015

Toothless Tottenham

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.

Key passes per 90 minutes - top five players (who have played more than 500 minutes) for each club in the Premier League's current top seven								Key passes per 90 minutes - top five players (who have played more than 500 minutes) for each club in the Premier League's current top seven

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.

Tagged ,

RSS Feed Google+s The Fighting Cock!
 
 

Twitter

Enjoy the match tonight, folks, especially those at the ground. Going to try to watch first thing without knowing the score... fat chance!
2 hours ago