Some thought about the testing leagues.

TotalFootballFan9 said:
Guys, I've spent a ton of time on creating different testing leagues for myself and I just want to share my experience and highlight some issues of "testing league", maybe it will be helpful to someone.

First of all, I want to say that all my experience tells me it's 100 times harder to create "the best testing league" than just "the best tactic" for FM and the guy who is able to create such testing league must possess insane amount of knowledge about FM and in this case he can easily deliver the best tactic for FM because it requires much more knowledge to create the testing league where all tactics have equal chances and the testing environment fully simulates the real game environment than just a very good tactic.

Over time I'll be updating the OP of this thread and adding additional sections.


THE CONFIGURATION OF THE TESTING TEAM / ASSISTANT MANAGER PICKS THE TEAM ​

I'm sure that no one would argue the fact when you test a tactic then the result of the test depends on what team will you pick to test it with. The better your team suites for the tactic the better result you'll get.

The guy who creates the tactic testing league always give advantage to some tactic when picks players for the testing team and it isn’t his fault because it’s inevitable and here’s why it's inevitable...

Even if all players of the testing team have the equal CA it doesn't creates equal conditions for all tactics because CA is just the amount of the points which players have to spend on their attributes but it can be spend very differently.

For example, the striker has spent his 140 CA budget on Physical and Mental attributes and there's nothing left for Technical attributes so all his Technical attributes are at very poor level and if the creator of the test chooses this striker he will give advantage to tactics which rely on a striker with high Physical and Mental attributes but tactics which rely on a striker with high Technical attributes will take a strong hit in this case.

You can ask why just don't pick all type of the strikers for the testing team? Even in this case it isn't possible to exclude the influence of the "middle" man because anyway the guy who makes test have to choose which players to pick but it's extremely hard to choose different type of strikers which are equally efficient and it’s just only the half of the problem and the second half is who will pick the team? Assistant Manager?

Heh, most the time I can't hold my laugh when I see what choices Assistant Manager makes when he picks the team in the tests. Heh, I understand the most of people only watch the result of the test and don't pay attention what players were used for the test but in this case they misses the most important thing which determines everything.

I made a test. I picked West Bromwich(Predicted 10th) and edited CA/PA of my assistant manager(David Kemp) to maximum 200! Then I went on “Holiday” and left him in change of everything. How do you think it ended? Just think this guy had 200 CA/PA so in this case could I expected to find West Brom at least somewhere in the middle of the table after the season?

Nope... I was fired after 30 matches and found West Brom at the bottom of the table! I just don't understand how anyone can believe in the competence of assistance managers when they pick the team or give tactical advises?

Even 200 CA/PA Assistant Manager is capable to do only the very simple stuff such as don't pick a striker to play a GK and if someone thinks that the assistance manager is able to analyze the tactic and pick the best suited players for it than I say it's just ridiculous.

Here you can download the game save and see how "competent" is 200 CA/PA assistant manager - Assistant Manager Test




I WANT TO BELIVE​

It's logical that most people cheers any tactics tests because they think that such tests could save their time and not all people see that such test could do opposite and waste their time.

For example, the poorly configured testing league could put the bad tactic on the top and make you waste your time on testing it.

I want to repeat once more what I said in the beginning of the post. In order to create the testing league which provides the results that could be taken into consideration you need to have a **** amount of knowledge about FM and if someone possess such knowledge he would have already provided us with a awesome tactic for FM.
Click to expand...
Brilliant write up mate
 
the saddest thing about this is that you actually play "against" these AI's, and the game loses its challange / gets boring once you start dominating after 2-4 seasons
 
If only the tactic tests used holiday as well.
 
Ignoring the holidaying......

Set up a tactic with a libero on attack duty and see who your assistant manager places there when you ask him to pick the team?

Screenies welcome :)
 
callamity said:
Ignoring the holidaying......

Set up a tactic with a libero on attack duty and see who your assistant manager places there when you ask him to pick the team?

Screenies welcome :)
Click to expand...
did a search on the data editor yesterday and there are only 2 players in the whole database with over 130 CA who can play sweeper.

will load up some of my saves and see.
 
Will1981 said:
did a search on the data editor yesterday and there are only 2 players in the whole database with over 130 CA who can play sweeper.

will load up some of my saves and see.
Click to expand...


If a player has the stats to play the role he will do it well enough till he learns it.

EDIT: I use defensive midfielder in this role as I think it performs the half back role better than a half back does.
 
Last edited:
Here is my opinion You can have a great tactic, great players, follow the instructions down to a T. However if you draw a game, or morale slips or if you mismanage a player grievance, it can all fall to ****.

Hence why in tactic testing leagues teams go on a "bad run". Any human player would be having team meetings, chats with layers and squad rotation etc to rectify.

As TTF also says, players not being picked in right positions either can be a ****. The palace save I am testing in the now will have the ***.Man pick Zaha as my inside forward every single game if I let him over Puncheon who is clearly the better not just for the role, but the specific instructions I have set.

The problem isn't with the testing leagues but actually with the human element involved. As you can have all the variables perfectly tee'd up but make bad decisions, a poor choice of team talk or no remedial action to rectify matters when they go south.

Plug and Play is dead. I say this as an author of tactics over 10 years who tries to plug and play. These tests rely on plug and play conditions, all of them do. The game has evolved since then massively to SI's credit.

No point hating the testers. Not one of them is crooked, its just that NO MATTER what they do, as TTF indicates, there are variables that massively influence performance that can make things go **** up.

Want proof. Look at all teams tested, the vast majority go through those "dips" in forms I referenced above. Be it two/three games or six. I know, I've looked already.

I think what tactic creators need to do is up their game. Don't just release a tactic but release a guide or a "how to" for different scenarios. Not just fav, underdog and balanced but how to cope with all the scenarios the game throws up. It's not the testers fault, more the creators for being half arsed and lazy. I know because i'm guilty as charged. I know **** right folks won't mirror my results unless they play a similar way.

I'll release a tactic tonight giving what i perceive to be the best possible chance for success. It will be a mammoth opening post but maybe that's what us creators need to do now, instead of ******** that folk "don't get it". Ideally every post from a creator on his/her thread should be "refer to OP", if not, they themselves have ****** up.

My two cents for what it was worth.
 
Last edited:
I've updated the OP and added this seaction:

Btw, guys all what I've posted so far is only about 20% from what you should take into consideration when crating a testing league. Heh, as I said it's like a rocket science and if you ignores all this stuff then you can throw away the result of the tests because they won't have any use in real game.


THE CONFIGURATION OF THE TESTING LEAGUE​

The results of the tests fully depends on the configuration of the testing league.

For example, you'll get very different results if you use 140 CA players and if you use 175 CA players because 175 CA players is much "smarter" than 140 CA player and 175 CA players are capable to do much more things on the pitch than 140 CA.

Let's say your tactic has extremely high D-Line and if you play with such high D-Line against 140 CA players then probably you’ll get away with that because 140 CA players don’t have high enough Mental and Technical attributes to exploit the huge space behind your D-line. Speaking other words 140 CA players don’t see space behind your D-line and even if they see it they don’t have enough technique to make such passes so your tactic could work just fine against 140 CA players but if you play against 175 CA players they will hammer you.

As you can see CA of the players in the testing league has crucial importance and determines the result of the tests.
 
Chrissy Dude said:
Here is my opinion You can have a great tactic, great players, follow the instructions down to a T. However if you draw a game, or morale slips or if you mismanage a player grievance, it can all fall to ****.

Hence why in tactic testing leagues teams go on a "bad run". Any human player would be having team meetings, chats with layers and squad rotation etc to rectify.

As TTF also says, players not being picked in right positions either can be a ****. The palace save I am testing in the now will have the ***.Man pick Zaha as my inside forward every single game if I let him over Puncheon who is clearly the better not just for the role, but the specific instructions I have set.

The problem isn't with the testing leagues but actually with the human element involved. As you can have all the variables perfectly tee'd up but make bad decisions, a poor choice of team talk or no remedial action to rectify matters when they go south.

Plug and Play is dead. I say this as an author of tactics over 10 years who tries to plug and play. These tests rely on plug and play conditions, all of them do. The game has evolved since then massively to SI's credit.

No point hating the testers. Not one of them is crooked, its just that NO MATTER what they do, as TTF indicates, there are variables that massively influence performance that can make things go **** up.

Want proof. Look at all teams tested, the vast majority go through those "dips" in forms I referenced above. Be it two/three games or six. I know, I've looked already.

I think what tactic creators need to do is up their game. Don't just release a tactic but release a guide or a "how to" for different scenarios. Not just fav, underdog and balanced but how to cope with all the scenarios the game throws up. It's not the testers fault, more the creators for being half arsed and lazy. I know because i'm guilty as charged. I know **** right folks won't mirror my results unless they play a similar way.

I'll release a tactic tonight giving what i perceive to be the best possible chance for success. It will be a mammoth opening post but maybe that's what us creators need to do now, instead of ******** that folk "don't get it". Ideally every post from a creator on his/her thread should be "refer to OP", if not, they themselves have ****** up.

My two cents for what it was worth.
Click to expand...
Two great ****** cents worth mate
 
Nobody uses holiday in their tests.
As far as for everything else, I can see your point, but I think the whole point of those leagues is to give results within 15-20% of accuracy. People use instant result, AI teamtalks and AI squad selection, because it creates common pattern and field for testing. The point of any test is to make conditions equal for each tactic. Harder to do it when doing all micromanagement by yourself.
Watching full matches might be more accurate, and AI may indeed do small tweaks such as changing player roles, but who does have the time to run 10+ seasons watching matches?
I have tried most of the tactics that are being tested in those leagues, and result are pretty accurate, as I said within 15-20%. I am more than happy with that, and not only me. I can go visit a thread, then open several most popular tactics, pick the one that suits my team and my playing style, and be confident that with right management I will succeed.
No offense to you, you are one of the best tactitians here, but I think you take those tests too seriously. Nobody is going to put days and weeks creating ideal environment and testing countless tactics, tweaks, new versions and etc. Your executioner tactic works for many many people, it is a wonderfull tactic, and test results only prove that. Sure, they might be not as accurate, but at least they make it easy for casual people here to make right decision, isn't that the whole purpose?
 
played sterling as striker, henderson and alen as DM
whoever AM pick i dont care
atribute filter is smart way to tell "get a better player so this tactic will work" (except ppm filter)
 
using holiday to test = the AM will change the role depend on his(SI game knowledge) preferences
 
pavleg said:
Nobody uses holiday in their tests.
Click to expand...

wkdsoul said:
If only the tactic tests used holiday as well.
Click to expand...

My point was to show how the incompetence of the assistant manage even if he has 200 CA.

Of course during the test you won't use "Holiday" mode because it would be just ridiculous but assistant manage still picks the team!

I remember how assistant manager in some of Will's test picked Angel Di Maria to play as BWM! It was just hilarious! :)
 
Why don't you'se all get together and make the best possible tactic?
 
You must log in or register to reply here.

FMBase

Beta

Resources

Back
Top