FM Data Analytics: use data visualizations to review your tactics

[gam945]

Right after the winter update, I launched my save with my beloved Calcio Padova in Serie C, the 3rd tier of Italian football. The club is projected to be 2nd in the league, and our objective is the playoffs, meaning we should be at least 10th. My goal with this save is to take the club to the upper levels of Italian and European football, but also to create an identity of vertical possession, high tempo and high pressing playing style - think of Ajax in 18/19 or Bayern 11/13. In possession, I want to achieve a 3-3-1-3 while defending in 4-4-2/4-4-1-1, so here is my tactics and how I want it to work. I am no tactical genius, but as any manager, I'll try to adjust it by seeing how it works out.

Tactic

This is what I'm trying to achieve: a 4-4-2/4-4-1-1 that morphs into a 3-3-1-3 in possession.

My interpretation is as follows (subsitutes in black under the starter):

 

I will not take too much time explaining it as it is not really the goal of this thread, but it will quickly break down what I expect from this tactic.

CD-de BPD-co FB-su form my back 3. On the left, the central defender is instructed to go wider and although he participates in playing from the back his main task is to defend. He only needs to play simple passes to the DLP and IWB on his side. The BPD acts like a sweeper, sometimes goes forward and participates a lot in playing from the back. Think of Barcelona Dream Team's Koeman. The FB has PIs so he acts the most conservative possible (Hold position, Sit Narrow, etc) but helps in the building play phase and supports the attacks. Think of a very conservative Lahm. 

IWB-su DLP-de CM-su form the second layer of 3 in the 3-3-1-3, so basically the tip and sides of the diamonds. Their roles are similar to corresponding players in a 3-3-1-3. The DLP is the key player for build up, and the IWB and CM are hard working creative players overall. Think of Ajax's De Jong for the DLP, and Kimmich for the IWB and CM.

AM-at is the tip of the diamond. His role is not much of a playmaker, but more of a "space investigator"  . I expect him to help in the late stages of build up, but he should concentrate on sneaking in the opposite team's penalty box and grab a lot of goals and some assists. He should be perfect at exploiting spaces in the third tier of the pitch. Think of Ajax's Van de Beek or Thomas Muller.

IW-su F9-su W-su form the last layer of 3 in the 3-3-1-3 and are my attacking threats. IW and W focus mainly on getting into dangerous spots from the wing and provide assists and score some goals. They should profit from 1 on 1 situations a lot, think of Robben & Ribery. Lastly, F9 is the star of the team, he acts as the real advanced playmaker and provides assists but he should also get a lot of goals. Think of no less than 10/11 Messi or Cruyff.

We are now in early November in my first season and I need to review my team's performance before planning my pre-contract shopping that will start on January 1st. I will use my app, FM Data Analytics, to analyze the performance compared to what I expect.

First some data from FM21:

A lot of draws, still unbeaten but an evident issue with scoring more than 1 goals in a match...

Quickly, I can see that we are good in possession, have poor finishing but lots of chances, a busy and quite leaky defense. So the only area I should be satisfied is midfield and build up play in general. Let's see if a more thorough analysis will confirm that.

General

20 goals in 19 matches, that is indeed poor. First thing I notice is that Halfredsson (AM-at) is not contributing as much as he should. Otherwise, goals are coming from my wingers and my front AM/F9, and assists are spread between my midfielders mostly. A lot of room for amelioration, and I shall keep an eye on shooting in the next part of my analysis, as I can see that we are quite bad in this area. Also, we had 6 penalties, spread between Beretta/Santini/Paponi, which lowers even much their real contribution.

I will analyze 2 areas: building up (passing) as it is critical for my playing style and finishing (shooting) as I notice that we are scoring a lot less than we should.

First, I will set my DNA Model. This will color each player's point in the visualization (bad to good goes like this: red-orange-yellow-light green-dark green) and I will be able to see how players with the attributes I need are performing. Keep in mind that we're in the Italian 3rd tier, so an average of 60 (/100, equivalent to 12/20 in FM) for key attributes is pretty good.

Here is a filter page that also acts like a quick summary. Not a lot of players with a good DNA Rating...

Let's dive into the real analysis!

Build up and creativity (Passing)

1 - Attempted vs Completed Passes

Most of my players have a more or less 85% passing completion rate as I can see, and the two players leading the pack are actually my two DLPs, so makes sense it confirms our possession/passing stats. Let's see who actually makes useful passes...

