alex said: I know . But that tweak includes also remove sit narower on if's and get further fwd on dm's. Do you think this could have impacted the result more? I am curios how it does with only that change Expand
I removed sit narrower because I didn't want the IFs and DMs occupying the same "space". I'll run a test soon and change just overlap to underlap and ill post the results later
alex said: @CBP87 have you considered testing it underlaps instead of overlaps? With only this modification, I know you did it with a couple of other modifications Expand
I can say that so far for FM24 we've simulated about 2,000,000 matches.
The gathered data tells us that the score usually changes in the following manager:
1,200 -> 2,400 | most of time the score changes by -/+ 1-2 points (3 points happens but very rarely)
2,400 -> 4000 | most of time the score changes by -/+ 0-1 points (2 points happens but very rarely)
This tactic was tested for 2,400 matches and got 62 points according our statistic if we test it further up to 4,000 matches then the score won't change or the change will be not more than 1 point, so at the best its score can increase to 63 points, which would be nothing special.
I'm sure you understand that it doesn't make sense spending computing power to test any tactic further if there's a very high probability that it won't produce any "special".
The below you can find the details about our testing algorithm:
- all tactics by default get tested for 1,200 matches.
- the top tactic gets an additional testing up to 4,000 matches.
- if a tactic gets into HoF after 1,200 matches then it gets an additional testing up 2,400 matches.
- if a tactic gets a score that is 2 points less than the top tactic after 1,200 matches then it gets an additional testing up to 2,400 matches
- if a tactics after 2,400 matches get a score that is 1 points less than the top tactic then it gets an additional testing up 4,000 matches Expand
Is there any chance this can be tested for the 4000 matches please? I know you don't normally do it for a 62 pointer but would it be interesting to see how it performed considering it got a 65 on the first run and then dropped 3 points on the 2nd run
Gianaa9 said: Exactly, that's why i always used Hold Up on all my Box Levante x Pirate tactics, and also in 4231 ones, glad that someone understood its potential ahah Expand
Trying to see if I can get it to work on a 433, proving very tricky
Gianaa9 said: well done! not so hopeful about balanced one, cause attacking mentality showed better results almost everytime, but so curious about Underlaps or even without both Underlaps and Overlaps. I was wondering also if Dribble More on FBs could be positive or not, cause often they make risky passes to DMs losing the ball in open spaces for the opponents (so in this cases i'd prefer them to dribble past the opponent instead of trying passes) but when it comes to pass to IFs in possession phase i'd like to have less dribble attempts by defenders.
By the way very good job, we'll see! Expand
Yeah I don't think the balanced one will reach the heights of the attacking mentality but I am intrigued to see how the underlaps work, Interesting point about underlaps/overlaps, will have to test that. I think hold up ball works hand in hand with those options though
Han106 said: Took the best set pieces from @Mark's list and I Frankenstein all of the best performers combined into one with credit labeled onto each set piece.
@Delicious's Attacking Corners Zimsmula's Defensive Third Defending Throw-ins and Attacking DFK @AideNGomiinaM's Attacking IFK @CSTG KANE's Defending Corners and Defending DFK Chou's Defending IFK
This would have to be tested to see if it has the best set of set pieces, but it's my best interpretation and attempt at it. Expand
what do you do in terms of personel list? do you leave that to your assistant manager?
Sane said: Have you considered testing 2 BWM-S instead of 2 DM-S and SS instead of AM? Perhaps, BWM will provide a slightly better option? And also, have you thought about testing force opposition outside/inside and block crosses? Expand
forcing opposition outside/inside doesn't seem to be working, well not in my tests anyways. It leaving them options blank always seem to produce better results defensively.
In regards to the BWM, I've not tried it but I'm not a big fan of roles where the players wander, so roles like BWM, RPM, BBM.
Changes-
Mentality to very attacking
Changes -
Mentality to Balanced
Wow, good going that, thanks for testing pal
Not too shabby, 96 pts and 4 cups in the bag. Thank you, I'm glad you're enjoying it
I removed sit narrower because I didn't want the IFs and DMs occupying the same "space". I'll run a test soon and change just overlap to underlap and ill post the results later
https://fm-arena.com/thread/9042-katana-4231-104p-v3-3/ already done it pal, achieved 65 points on 1200 matches but unfortunately dropped 3 points on 2400 matches
I can say that so far for FM24 we've simulated about 2,000,000 matches.
The gathered data tells us that the score usually changes in the following manager:
1,200 -> 2,400 | most of time the score changes by -/+ 1-2 points (3 points happens but very rarely)
2,400 -> 4000 | most of time the score changes by -/+ 0-1 points (2 points happens but very rarely)
This tactic was tested for 2,400 matches and got 62 points according our statistic if we test it further up to 4,000 matches then the score won't change or the change will be not more than 1 point, so at the best its score can increase to 63 points, which would be nothing special.
I'm sure you understand that it doesn't make sense spending computing power to test any tactic further if there's a very high probability that it won't produce any "special".
The below you can find the details about our testing algorithm:
- all tactics by default get tested for 1,200 matches.
- the top tactic gets an additional testing up to 4,000 matches.
- if a tactic gets into HoF after 1,200 matches then it gets an additional testing up 2,400 matches.
- if a tactic gets a score that is 2 points less than the top tactic after 1,200 matches then it gets an additional testing up to 2,400 matches
- if a tactics after 2,400 matches get a score that is 1 points less than the top tactic then it gets an additional testing up 4,000 matches
That makes sense, thanks for explaining @Droid
Changes -
Hold up added to IFs
Added focus down flanks
Added underlap
Removed overlap
Is there any chance this can be tested for the 4000 matches please? I know you don't normally do it for a 62 pointer but would it be interesting to see how it performed considering it got a 65 on the first run and then dropped 3 points on the 2nd run
Thanks
If you can't see it pal it will normally indicate that the width is set to standard, middle of the bar
Changes -
Removed overlap
Added underlap
Removed sit narrower from IFs
Added get further forward to DMs
There ya go pal but just in case it doesn't work, here is the link for it mate https://www.fmscout.com/a-fmscout-essential-fm24-views.html
Trying to see if I can get it to work on a 433, proving very tricky
By the way very good job, we'll see!
Yeah I don't think the balanced one will reach the heights of the attacking mentality but I am intrigued to see how the underlaps work, Interesting point about underlaps/overlaps, will have to test that. I think hold up ball works hand in hand with those options though
@Delicious's Attacking Corners
Zimsmula's Defensive Third Defending Throw-ins and Attacking DFK
@AideNGomiinaM's Attacking IFK
@CSTG KANE's Defending Corners and Defending DFK
Chou's Defending IFK
This would have to be tested to see if it has the best set of set pieces, but it's my best interpretation and attempt at it.
what do you do in terms of personel list? do you leave that to your assistant manager?
forcing opposition outside/inside doesn't seem to be working, well not in my tests anyways. It leaving them options blank always seem to produce better results defensively.
In regards to the BWM, I've not tried it but I'm not a big fan of roles where the players wander, so roles like BWM, RPM, BBM.
Ill test it and share the results with you
These are the ones I will be uploading tonight hopefully
Balanced -
Underlap -
Predicted, 17th