Tokyo Verdy – Using Data to Aid Recruitment

With promotion to the J1 League secured (check out the season summary here), now is the time to kick on and improve the squad. We did a great job last season, but the squad balance wasn’t great and there was a significant drop off in quality between the starters and the replacements in the majority of positions. I’m all for giving players that earn a promotion the chance to earn their place in a higher division, but we aren’t going to survive without some additions.

In this post, I’ll talk through the system I’ve put in place for recruitment this year, with the vast array of data available in-game aiding decision making in the absence of specific attributes. This system is broken down into three parts:

1. In-Game
2. Exporting Data
3. Analysing Data

In-Game

The in-game portion of the recruitment system is fairly standard, using recruitment focuses to gain knowledge on a large pool of players, and then using scouting and shortlists to narrow down that pool into a refined list of potential targets for a set of different scenarios.

Recruitment Focuses

Early on in the save the sole scouting focus will be in Asia. There are a couple of reasons for this. Firstly, there is a cap of 5 foreign players per match day squad. Scouting extensively in Japan is obviously a must, but as the below nations also qualify as non-foreign for work permit purposes any players we can find here to supplement the squad will be valuable additions.

In order to avoid any conflicts I’m going to limit myself to 5 foreign players in the squad in general at a time, although if an opportunity is presented to us that feels too good to pass on then I’ll make an exception with a view to moving a player on. 

The second reason is simple, it’s because it’s what we can afford. In season one we could only scout Japan, which has now been increased to East Asia as the budget was increased. Over time I expect us to scout worldwide, but for now we’ll do our best with what is available.

I will have four recruitment focuses that I will use, and will rotate the assigned scouts and analysts in order to try and develop the biggest pool of players possible. These focuses will be based on the status I expect the players found to have within the squad if signed, and therefore there will be no criteria based around the position or role they play in. 

First Team – No age restriction, minimum 3* Current/Potential Ability

In order to give us the biggest potential pool of potential first team ready additions, I’ve purposely left this focus very vague. If a young player is good enough to play in the first team then I’m willing to sign them and put them in. If an older head can come in for a year and improve us, great. I’m trying to avoid signing a player just to come in as a back-up, hence the 3* minimum. The aim is to be constantly improving, so the plan is for a new signing to be an improvement on the current incumbent, who then becomes the back-up.

Prospects – Aged 15-20, minimum 1* Current Ability / 3.5* Potential Ability

The prospects recruitment focus is still fairly vague, with the obvious addition of an age restriction. I’m happy to sign young players that are nowhere near first-team level yet with the aim of improving them alongside our intake players, but I am looking for a fairly high ceiling to ensure that I’m not signing a high number of players that will potentially be good enough for where we are now, but not where we hope to be by the time they’ve developed. 

Out of Contract – No age restriction, minimum 1* Current Ability / 3* Potential Ability, Contract Expiry < 1 year

The Out of Contract recruitment focus will find the most players that we aren’t interested in, as its been set up to cater for both the First Team and Prospect criteria. However, the extra admin of filtering these players out will be worth it to find players who can supplement the squad without spending out on transfer fees. The money doesn’t seem too bad in Japan, having been given £5 million to spend on transfers in this window, but being able to improve on the cheap will never be a bad thing. 

Loans – High priority focus a month before and during transfer windows, no age restriction,  minimum 2.5* Current/Potential ability

The loans recruitment focus is one that will be run at every transfer window, but started a month before to get an idea of who may be available. I don’t intend to rely on loans a huge amount, but if a quality player can come in and improve us for a year then I’ll definitely explore it. 

Each player will receive a cursory glance at the attribute information I have available to see if they pass the eye test (plus I’m nosey), and those with a good scout’s recommendation (ideally a B- or higher) will be added to our shortlists.

Scouted Players Screen

As the players who are scouted but don’t make the shortlists will remain listed in the Scouted Players screen, it seems a shame not to make use of it. I’ve created a custom view showing some of the more generic stats available, that aren’t related to a specific position or role.

