FM-Arena Tactic Testing League

by Alexander, Nov 12, 2020

Here are some highlights of FM-Arena tactic testing league:

- All players in the league have their morale and conditions set and "frozen" at the highest possible level by an in-game editor, so the players' morale and physical conditions don't change and stay at the maximum level throughout the testing process. Injuries are removed when they occur. This way we greatly reduce the game randomness and the results become significantly more consistent compared with testing during normal game.

- All tactics regardless their formations/roles/duties are tested with identical players that are equally good in doing different tasks on the pitch, it guarantees that the results aren't affected by difference in the players' quality.

- All teams have the "Team Cohesion" and "Tactic Familiarity" maximized.

- The players that are used to test tactics don't have any "Player Traits" and that guarantees they play according with the tactical instructions and not their "Player Traits".

- The variety of the AI tactics in the league ensures that tactics are tested againt the most optimal set of AI tactics.

- In our testing league the AI managers can't be sacked during the season due poor results and they stay in charge until the end of the testing so it means that every tactic is tested against the same set of AI managers.

- We test tactics for 912 matches and that guarantees a high accuracy.


The above are just some of FM-Arena tactic testing league features and there are much more.

+1

Hi!

Would it be possible to have xG's in the tactics test results?

Thank you B

0

KOniGun said: Hi!

Would it be possible to have xG's in the tactics test results?

Thank you B


Hi,

We thought about adding xG stat but decided not to do that because xG stat could be useful only if you test tactics for a small amount of matches, like 10 or 20 matches but we test tactics for 456 matches. When you test tactics that much then there's no need for xG stat.

0

Ok you're right

0

What is the average CA of Team A and Team B? It baffles me that the testing results of FM-Arena are SO vastly different from FMBase and from Holiday Testing with Frozen players.

Not saying one is better than the other but the stark differences go well beyond variance.

0

healmuth said: What is the average CA of Team A and Team B? It baffles me that the testing results of FM-Arena are SO vastly different from FMBase and from Holiday Testing with Frozen players.

You can download a game save from any test and get any data you wish :)

0

healmuth said: What is the average CA of Team A and Team B? It baffles me that the testing results of FM-Arena are SO vastly different from FMBase and from Holiday Testing with Frozen players.

Not saying one is better than the other but the stark differences go well beyond variance.


I downloaded the save file from FM Base, and can conclude that their test will never be correct for unconventional tactics like strikerless. For example, Phoenix works best with fast SS who have good finishing, but the teams given in FM Base test do not have that kind of players. Meanwhile, the teams in FMArena test offer all kind of players for every role, and the tactic just works flawlessly.

+2

FMAdept said: I downloaded the save file from FM Base, and can conclude that their test will never be correct for unconventional tactics like strikerless. For example, Phoenix works best with fast SS who have good finishing, but the teams given in FM Base test do not have that kind of players. Meanwhile, the teams in FMArena test offer all kind of players for every role, and the tactic just works flawlessly.

Also, the assistant manager picks the team in the FM Base testing but any assistant manager has no clue how pick the team for strikerless tactics. In general, any testing when the assistant manager picks the team is flawed because the AI doesn't know how properly pick the team for many tactics.

+2

healmuth said: It baffles me that the testing results of FM-Arena are SO vastly different from FMBase and from Holiday Testing with Frozen players.
In FM-Base testing when the season isn't over yet many teams change their managers due poor results and that has great impact on the result for every tactic but that doesn't happen in FM-Arena the AI teams don't change their managers until the end of the season.

0

FMAdept said: I downloaded the save file from FM Base, and can conclude that their test will never be correct for unconventional tactics like strikerless. For example, Phoenix works best with fast SS who have good finishing, but the teams given in FM Base test do not have that kind of players. Meanwhile, the teams in FMArena test offer all kind of players for every role, and the tactic just works flawlessly.

We clone players for Strikerless tactics prior to testing, besides each team has a crazy amount of players available for any role. Are you sure you've looked properly at the save game?


Poacher said: In FM-Base testing when the season isn't over yet many teams change their managers due poor results and that has great impact on the result for every tactic but that doesn't happen in FM-Arena the AI teams don't change their managers until the end of the season.

This might be accurate for you, but we use FMRTE to ensure staff can't move teams by giving crazy wages. It's something we noticed in the last patch.

-----------------------------------------------------

Any difference in results is solely down to the fact we test in different ways. FM Base use a sub-top team and an underdog team whilst FM Arena as far as I'm aware use two elite teams. There's no worse or better way to test due to RNG but we use some memory manipulation to try to remove as many variables as possible, something I'm sure Arena also do.

0

Liam said: This might be accurate for you, but we use FMRTE to ensure staff can't move teams by giving crazy wages. It's something we noticed in the last patch.

"Crazy wages" don't prevent the AI managers being sacked and it's basic knowledge

0

Markizio said: "Crazy wages" don't prevent the AI managers being sacked and it's basic knowledge

There are things we do behind-the-scenes. Mr L has been testing tactics since 2013. Do you really think we haven't seen these issues before? ;)

0

Liam said: There are things we do behind-the-scenes. Mr L has been testing tactics since 2013. Do you really think we haven't seen these issues before? ;)



This only works for human managers and it doesn't work for AI managers and that is also basic knowledge

0

Markizio said: This only works for human managers and it doesn't work for AI managers and that is also basic knowledge

Happy new year man and rest assured we are always looking to improve accuracy of testing.

+4

I think this data so good but it reflect only a part of game. I can play high press, tackling hard and dont care about fitness and injury ( much more risk but low cost ) - important if you play gegen ( like LIV in this season ). Top best tactic in this data got format: Attack, High press, direct pass.

0

May i ask something?  i ve seen that the section ''GFC'' goal from corners have been removed , why is that? so far these years it was the only forum that had this section and now has been removed.. this is so sad :/

0

The M1rakuru said: May i ask something?  i ve seen that the section ''GFC'' goal from corners have been removed , why is that? so far these years it was the only forum that had this section and now has been removed.. this is so sad :/

Hi,

Unfortunately, we had to stop delivering "gfc" and "poss" stats due some technical limitations. I hope we can bring those stats back at some point in the future. Until then I'm afraid the only way to get those stats is downloading game saves and checking the stats manually.

0

Is it possible to download save for tactic testing?

0

Calvino said: Is it possible to download save for tactic testing?

Hi,

At the moment we don't share it but that might change in the future.

+1
Create an account or log in to leave a comment