2 - Completed vs key Passes

So...two players are noticeably performing well: Curcio (starter - IWB) and Ronaldo (sub - CM). If I compare Ronaldo to the starting player in his position, Germano, I notice that they have more or less the same number of completed passes, but the sub has a more than double his key passes number. There was a fixture congestion from mid-September to mid-November with 2 matches per week each week, so I had to rotate a lot. Ronaldo got 9 games and Germano 10, so here their performance can be compared and I will be looking to give the starting position to Ronaldo.

My IW, Montiel, is also performing better than his competitors, and I notice that there is a cluster in the bottom left (#1) that consists of my wingers and AM/F9s. We seem to rely a lot on my midfielders for key passes, while it should be my wingers and F9 that should perform on this aspect. The cluster in the bottom center consists of my defenders, so it seems okay. In the bottom right, with a lot of passes completed but low key passes we find my DLPs, pretty normal again as they focus on early stages of build up.

Let's analyze created chances to see if I draw the same conclusions.

3 - Completed Passes vs Chances created

Again I notice the same thing. My starting IW Montiel is still overperforming his competitors in the wing; this time my starting W, Bifulco, also creates a lot of chances; I notice a cluster (#1) of poor wingers and AM/F9s in the bottom left; a cluster (#2) of defenders in bottom center and my DLPs have low created chances.

Also again, I notice that Ronaldo creates a lot more chances (about 4x more!) than Germano, the starting player in his position! Notice that Germano is even hidden in the cluster of defenders! From the three previous visualizations, I can already draw some conclusions:

• Performance-wise, Ronaldo (sub) is more creative than Germano (starter) and I need a creative player in this position. It seems I should give Ronaldo a chance as the starter in CM-su, and monitor his performances and his impact on results.
• Montiel and Bifulco are the only players from attacking positions that create chances. This is alarming as I said, my F9 should be the one who creates and assists the most. The subs for my wingers are also quite bad, and something has to be done tactic-wise to resolve this issue, as I do not have subs that can take the starters' places.
• My DLPs and my defense are doing their job, nothing alarming in this area.

Crossing-wise, my wingers are doing their job:

I won't dive into analyzing assists, as (1) as we saw in the summary we do not have a lot of goals nor assists so the data is less significant and (2) assists figures seem to be crippled by poor finishing. So let's analyze our finishing, as this seems the most urgent issue...

Finishing (Shooting)

1 - Attempted vs On Target Shots

So my attacking players are having the most shots per game, which seems normal, but it also seems we have an aiming issue...I cannot draw a lot of conclusions from that, but I can already imagine what's going on here: bad aiming + low quality shots = very poor finishing stats...

2 - On Target Shots vs Goals

Notice that only players with actual goals are represented here. I don't see anything abnormal...it seems that my attacking players have a normal conversion rate from on target shots. However Paponi (sub-F9) is a distaster: he shoots a lot and converts lower than anyone else. Actually at the start of the season he was my starter in F9, but it was so evident that he was performing badly that I gave Santini the starting position. He still got a lot of games because of the tight schedule over 2 months.

Basically until now, it seems that our aiming is really bad, but our conversion from on target shots is okay. Let's see the quality of our shots...

4 - Attempted Shots vs xG

Again, like the first shooting visualization, my attacking players (wingers/AMs/F9s) are having the most shots. However here the data is skewed, I know it for a fact. We have 6 penalties in the league, 3 more than the second team with most penalties. Santini/Beretta/Paponi took them, let's say 2 each...this lowers their xG per 90min a lot! But still, Montiel (starter - IW) and Beretta (sub - AM) seem to have a good xG. This can't be said for the others...

5 - xG vs Goals

Comparing the xG with goals show that if we remove penalties, it seems that we have poor finishing AND poor xG...

So what conclusions I can draw from this?

• My players aim very badly: poor On target/Attempted shots ratio
• High number of penalties (6 in 17 league games) skews the data and if we do not take it into account, it seems we have a very poor attack in all of it aspects except conversion from shots on targets
• The solution has to be tactical, but the players are either not good enough (which is hard to back up with their attributes for a Serie C team) or do not fit into my tactic (which is a lot more plausible as I play many of them out of their position). In either case I will need players that adhere to my playing style and fit into my tactic's roles.

I will finish my analysis here, but I've seen the defense statistics and they seem okay. From match observations, I noticed that we were pretty good in build up but bad in finishing and I wanted to confirm that. We are still unbeaten but we have drawn a LOT of games, so there is room for improvement.