The plan is to use the scouted players screen as a safety net of sorts; if during a transfer window I can’t find the right target in my shortlists, I’ll take a look back at my scouted players to see if anybody has slipped through the cracks. I’ll also potentially fill any gaping holes in the squad short-term with players from this list if my main targets are unobtainable or deals fall through at the last minute.

Just from taking a quick glance at the screenshot above, Marc Klok looks like a player who I’ve missed and should be shortlisted. He may be 30 but he’s Indonesian (counts as a domestic player), clearly is thought of highly by my scouts and doesn’t look to have an overly high market value or wage.

Shortlists

Rather predictably, and as shown in the flowchart, we will have four shortlists that will match our four recruitment focuses. Players can be added to more than one shortlist if they match the criteria for multiple, for example if a player found in the Prospects focus is also in the last year of their contract then they will be added to both the Prospects and Out of Contract shortlists. I’m splitting the shortlists into two groups, as how I will analyse players within them will differ.

First Team and Out of Contract – Players will be added to shortlists for one year

The First Team and Out of Contract shortlists will be the area where data will be used to aid recruitment the most, and therefore a bit of prep work has been done. I’ve created eleven custom views, plus a more generic view that I’ll use as my primary view when accessing the shortlist screen in game. Nine of the eleven views are designed to include all of the metrics I’d want to see for the profile of player I’m looking to recruit, with the other two being overviews of all attacking and defensive metrics to use as a general insight. The use of the word profile rather than position here is intentional; in some positions I may want to sign players that play in different ways, or I may be looking for a profile that could be effective in more than one position – either centrally or out wide, for instance. I won’t spam screenshots here of all eleven views, but for context they are:

• Goalkeeper
• Centre-Back
• Full-Back (Attacking)
• Full-Back (Inverted)
• Pivot Player
• Central Creator
• Secondary Goal Threat
• Wide Creator
• Centre Forward
• Attacking Output
• Defensive Output

Most of these are pretty basic, but for a couple of the more vague titles I’ll explain my rationale a little.

Pivot Players will primarily be defensive midfielders, but could also play as central midfielders. They will be the ones performing the less glamorous duties in the engine room; making tackles and interceptions, keeping things ticking over with the ball, running hard for ninety minutes. Goals and assists are a nice bonus from any player, but that’s not what I’m looking for from this profile, I just want a steady, reliable presence at the heart of midfield who’s comfortable on the ball and can stick a foot in. Right now this role in my tactic is a Ball Winning Midfielder, but in the future this could change to a Defensive Midfielder, Half Back or even an Anchor.

Secondary Goal Threats are players who could play as a number 10, out wide, or even as a Segundo Volante running from deep, but the expectation is that no matter what their role they will crash the box and take some of the goal-scoring burden off of our strikers. I want all of our players to be competent on the ball so I’m still looking for solid passing metrics, but instead of focusing on chance creation or key passes, I’m looking for out and out goal-scoring threat, so heading and shooting metrics are king here. 

Wide Creators are fairly self-explanatory; players who play out wide and look to create chances and lay on assists rather than putting the ball in the net. Obviously this will mostly be looking at wingers (I like to have a balance of a goal threat on one wing and a creator on the other), but I’d also take a look at these metrics when assessing full-backs, as my current tactic currently includes a wing-back on an attack duty. 

I’ll go into more detail about how the data from these views will be used later in the post.

Prospects – Players will be added to shortlist indefinitely but will be removed at 21 if not approached

For the most part, young prospects will be assessed in the same way as players in the Scouted Players screen – by the eye test. There are a few reasons for this. Firstly, a young player may not necessarily be a first team regular at their club, and therefore would have little or no data to analyse. Secondly, a young player may be performing one role for their club, but possess the skill-set and adaptability to be trained to perform another for mine while they’re still developing. Lastly, I’ll always take a punt on a young player who excels in at least one area. If a winger is lightning quick but below par technically, they can still offer a threat. Even if their game never develops to the level we need, they’ll still have that X-factor that may convince another side to offer us a profit.

