Set Pieces

by Mark, Feb 8, 2021

Any thoughts on taking Set Pieces routines from some of the top tactics and testing them with Phoenix to see which is best?

0

Mark said: Any thoughts on taking Set Pieces routines from some of the top tactics and testing them with Phoenix to see which is best?

Only Corners or Throw-ins makes any meaningful difference and other set pieces are irrelevant.

0

Grimlock said: Only Corners or Throw-ins makes any meaningful difference and other set pieces are irrelevant.

Do you have a set that works well?

0

Mark said: Do you have a set that works well?

I'm not an expert when it comes to set pieces, when I need good set pieces I usually take them from highly rated tactics.

0

If anyone wants to test set pieces I have downloaded the routines for the following tactics:

Phoenix 3.0
ZaZ Blue 2.0
Viola V1
Fireball Stepback
Killer
Motion
Void
Knap Holy Ghost

All rated pretty highly. Put Knap in as the first non strikerless tactic.

Also included TFF as many used his set pieces early on, and one that claimed to have the best set pieces on FM Scout (Lie Cheat Steal) plus my interpretation of the set pieces from SP God.

ZaZ Blue 2.0.fmf
Downloaded : 605 times
Uploaded : Feb 9, 2021
Void.fmf
Downloaded : 460 times
Uploaded : Feb 9, 2021
Viola V1.fmf
Downloaded : 393 times
Uploaded : Feb 9, 2021
TFF Routine Collection.fmf
Downloaded : 431 times
Uploaded : Feb 9, 2021
SP God.fmf
Downloaded : 428 times
Uploaded : Feb 9, 2021
Phoenix 3.0.fmf
Downloaded : 435 times
Uploaded : Feb 9, 2021
Motion.fmf
Downloaded : 641 times
Uploaded : Feb 9, 2021
Lie Cheat Steal.fmf
Downloaded : 417 times
Uploaded : Feb 9, 2021
Killer.fmf
Downloaded : 390 times
Uploaded : Feb 9, 2021
HolyGhost.fmf
Downloaded : 444 times
Uploaded : Feb 9, 2021
Fireball.fmf
Downloaded : 441 times
Uploaded : Feb 9, 2021
0

What do you think @Zippo?

0

Mark said: What do you think @Zippo?

I find that testing Set Pieces is a highly complicated thing to do. You can't just take one tactic set pieces and put them into some other tactic and expect to see it works the same in both tactics. I can tell you it doesn't work this way and if you try to do that then you'll end up with a complete mess because Set Pieces are tied to the formation/roles/duties so you have to tune them for every tactic and you must be good at doing that and understand what are you doing so it requires a lot of efforts and knowledge.

1

Also, stats from set pieces can be misleading. For example, a tactic can have several goals from corners not because the set pieces are good, but because it shoots a lot and get lots of corners. Same for free kicks, getting lots of fouls because your players drible more in dangerous are. For defense, a tactic can have very good results because it makes most fouls further from the box, making you feel like it concedes less from free kicks. That being said, it's very hard to test.

1

Zippo said: I find that testing Set Pieces is a highly complicated thing to do. You can't just take one tactic set pieces and put them into some other tactic and expect to see it works the same in both tactics. I can tell you it doesn't work this way and if you try to do that then you'll end up with a complete mess because Set Pieces are tied to the formation/roles/duties so you have to tune them for every tactic and you must be good at doing that and understand what are you doing so it requires a lot of efforts and knowledge.

Fair enough. I thought it was worth a punt as there really are no forums on this around, unlike my last game which was FM19 which had a set piece flaw with throw ins

0

Mark said: Fair enough. I thought it was worth a punt as there really are no forums on this around, unlike my last game which was FM19 which had a set piece flaw with throw ins

It's still possible to test for specific tactics. For example, you can pick Phoenix or Blue and plugin each set piece, testing like three times each to see if there is significant difference. However, you still have to adjust the set pieces so the positions are right, because they can get messed if you change the formation. For example, if I plug in Phoenix set pieces in Blue, it will have wrong positions in set pieces because of how the tactic was set initially. Another problem is that some tactics just have different positions, like some with more midfielders, more strikers or without wingers, which would need to be adjusted accordingly. Anyway, it can be done for a specific tactic, but it will also need those kind of adjustments.

0

Ok, I think I will try them in the save I had testing lower league management. All the Set Pieces are coming from the successful tactics for FM21 on FM Arena. The are very much all similar with most being 4-1-3-2 strikerless and a few 4-1-2-3 strikerless. Mainly MC moving to AMC. I will adjust where positioning/roles are substantially different. I understand the complexities on role and duties but will see what it throws up and then test the winner in my real save or against the lower league management save I had with Viola and ZaZ Blue. On the comments from @ZaZ, I think have the same team means these things would even out. Anyhow, still keen to see if any of the set pieces deliver a dominant result after 3 runs in 5 divisions for each routine versus the original Phoenix v3.0 results. Let me know if you want to know the results

1

Mark said: Ok, I think I will try them in the save I had testing lower league management. All the Set Pieces are coming from the successful tactics for FM21 on FM Arena. The are very much all similar with most being 4-1-3-2 strikerless and a few 4-1-2-3 strikerless. Mainly MC moving to AMC. I will adjust where positioning/roles are substantially different. I understand the complexities on role and duties but will see what it throws up and then test the winner in my real save or against the lower league management save I had with Viola and ZaZ Blue. On the comments from @ZaZ, I think have the same team means these things would even out. Anyhow, still keen to see if any of the set pieces deliver a dominant result after 3 runs in 5 divisions for each routine versus the original Phoenix v3.0 results. Let me know if you want to know the results

I am interested. Whatever works for Phoenix will work for Blue, since they both have a similar structure and play style.

0

I will post when I complete the runs. I have found duplicates so not sure how many there will be. Should be done in a couple of days or so.

3

Mark said: I will post when I complete the runs. I have found duplicates so not sure how many there will be. Should be done in a couple of days or so.

I think it is a great idea. Set pieces are essential in real football. I love being able to work on these situations in my games.
Get a solid tactic in defense, where set pieces can win the game or open it.
I want to see the results of your tests.

0

I have done 10 test runs so far, all across 5 Divisions. I am doing 3 test runs across 5 Divisions for each Set Piece Routine, only using the Set Piece Routine on the lowest ranked team of each Division. I have definitely ruled out using Lie Cheat Steal as it returned 20% less points per game than the current leader. Also going to exclude Fireball which 8% below current leader. Motion is the current leader and sits 5% above the base Phoenix results. The tests continue.

2

Mark said: I have done 10 test runs so far, all across 5 Divisions. I am doing 3 test runs across 5 Divisions for each Set Piece Routine, only using the Set Piece Routine on the lowest ranked team of each Division. I have definitely ruled out using Lie Cheat Steal as it returned 20% less points per game than the current leader. Also going to exclude Fireball which 8% below current leader. Motion is the current leader and sits 5% above the base Phoenix results. The tests continue.

Make sure to divide the table in assists from throw ins, corners and free kicks, as well as direct goals from free kicks. If possible, do so for both goals scored and conceded, so we can analyze attack and defense. Some set piece can be better for corners but worse for free kicks, for example.

P.S.: Instead of using absolute number of assists by set pieces, it's probably better to use the % of total goals, to make it more precise.

1

Have you tried this one?
https://fm-arena.com/find-comment/3307/ Ca

0

blackmoix said: Have you tried this one?
https://fm-arena.com/find-comment/3307/ Ca


Not yet but it has been added to the list

0

ZaZ said: Make sure to divide the table in assists from throw ins, corners and free kicks, as well as direct goals from free kicks. If possible, do so for both goals scored and conceded, so we can analyze attack and defense. Some set piece can be better for corners but worse for free kicks, for example.

P.S.: Instead of using absolute number of assists by set pieces, it's probably better to use the % of total goals, to make it more precise.


Where can I source this information. I have all the game saves from the end of each season

0

Mark said: Where can I source this information. I have all the game saves from the end of each season

Team Report -> Analyst Report -> Scoring/Conceding -> Assists. You need a performance analyst, though. The lower teams probably don't have and didn't collect that data.

