nestorFM

AboutFM18FM19FM20

nestor on 28 May, 2020

Goal Chains – postprocessing an amazing book

Starting the save at Everton got me excited. It was something completely new, as in term of FM I haven't played in Southamerica on a proper save, and actually never managed in Chile. I was always managing in big countries/leagues like Brazil or Argentina. Chile was something different, as it's a much smaller league. My initial idea was to fill up my squad with young but talented chileans as we hadn't that much money, and there's actually some rigid registration rules towards young players in chile.

Every club is allowed to have a total of seven foreign players, but with the following limitations:

  • 5 players who are 21y+ old
  • 2 players with an max age of 21

With Chile being a small footballing nation, I thought bringing in the best foreigner would allow us to give us a huge jump towards improving our league positions. When I took the job, they where a bit above the relegation battle in the first season, managing to finish directly on a Copa Sudamericana place (7th to be specifically). And a 2nd of place in the following season.

tactic

I think we've achieved this mainly due too, tweaking my 3-5-2 tactic I was already using at Santander and replacing most of the foreign players. Following Players have been started the save:

# Player Screenshot Replaced by Screenshot
1 Mauricio Álvez alvez - -
2 Alexis Gazzaniga gazzaniga - -
3 Jonathan Serrrizuela serrizuela -
4 Martin martin Gerardo Durán duran
5 Marcelo Caraballo carabello Raúl Riera riera
6 - - Pedro Penayo penayo
7 - - Federico Cobián federico

After Reading “Football Hackers – The Science and Art of a Data Revolution” by Christoph Biermann, (which I found a really amazing book, recommended by Lee/fmanalysis), I was asking myself: “Can we translate such analysis to the world of Football Manager?

A first try

During the midseason break in my 2nd season, I have snapped up a quite amazing Mexican wonderkid, Federico Cobián (who is linked above), and playing him as PF(A) improved our performances so drastically, that we finished up runners-up behind league favorites Club Universidad de Chile. Subjective, I felt that he was the main reason we played so much better. So, I wanted to see if I could quantify Federico’s impact with some calculations. One of stats that could be easily adapted into FM, was the expected Goal Chain (xGC) value, which is described in the following way:

«You can go one step beyond Expected Assists and have a look at the players involved in the passing chain leading up to a shot at goal. It sounds complicated but isn’t really: whoever was involved in an unbroken spell of possession is assigned the resulting shot’s expected goal ration”

Without having the actual expected Goal (xG) value, I decided to use the following Events during a football game my save: Goals + Clear cut chances + Half Chances (and give them all a value of 1).

After around six months at Everton, as mentioned above, surprisingly we managed to qualify to the Copa Sudamericana in South America, the 2nd best continal competition. We played only two as we went out in the first round we played in after a horrendous away performance and conceding four goals in the last 30miuntes. The games vs Independiente Medellín finished 1-4 (A) and 1-0 (H).

So, I decided to start on the analysis, as described above to calculate each touch per player of any of the chances created in the games:

sudamericana

My first conclusion on these were so far:

  • In my opionion, the "playing out the defence" TI was mirrored in the GK and all Central Defenders having contributed on 1 - 2 chances per game
  • Víctor González, Bastián Santis and Fernando Aguilera’s high stats were no surprise. The last two lads where playing up top as PF(A) respectively DLF(A) and Víctor, was one of the key players playing as IW(A) on the MR position, on my first months. So, nothing unexpected there.
  • Carlos Torres, playing as winger on ML, was also having some good numbers with taking part in 2 chances during that game he played in. But according to the match stats with the give key passed, he was doing much better, but just barely taking part in the scoring opportunities (or the other players wasted them), which I would think as a left winger:

medellin

  • Same goes for Jonathan Serrizuela, one of the Argentinean players at Everton, who had 4 key passes in that game, without any of it creating a chance.

As a next step a decided to give the 30 games a go we played in Chilean First Division:

The Defense

The impact of the Goalkeeper and the Defender on the chances created during the season was a bit lower, around 1 chance per game that everyone of the three defenders and the Goalkeeper participated. Maybe I would have expected a bit higher values here:

Games Played Part of # Chances xGC/90 Position Player
29 14 0.83 Goalkeeper Riera
25 16 1.08 Central Defender Alvez
30 19 1.03 Central Defender Cisternas
2 1 1.00 Central Defender Mirko Gómez
29 14 0.83 Central Defender Gonzalez
12 4 0.75 Central Defender Figueroa
5 2 0.60 Central Defender Henriquez
5 2 0.40 Central Defender Ortega

The Engine

The starting midfielders (Flores, Medina and Penayo) all had a bit less than 2 chances per game created:

Games Played Part of # Chances xGC/90 Position Player
13 16 2.08 Central Midfielder Gazzaniga
29 41 1.90 Central Midfielder Flores
29 30 1.72 Central Midfielder Medina
29 30 1.45 Central Midfielder Penayo
7 2 0.43 Central Midfielder Duran
16 3 0.38 Central Midfielder Serrizuela
8 3 0.38 Central Midfielder Baeza

Gazzanniga up top with a the best ratio, creating a lot of chances, but mainly after corners/free kicks. Baeza and Serrizuela have not been creating much, but they were on my unwanted List anyhow. Maybe the mexican CM Duran would have been sold in the future as well....

The Wideplayers

The two main wingers (Victor Gonzalez and Gabriel Torres) participed in around 2 chances per game. To be honest I would have expected a bit more here, as I felt both players were really crucial and would take part and much more important role.

Games Played Part of # Chances xGC/90 Position Player
21 22 1.95 Wide Midfielder Gonzalez
25 27 1.88 Wide Midfielder Torres
11 12 1.64 Wide Midfielder Parada
1 1 1.00 Wide Midfielder Salazar

Parada, a 20y old chilean winger, stepped up really good when Victor Gonzalez got injured for a few weeks, and have almost the same productivity as the key player had. Somehow there wasn't really an impact that Parada was playing as W(S) on the right and Gonzalez as IW(A).

The strike force

All three main striker have been having a really good impact on the game and the result. Santis with a bit better numbers than Aguilera, but he was also shooting most of the penalties this season. So I'd say they were about the same.

Games Played Part of # Chances xGC/90 Position Player
14 30 2.86 Striker Cobian
25 33 2.08 Striker Santis
30 39 2.03 Striker Aguilera
6 2 1.17 Striker Millar
2 1 0.50 Striker Munoz

As expected did Federico Cobián have really great stats, just a bit shy of 3 chances per game.

Next Step

At the end of the season Federico left again for Chivas, as they wanted “give the player a chance on the first team”. Which he barely got, like always in FM. A few month later Chivas was lying on the 18th place, which was quite a surprise as they were one of the league contenders (and are a much better club than, Murciélagos FC or Dorados de Sinaloa). So, a surprising offer popped up in my inbox:

chivas

So, I decided to follow my former star striker. As I have somehow lost the interest in managing in chile after my 2nd season (I already got an offer from Indenpendiente in Argentina, while sitting on the first which would have be some fun, but they opted to go for Mauro Icardi as manager instead, “you’re just not the right guy”).

¡No manches, wuey!

Tweet this article:
developed with love by nestor
React, GatsbyJS, Github and Netifly are used to provide this site