Google Docs
Dropbox
Introduction
I have been in and around the FM scene since the Championship Manager days. I was a regular contributor on many FM related forums. However I stopped playing in 2012/13 as the game was taking up too much of my time. I have returned this year however I no longer have access to my original login details/email etc. My day job involves testing databases and collating data, so I was naturally interested in creating a testing DB for FM in which I would attempt to remove as much randomness as possible without changing the core gameplay. Below I have outlined my attempt.
Please note this is the final testing DB, which I have tested hundreds of times making small modifications until I was happy with the results and confident enough to publish them here.
Test Team
The test team for the first round of testing is West Ham, original West Ham team, predicted to finish 8th in the test league so replicating a solid mid table/top 10 team. No transfers as transfer embargo is in place. West Ham are a good mid table team and favourites against some teams and underdogs against the others. The tactical attributes of the test team have been set to 10 and the Assistant Manager tendencies have also been set to 10. The Assistant Manager has a CA of 150. No other changes to staff a part from extending contracts of all players and staff.
For the second round of testing, I have slightly modified my original db, hopefully it is now even more accurate. I will re-test all the tactics which scored 135+ points in the West Ham test with Liverpool. I have again slightly modified the Liverpool team by changing a few roles and mirroring players feet to ensure fairness to all tactics. The Liverpool team is predicted to finish 4th in the test league, so this test will replicate the usage of these tactics with a top 4 Premier League team. All other attributes remain the same as explained below. As with the previous test team, all PPM's have been removed.
Changes made to test teams - Removed all PPM's. Mirrored some players feet to other foot to make sure all positions have natural foot players. ie. Payet originally had 12 left foot and 20 right foot. I mirrored this to 20 left foot, 12 right foot, so he can play as right inside forward or left winger. No players have 20 both feet, I only mirrored some players feet so all formations will have a fair chance during the testing process. Also changed some player role to again ensure fairness to all tactics. No others attributes changed a part from those listed below which were changed for all AI players and test team players.
AI Players
All players are retired from international duty and will retire at the end of the season. This is to ensure players don't complain about not playing. All players have had the nationality of the division and language assigned to them as a second nationality. No AI players attributes changes, no PPM's removed. The only attributes which have been fixed are below, which were changed for AI and test team players. Morale and condition is frozen for ALL teams using Panos's Fitness Coach (thanks to Panos).
Fixed attributes:
Consistency - 15
Dirtiness - 10
Important Matches - 10
Injury Proneness - 1
Natural Fitness - 15
Stamina - 15
Versatility - 15
Adaptability - 15
Loyalty - 15
Professionalism - 15
Pressure - 10
Training facilities changed to 20 for all test division teams.
All rivalries removed to ensure all games are played on an equal footing.
League Structure
The test division is a two tier division with the top 10 Premier League teams sorted by reputation in the second tier and bottom 10 Premier league teams in the top tier. We are not concerned with the top tier as the reason for this setup is to ensure all games are competitive. The top two teams in the division receive automatic promotion, the next four teams go into a playoff where they will fight for the one remaining promotion place. This ensures all teams have something to fight for until the end of the season. All teams will play each other 10 times, so a total of 90 games, which is the equivalent of 2 and a bit Premier League seasons. (The idea for this structure came from a post made by TFF, so credit to him).
I have assigned all the best referees in the game to the league and there are no suspensions, so all teams can field their best starting 11. There are no suspensions but with the best possible referees, the data collated for yellow and red cards should be accurate.
AI Managers
Luis Enrique, Guardiola, Mourinho, Conte etc have all been assigned to teams which didn't have world class managers. This is to ensure formation diversity as some managers (Klopp) will always attack, and some managers (Mourinho) will mostly defend, so it doesn't matter if the test team is a top team or an underdog team.
The AI uses a different approach depending on the reputation, ability etc of the test team. This means a tactic that does well with an underdog team might not do so well with a strong team. To mitigate this I have assigned some of the most popular formations to the AI managers and changed their tendencies to make sure some teams will always use an attacking mentality and other teams will adopt a defensive mentality. I believe this will produce balanced results which can be replicated by an underdog or strong team in actual gameplay.
I have also changed formation tendencies for some managers to ensure all teams don't use the same formations (again this is for formation and tactical diversity). I have also changed board confidence for all AI managers to 200 to mitigate early sackings.