I will now summarize how I will act on my analysis:

• Ronaldo will be given a chance as the starter in the CM role instead of Germano
• F9 and AM-at his clearly unsuited for my players, I will look into a front two combination that will use more their strengths without compromising too much my 3-3-1-3 shape in possession (but that may be inevitable and I may obtain a 3-3-4 which will be more direct as I will lack the link between midfield and attack)
• Will concentrate my transfer activity on a AM & F9 and in a lesser extent on wingers.
• My midfield and defense will mostly get 1 year renewal for the 30+ and 2 year for the 25+ as most of them have contracts finishing in June 2021 and winter transfer period is approaching. This will give me more time to concentrate on shopping for attacking players.

Clearly, we are still in early stages of my project. However, I want to be in Serie A in 5 years. My save is scum save free, and I try to play as realistically as possible. Padova is already a contender for the Serie C title in the first FM21 season, so it is realistic to aim the Serie A in 5 years max. Also, I am aware that I may not fully develop my playing style in Serie C and but 5 years is enough for me to build a team that will satisfy me. 

Feel free to use the app, which is described thoroughly in my thread from the General discussion forum (see the start of my post), and discuss your analysis

[ericstpeter]

Great post. I know it’s not the point but I’d love to give this system a shot. It looks fun. Would you mind giving us all the PI’s you use? 

[gam945]

4 hours ago, ericstpeter said:

Great post. I know it’s not the point but I’d love to give this system a shot. It looks fun. Would you mind giving us all the PI’s you use? 

F9 & AM: Roam
IW & W: Get Forward
CM: Go wider/Roam
IWB: Shoot less often
CD: Take fewer risks/Go wider
BPD: Ease off tackles
FB: Take fewer risks/Sit narrower/Cross less often/Cross from deep/Hold Position

However I'm not getting as good results as I'd wish...as I said, a lot of draws, and a big issue on scoring goals. From what I see in matches though, build up is excellent.

[ericstpeter]

46 minutes ago, gam945 said:

F9 & AM: Roam
IW & W: Get Forward
CM: Go wider/Roam
IWB: Shoot less often
CD: Take fewer risks/Go wider
BPD: Ease off tackles
FB: Take fewer risks/Sit narrower/Cross less often/Cross from deep/Hold Position

However I'm not getting as good results as I'd wish...as I said, a lot of draws, and a big issue on scoring goals. From what I see in matches though, build up is excellent.

Thanks. I’ve seen a lot of people try to replicate that 3-3-1-3 shape in attack and it’s never quite there. This one seems to be on point.

[gam945]

33 minutes ago, ericstpeter said:

Thanks. I’ve seen a lot of people try to replicate that 3-3-1-3 shape in attack and it’s never quite there. This one seems to be on point.

Screenshot of a build up phase...this is the same 4-4-1-1 I've posted. It's a little skewed to the right side, but I have to accept this if I use a back four to get a back 3... also when the ball goes in to the left side, the whole block shifts to the left so there's no space left wide open.

 

[ericstpeter]

6 minutes ago, gam945 said:

Screenshot of a build up phase...this is the same 4-4-1-1 I've posted. It's a little skewed to the right side, but I have to accept this if I use a back four to get a back 3... also when the ball goes in to the left side, the whole block shifts to the left so there's no space left wide open.

 

Yeah. I like it. Playing with it now as Arsenal in 2029. Defensive great so far but low scoring like you said. Tried the AM-A as a SS-A?

I’d also be curious about lining the AMC and ST off vertically when you try the SS-A instead of offsetting them. I think it would add a little more unpredictability to the attack. Maybe the confusion leads to more goals

Edited February 27, 2021 by ericstpeter

[gam945]

On 27/02/2021 at 18:33, ericstpeter said:

Yeah. I like it. Playing with it now as Arsenal in 2029. Defensive great so far but low scoring like you said. Tried the AM-A as a SS-A?

I’d also be curious about lining the AMC and ST off vertically when you try the SS-A instead of offsetting them. I think it would add a little more unpredictability to the attack. Maybe the confusion leads to more goals

I wanted to give each of the AMC and ST a clear space to operate in, thus the offset, but have the Roam From Position PI so it's worth trying to use a true 4-4-1-1.

I'm now using this tactic and it's pretty close to the 3-3-1-3 in attack as the F9 comes deep quite often. F9 and the Poacher have the Roam from position PI, so despite being offset to one side, they actually aren't fixed to one side of the pitch. I also removed the Stay Wider PI from the right center midfield, which is now a BBM, and the wingers have an attack duty. Finally, IWB also has the Get Further Forward PI. I'll post some screenshots later but it's producing more goals than the previous one.

Edited February 28, 2021 by gam945