0

Mark said: Not yet but it has been added to the list

Disclaimer, I don't know whether it was a fluke but when I tested the Rego pyramid long throw from fm19, the first game I scored 3 goals from long throws, I noticed that in highlights some players set as go forward run towards the guy throwing as if they are set on come short and that draws some opposition with them leaving the guy on the far post with fewer markers and he usually scores if it doesn't get blocked by the keeper (which happens a lot on fm21) I am not a setpiece expert, but I reckon that if one can identify which positions are drifting towards the "come short" and leave them with the go-forward instructions, and adjust the ones in the box maybe this could be exploitable. Just a thought.

0

ZaZ said: Team Report -> Analyst Report -> Scoring/Conceding -> Assists. You need a performance analyst, though. The lower teams probably don't have and didn't collect that data.

I think I can access this info for most if not all the teams but cant download the data - so will have to be a manual process at the end. I agree that his info is definitely of interest in the analysis and may allow a mix and match of best types of set pieces.

1

I wont be doing Knap or Killer as they look too similar to some of the other set piece routines. I have added the FM19 one suggested by Madundwa, because I loved FM19 so much. That makes it ten routines. SP God is done, TFF last run is just going, and I should get through Void tonight. Havent downloaded the results yet but will probably do that depending on how exciting the Australian Open tennis is. It is not easy being an Aussie all round sportsman. That will leave FM19, and the best 3 tactics of ZaZ Blue, rerun Phoenix under this set up and Viola.

Just a quick update. Might have some more initial results tomorrow.

1

Here are the results:

Only 3 set piece routines worth looking at to be honest, Motion, Void and Phoenix. I am running with Void with a few modifications, I think statistically it looks more solid. Wont get a chance to test it all more for a week or so so wont post an update until then.

1

And here is my modified version of set pieces that I am using with my modified version of ZaZ Blue 2.0 where I have changed the Central Defenders to CD Defend and the Wingers to Wingers Support to suit my team. Working well so far

UPDATE: Did not test well

MDW.fmf
Downloaded : 335 times
Uploaded : Feb 14, 2021
0

I didn't know my set pieces were that bad. I mean, it's not bad, but the set pieces from Void, Phoenix and Motion seem much better. Are you making extra experiments on those three to avoid statistical noise? I am just waiting for your final results to update my tactics and give you credits.

0

Mark said: And here is my modified version of set pieces that I am using with my modified version of ZaZ Blue 2.0 where I have changed the Central Defenders to CD Defend and the Wingers to Wingers Support to suit my team. Working well so far
I think I have tried that but It doesn't make much diff with Winger and CD

0

ta2199 said: I think I have tried that but It doesn't make much diff with Winger and CD

My squad has very much higher ratings for roles for those positions. But you are probably right.

ZaZ said: I didn't know my set pieces were that bad. I mean, it's not bad, but the set pieces from Void, Phoenix and Motion seem much better. Are you making extra experiments on those three to avoid statistical noise? I am just waiting for your final results to update my tactics and give you credits.

I am not sure whether there is that much certainty in this stuff. The tests seem to be quite variable. I will see how it plays in my save using the modified version and then test later in the week. Was just a bit of fun looking at it all and then comparing the ones that seemed better against the ones that appeared not to be good. Didnt pay much attention to the ones in the middle.

1

Mark said: My squad has very much higher ratings for roles for those positions. But you are probably right.



I am not sure whether there is that much certainty in this stuff. The tests seem to be quite variable. I will see how it plays in my save using the modified version and then test later in the week. Was just a bit of fun looking at it all and then comparing the ones that seemed better against the ones that appeared not to be good. Didnt pay much attention to the ones in the middle.


I guess it's pretty clear that my set pieces are far from the best and can be improved. I will run some experiments with those three top set pieces (and mine) and see what I can copy from each type of set piece. Then  I will see if I can improve it even further.

About the roles, I wouldn't care much about the assistant opinion. I actually doubt the number of stars make any difference at all. I think the only things that matter are tactical affinity and position proficiency.

0

Fair enough. I looked up their current role abilities on FM Scout. I might revert back to the standard Blue if the positional changes dont make much difference.

I will be interested to see how you go with set piece experiments.

0
Create an account or log in to leave a comment