Testing Process
I don't arrange any friendlies as I use FMRTE to change match preparation to 100 for each tactic so it is fully fluid when the league commences.
Training is set to balanced and match preparation is set to defensive positioning.
No team talks, press conferences, team meetings etc.
Simulate match (no game plan) is used for all games.
OI's as instructed by the author. Please see Dropbox screenshots. I will omit OI's if the author has requested the OI's to be omitted.
I will pick the team myself to ensure the role ability is as high as possible. I will also follow the author's instructions regarding the type of player on each wing etc. ie left footed player on left wing or right footed etc. Once the team is picked it will remain the same throughout the season as there are no injuries, suspensions etc.
I will use my best judgement when selecting setpiece takers unless specified by the author.
I will use the same saved game file for each and every test. The fixture list is exactly the same for all tested tactics. This ensures fairness for all tested tactics.
Results
I will update the thread when a new tactic has been tested. I will post a link to the Google docs spreadsheet. I will upload various screenshots to my Dropbox account, again I will share the Dropbox link, so you can view all the screens yourself. If there is enough demand I will make the final save file available for analysis. Please note you will not be able to play with this file as the players will retire at the end of the season. Save files will be for analysis only. (Spreadsheet originally posted by Mr Langvatn, modified by myself).
The results from your testing are different to others?
I have been open and honest in how I have setup the testing DB. I have made every effort to remove as much randomness whilst trying to retain the core gameplay. I could quite easily modify and remove other bits and pieces, however those results will not mean anything when compared to how we actually play the game.
In closing it is up to the user to decide which testing process to put their faith in. My job is to test, your job is to discuss.
07 May - I have tested two of the tested tactics a further three times and across all the tests there has been a 7 point deviation with one and 9 points with the other. I am quite happy with this as it is impossible to remove all the randomness from the game. However there was a big deviation with the number of yellow cards.
09 May - I have refined my original db which should be even more accurate than before. I will re-test all the tactics which scored 135+ points in the first test with Liverpool, who are predicted to finish 4th in the test league. West Ham were predicted to finish 8th. So this test will replicate the usage of these tactics with a top 4 team. I have amended my original post and full details can be found there.
Dropbox
Introduction
I have been in and around the FM scene since the Championship Manager days. I was a regular contributor on many FM related forums. However I stopped playing in 2012/13 as the game was taking up too much of my time. I have returned this year however I no longer have access to my original login details/email etc. My day job involves testing databases and collating data, so I was naturally interested in creating a testing DB for FM in which I would attempt to remove as much randomness as possible without changing the core gameplay. Below I have outlined my attempt.
Please note this is the final testing DB, which I have tested hundreds of times making small modifications until I was happy with the results and confident enough to publish them here.
Test Team
The test team for the first round of testing is West Ham, original West Ham team, predicted to finish 8th in the test league so replicating a solid mid table/top 10 team. No transfers as transfer embargo is in place. West Ham are a good mid table team and favourites against some teams and underdogs against the others. The tactical attributes of the test team have been set to 10 and the Assistant Manager tendencies have also been set to 10. The Assistant Manager has a CA of 150. No other changes to staff a part from extending contracts of all players and staff.
For the second round of testing, I have slightly modified my original db, hopefully it is now even more accurate. I will re-test all the tactics which scored 135+ points in the West Ham test with Liverpool. I have again slightly modified the Liverpool team by changing a few roles and mirroring players feet to ensure fairness to all tactics. The Liverpool team is predicted to finish 4th in the test league, so this test will replicate the usage of these tactics with a top 4 Premier League team. All other attributes remain the same as explained below. As with the previous test team, all PPM's have been removed.
Changes made to test teams - Removed all PPM's. Mirrored some players feet to other foot to make sure all positions have natural foot players. ie. Payet originally had 12 left foot and 20 right foot. I mirrored this to 20 left foot, 12 right foot, so he can play as right inside forward or left winger. No players have 20 both feet, I only mirrored some players feet so all formations will have a fair chance during the testing process. Also changed some player role to again ensure fairness to all tactics. No others attributes changed a part from those listed below which were changed for all AI players and test team players.
AI Players
All players are retired from international duty and will retire at the end of the season. This is to ensure players don't complain about not playing. All players have had the nationality of the division and language assigned to them as a second nationality. No AI players attributes changes, no PPM's removed. The only attributes which have been fixed are below, which were changed for AI and test team players. Morale and condition is frozen for ALL teams using Panos's Fitness Coach (thanks to Panos).