Loans – Players added to shortlist for three months but shortlist cleared at the end of every transfer window

Loan targets will be assessed mostly on a combination of generic metrics and the eye test, as they’ll only be at the club short-term and therefore I’ll take more stock of recent performance over a longer term output. I don’t mind if they’re overperforming and not likely to sustain their level forever, as long as I think they’ll sustain it for long enough to benefit us!

Exporting Data

Once I’ve identified the key areas that need improvement (I’ll talk about this specifically shortly) the next step is to export the data relevant to those areas from the game so that I can make use of it.

This isn’t going to be a full tutorial on the process as this post is already getting fairly lengthy, (if there is interest I could definitely provide one in the future) but the aim here is to save the relevant data as a ‘web page’ file in order to import it into the spreadsheet software of your choice (Excel, Google Sheets, WPS Spreadsheet etc.).

To export data, select all of the rows currently showing by pressing CTRL+A (Command+A on Mac) and then press CTRL+P (Command+P). The below should appear, and from there save the file as a Web Page somewhere that you will be able to find it later. 

My weapon of choice is Google Sheets, where I’ve created a file with a page for each of my eleven custom views. Once I’ve imported my data and cleaned it up a little (I’ve added a couple of metrics not included in the game so need to make sure the columns line up etc) it’s time to import that spreadsheet into a program where I can create some visualisations. 

Analysing Data

I’ve created my visualisations in Looker Studio, but Tableau is also an excellent piece of software for this. Luckily, the mastermind that is Steinkelsson FM created a tutorial back on FM21 for creating visualisations in Tableau that is easy to follow and will produce great results. 

Again, this isn’t a full tutorial, my knowledge of Looker Studio is very basic and just about does the job for what I need to create. This is about Tokyo Verdy’s pre-season, and making the moves we need to stay up. 

Some early work was done in the window, securing a few players on free transfers that I felt we couldn’t pass up. With Goalkeeper Matheus leaving us on a free transfer, we had a foreign spot free and I wanted to fill it with the most talented player I could attract. Eleven players came in on trial, in a variety of positions, with the logic being that by signing the best player I could, no matter which position they played, I could build the rest of the transfer window around that. I settled on Léo Artur, a 28-year-old Brazilian who can play as a Secondary Goal Threat from the left hand side or as a Shadow Striker. 

After Léo Artur was confirmed, I had a dilemma. Season one cemented in my mind how crucial our WB-A at left-back was to our system. Both Ren Kato and Arhan played consistently well and racked up plenty of assists (7 each) when called upon. However, in terms of ability they appeared to be two of the players least likely to make the step up, and in fact I actually let Kato’s contract expire at the end of the season. The dilemma is that one of the players we brought in on trial was a very talented left-back, Héctor Quiñónes. I almost chose to sign him ahead of Léo Artur, due to my desire to have a really strong left-back, but in the end decided Léo Artur’s talent made him worth it. Now I was in a position where I could still sign him, but I would have six foreign players for five spots, something I said I wanted to avoid. However, with Héctor’s obvious quality and the complete lack of performances from Stanley Ohawuchi I made the signing and decided I’d try to offload Stanley as soon as possible. 

The last early move was to snap up Shuto Minami, a solid holding midfielder that offers depth, but crucially is a homegrown Tokyo Verdy player, having played for the club from 2009-2016. This gives us another option to fill out the bench should injuries hit, although Minami has plenty to offer to the side and should see plenty of playing time.

With the early moves made, a transfer budget of £5 million and around £35k a week in available wage budget, it was time to strategize. In terms of first-team additions, I identified three key areas where we needed extra quality.

Right-Back – Kohei Yamakoshi was solid enough as an Inverted Full-Back last season, but wasn’t overly impressive and we lack any real depth behind him, with Kazuya Miyahara finding his place in the pivot. 

Wingers – Despite already having the signing of Léo Artur secured, the wide areas are a cause for concern. Quality rotation options weren’t really a thing last season, relying on the likes of Yuji Kitajima who was plucked from the reserves and a really poor Stanley Ohawuchi to make an impact when the starters weren’t available. Kosuke Saito also played a lot of his football out wide when he is primarily a 10, which hindered the impact he could make somewhat. 

Striker – We had two really good strikers last season, but sadly only one at a time. Michel Douglas looks more than capable of making the step up to J1, but behind him depth is non-existent. Itsuki Someno, Keito Kawamura, Toyofumi Sakano and poor old Stanley all failed to show any capability of coming into the side and consistently performing, with just Stanley now left at the club (for now). 

So with these three key areas in mind, it’s finally time to take a look at the data in order to try and whittle down our shortlists to a list of top targets. With my foreign player spots already decided, I’ll be focusing exclusively on Japanese players in this transfer window, and will be looking at players who played at least 750 minutes last season to ensure that the data gives a more representative view of a player’s strengths and weaknesses.

Right-Back

I don’t need the right-back to do anything particularly complicated in the system we currently play. As the left-back is tasked with bombing on and contributing offensively, the right-back just needs to be comfortable tucking into a back three in possession, be safe on the ball and win their duels when defending out wide. 

Immediately, I’m concerned by a few of our potential targets.  The Possession Loss % metric in the bottom right chart shows how often a player loses the ball compared to winning it. A percentage over 100% indicates that the player loses possession more often than they win it, which is obviously something we are keen to avoid. This immediately rules out Nanasei Lino, Takumi Nakamura and Shohei Kawakami.

As I’m looking for a player safe in possession, and not somebody to necessarily break the lines and unlock defences, I’m also steering clear of Riku Saga. Although I appreciate his Progressive Passing rate of just over 13%, his pass accuracy being comfortably the lowest, combined with him having the highest remaining Possession Lost % leaves me with concerns that he may play us into trouble a little too often.

This leaves three players; Naoki Otani, Rei Okada and Takayuki Mae. Rei Okada offers an interesting profile – he makes by far the most progressive passes and retains a high level of accuracy, he is the safest in terms of Possession Lost %, and attempts and wins the most pressures. However, his tackling success leaves a lot to be desired, and from taking a look at his profile in-game, it seems he is more suited to playing a traditional full-back role where his attacking intent can be utilised. 

Naoki Otani is interesting because in some aspects he looks to be poor compared to others – most notably his considerably lower number of tackles and pressures per 90. However, I’m willing to take this with a pinch of salt as Otani played his football for Tochigi last season exclusively as a centre-back and so I wouldn’t necessarily expect him to be completing as many actions as the others. A look at his attribute ranges shows me he’s a similar profile to Yamakoshi who he would be direct competition for, and therefore I think he’d be capable of performing the role I need him to, while already being able to see that he’s safe on the ball without shying away from having it.

Takayuki Mae is the last player who looks capable, with impressive pass accuracy and progressive pass %. He loses the ball slightly more often than Otani, but this can be accounted for as he has a more progressive passing style which I’m not opposed to. He also completes 2.4 pressures per 90 at a rate of 27% which gives him a nice compromise of being fairly busy but also effective. Another big plus for Mae is his versatility, as he could feasibly feature for us at RB, CB, DM or even as a 10 if really necessary. 

Based on what I can see here, Otani is the key target. I have first hand experience of a CB playing out as the Inverted Full-Back and like the added security of a natural CB creating the back three in possession. Combined with the fact that the data shows he’s good in his pressures and safe on the ball, and the eye test showing he looks to be a more capable player than Kohei Yamakoshi and should offer a strong upgrade if we can get him. I’d also be interested in picking up Takayuki Mae if I can, I see him as a player who would probably provide a rotation option, but his versatility could be key from the bench and in games where we may dominate the ball his progressive nature may be a secret weapon. 

ADDED TO KEY TARGETS LIST – Naoki Otani and Takayuki Mae ✅

Wingers

As I mentioned above, the lack of depth out wide is a major concern. I’m looking for at least two wide players to come in, providing they’re the right profiles. With the addition of Léo Artur there will be a tweak of the role on the left, with a shift to an Inside Forward. With this in mind, I’ll be analysing my potential targets as both a Wide Creator and Secondary Goal Threat, with the hope that I can find a suitable player of each profile to offer depth and variance to the attack.

Wide Creator

Well, this one is fairly straightforward. Ten Miyagi is the clear stand out here, with the graphs showing his ability to take on his man and get a cross in, his impressive numbers for not only chances created per 90 but also the quality of his chances (chance quality % is calculated by dividing total xA by Open Play Key Passes, a metric I’ve shamelessly taken from Mustermann FM and this brilliant video), and also his relative security in possession.

Secondary Goal Threat

With Léo Artur already a confirmed signing and looking capable of both scoring and creating, and Ten Miyagi identified as the primary target to come in as a creative wide player, it feels like one extra player who can pop up with the odd goal would round out the squad and give us some variation when we rotate. 

Although there aren’t any mind-blowing metrics here, I like the look of Koki Sugimori. With 5 goals in 13 starts (1475 minutes total) he’s clearly capable of making some sort of impact, but it’s by taking a slightly deeper look that he starts to become a more viable option. 

On the face of it his metrics aren’t that impressive, with both Ten Miyagi and Tatsuya Tanaka shooting and dribbling more than him per 90. Sugimori matched his xG last season with a performance of 0.31 goals per 90, but has the lowest conversion rate of the 4 shortlisted players. However, a look at his xG per shot of 0.13 shows that he isn’t being presented with high quality opportunities by his teammates. With this in mind, his high shot on target % and shots on target per 90 look more impressive, and lead me to believe that if we could provide him with higher quality chances then we could improve his output, even with a jump in division.

Ironically, this same logic also applies to Ten Migayi, and in a more impressive fashion. Ten also has an xG per shot of 0.13, but more than doubles Koki’s conversion rate and outperformed his non-penalty xG by 0.2 per 90. 

ADDED TO KEY TARGETS LIST – Ten Miyagi and Koki Sugimori ✅

Striker

As I mentioned earlier in the post, a reliable back-up for our primary forward was nowhere to be seen last year. There was no overlap between Mario Engels and Michel Douglas, and the rotation options were very scattergun and ineffective. This year, I want two dependable strikers who can rotate and bring the best out of each other. I don’t use my strikers a huge amount in build up, almost always preferring an Advanced Forward, and so good goal-scoring metrics are what I’m looking for.

Hiroto Yamada is objectively the most effective player shortlisted. He isn’t the busiest of strikers, averaging less than a shot on target per 90, but boasts an impressive conversation rate of 26%, gets more than half of his shots on target and outperforms his non-penalty xG by a decent amount. Mizuki Ando offers a similar profile, but as he is generally presented with higher quality chances (0.24 xG per shot) but gets less of his shots on target and has a lower conversion rate, Yamada would be the clear first option. 

ADDED TO KEY TARGETS LIST – Hiroto Yamada and Mizuki Ando ✅

So there we are, with the data exported and charts analysed, we are armed with a shopping list to try and keep Tokyo Verdy in J1, a league I feel they should definitely be a mainstay in.

Next up will be a review of season two, where we’ll see how many of these targets we managed to bring in, if any, and how we get on back in the big time. From there, it’ll be a look at the training I’ve implemented, based loosely on the idea of Tactical Periodisation. If you’d like to see a more detailed walkthrough of the process to export data and create visualisations in Looker Studio then let me know in the comments! Thanks for reading.