Fixed attributes:
Consistency - 15
Dirtiness - 10
Important Matches - 10
Injury Proneness - 1
Natural Fitness - 15
Stamina - 15
Versatility - 15
Adaptability - 15
Loyalty - 15
Professionalism - 15
Pressure - 10
Training facilities changed to 20 for all test division teams.
All rivalries removed to ensure all games are played on an equal footing.
League Structure
The test division is a two tier division with the top 10 Premier League teams sorted by reputation in the second tier and bottom 10 Premier league teams in the top tier. We are not concerned with the top tier as the reason for this setup is to ensure all games are competitive. The top two teams in the division receive automatic promotion, the next four teams go into a playoff where they will fight for the one remaining promotion place. This ensures all teams have something to fight for until the end of the season. All teams will play each other 10 times, so a total of 90 games, which is the equivalent of 2 and a bit Premier League seasons. (The idea for this structure came from a post made by TFF, so credit to him).
I have assigned all the best referees in the game to the league and there are no suspensions, so all teams can field their best starting 11. There are no suspensions but with the best possible referees, the data collated for yellow and red cards should be accurate.
AI Managers
Luis Enrique, Guardiola, Mourinho, Conte etc have all been assigned to teams which didn't have world class managers. This is to ensure formation diversity as some managers (Klopp) will always attack, and some managers (Mourinho) will mostly defend, so it doesn't matter if the test team is a top team or an underdog team.
The AI uses a different approach depending on the reputation, ability etc of the test team. This means a tactic that does well with an underdog team might not do so well with a strong team. To mitigate this I have assigned some of the most popular formations to the AI managers and changed their tendencies to make sure some teams will always use an attacking mentality and other teams will adopt a defensive mentality. I believe this will produce balanced results which can be replicated by an underdog or strong team in actual gameplay.
I have also changed formation tendencies for some managers to ensure all teams don't use the same formations (again this is for formation and tactical diversity). I have also changed board confidence for all AI managers to 200 to mitigate early sackings.
Testing Process
I don't arrange any friendlies as I use FMRTE to change match preparation to 100 for each tactic so it is fully fluid when the league commences.
Training is set to balanced and match preparation is set to defensive positioning.
No team talks, press conferences, team meetings etc.
Simulate match (no game plan) is used for all games.
OI's as instructed by the author. Please see Dropbox screenshots. I will omit OI's if the author has requested the OI's to be omitted.
I will pick the team myself to ensure the role ability is as high as possible. I will also follow the author's instructions regarding the type of player on each wing etc. ie left footed player on left wing or right footed etc. Once the team is picked it will remain the same throughout the season as there are no injuries, suspensions etc.
I will use my best judgement when selecting setpiece takers unless specified by the author.
I will use the same saved game file for each and every test. The fixture list is exactly the same for all tested tactics. This ensures fairness for all tested tactics.
Results
I will update the thread when a new tactic has been tested. I will post a link to the Google docs spreadsheet. I will upload various screenshots to my Dropbox account, again I will share the Dropbox link, so you can view all the screens yourself. If there is enough demand I will make the final save file available for analysis. Please note you will not be able to play with this file as the players will retire at the end of the season. Save files will be for analysis only. (Spreadsheet originally posted by Mr Langvatn, modified by myself).
The results from your testing are different to others?
I have been open and honest in how I have setup the testing DB. I have made every effort to remove as much randomness whilst trying to retain the core gameplay. I could quite easily modify and remove other bits and pieces, however those results will not mean anything when compared to how we actually play the game.
In closing it is up to the user to decide which testing process to put their faith in. My job is to test, your job is to discuss.
07 May - I have tested two of the tested tactics a further three times and across all the tests there has been a 7 point deviation with one and 9 points with the other. I am quite happy with this as it is impossible to remove all the randomness from the game. However there was a big deviation with the number of yellow cards.
09 May - I have refined my original db which should be even more accurate than before. I will re-test all the tactics which scored 135+ points in the first test with Liverpool, who are predicted to finish 4th in the test league. West Ham were predicted to finish 8th. So this test will replicate the usage of these tactics with a top 4 team. I have amended my original post and full details can be found there.
Last edited: