"Shoots With Power" trait could be the real deal.

by Zippo, Sep 28, 2020

Hey guys,

I've made a small test to see how "Shoots With Power" trait works.

I edited the "Finishing" and "Long Shots" attributes of Mane and Salah and set the "Finishing" to "1" and "Long Shots" to "20".




Also, I edited their Player Traits and gave them "Shoots With Power" trait




They started to score like crazy after that :)





So if someone of your attacker's got low Finishing attribute then "Shoots With Power" could be the real deal for him.

+2

@Zippo thanks for the tip!

After I applied the trait to Rashford, he scored a hat-trick in the next match :)

0

Yeah, that makes sense. Saw one of the bust the net videos not too keen on the trait, but to me it makes logical sense that if you can't finish well and also lack composure in front of goal, it's not a bad idea to just put your foot through it :)

0

I've tried it on my young striker and it worked like a charm.

Thank you! :thup:

0

This PPM has resurrected my dead striker... thanks! :D

0

It seems to be working very well for my striker. The goals has been increasing since I trained him that PPM.

0

How long does it usually take to train it?

0

I did a bunch of PPM testing for strikers one time, and I thought I'd share my results. This is a little long, so tl;dr almost nothing mattered. Sorry to disappoint. Just a disclaimer -- I only did the test with one striker, and it's possible that the PPMs interact with player traits. As I'll explain, if there is such an interaction, I believe it had no effect on my test.

So I had a newgen striker that was amazing, Michael Baroni. Very composed, very fast, good movement. Basically he was good at just about everything, maybe CA 180 or even greater. But he wasn't scoring like I thought he should. It was the usual story -- lots of one-on-ones fired right into the keeper. After reading all the manuals and forums, I thought he would do well with "places shots," and trained him for that. Afterwards, I was second guessing that decision, so hence the motivation for my test.

What I did was fire up a new save in Ligue 1, which has 20 teams. I created 20 clones of Michael Baroni, and put one on every team in Ligue 1. (As an aside, it was interesting to see the salary that the game gave to each clone. It ranged from $22mm/yr for PSG down to $1mm/yr for the poor teams.)  I then randomly assigned finishing and movement traits to each clone. I tried four finishing traits -- places shots, shoots with power, lobs keeper, and rounds keeper. I also interacted those with five other traits for movement: no movement trait, beats offside trap, moves into channels, runs with ball often. I was also curious about the "curls ball" trait, so I tried that too. So that's 20 possible combinations: 4 finishing X 5 other traits (4 movement +curls ball). If I had it to do over again, I would have had one test Baroni with no finishing trait, rather than one Baroni with no movement trait. But "lobs keeper" I figured would be basically useless, so would serve as a baseline for comparison for the other finishing traits.

I then simulated the same season four times, rotating which Baroni gets which trait so that it wasn't always a crappy team getting the "shoots with power" trait or whatever. So if you're counting, that's 80 Baroni seasons, with 20 for each finishing trait. I was the Amiens coach, but on vacation mode. The rest I left to the AI. Baroni was good enough to be the first-choice starter on all teams, including PSG. In the first season, I made the mistake of not adjusting injury susceptibility -- it was around 13 or so, and injuries therefore added some noise to the simulation. In subsequent simulations, I set it to 1 for all the Baronis. It was surprising how much time was lost to injury even for the proneness = 1 Baronis: league games missed for the average injury-proned Baroni was 6.5, while for the average non-injury proned Baroni it was 3.68.

On to the results. The outcome variable was goals/90 minutes. Here are the means across the finishing traits:
Places: 0.45
Power: 0.45
Rounds: 0.48
Lobs: 0.49

If you are worried that the distribution of g90 is skewed, here are the medians:
Places: 0.44
Power: 0.47
Rounds: 0.49
Lobs: 0.47

Long story short, there is almost no difference, nowhere near statistically significant, either taken individually (T-test) or collectively (F-test). I then tested a few other hypotheses that I had, and nothing mattered. For instance, I thought that a player on a good team might face more bunkering and would have fewer one-on-ones. So I looked for whether the impact of the trait differed by media prediction. No dice.

What about movement (and curls ball)? Again, here is average g90:
No trait: 0.42
Beats offside trap: 0.44
Moves into channels: 0.53
Runs with ball: 0.48
Curls ball: 0.48

Not exactly inspiring, but "moves into channels" is in fact statistically significant compared to having no movement traits at all. This is the only trait that makes a difference statistically. I mentioned earlier that I dismissed that the PPMs was interacting with attributes. It's possible that such interactions matter, I do not dispute that. But this is a striker who is good at everything, so such interactions are unlikely to matter for this particular exercise.

+4

runitout said: I did a bunch of PPM testing for strikers one time, and I thought I'd share my results. This is a little long, so tl;dr almost nothing mattered. Sorry to disappoint. Just a disclaimer -- I only did the test with one striker, and it's possible that the PPMs interact with player traits. As I'll explain, if there is such an interaction, I believe it had no effect on my test.

So I had a newgen striker that was amazing, Michael Baroni. Very composed, very fast, good movement. Basically he was good at just about everything, maybe CA 180 or even greater. But he wasn't scoring like I thought he should. It was the usual story -- lots of one-on-ones fired right into the keeper. After reading all the manuals and forums, I thought he would do well with "places shots," and trained him for that. Afterwards, I was second guessing that decision, so hence the motivation for my test.

What I did was fire up a new save in Ligue 1, which has 20 teams. I created 20 clones of Michael Baroni, and put one on every team in Ligue 1. (As an aside, it was interesting to see the salary that the game gave to each clone. It ranged from $22mm/yr for PSG down to $1mm/yr for the poor teams.)  I then randomly assigned finishing and movement traits to each clone. I tried four finishing traits -- places shots, shoots with power, lobs keeper, and rounds keeper. I also interacted those with five other traits for movement: no movement trait, beats offside trap, moves into channels, runs with ball often. I was also curious about the "curls ball" trait, so I tried that too. So that's 20 possible combinations: 4 finishing X 5 other traits (4 movement +curls ball). If I had it to do over again, I would have had one test Baroni with no finishing trait, rather than one Baroni with no movement trait. But "lobs keeper" I figured would be basically useless, so would serve as a baseline for comparison for the other finishing traits.

I then simulated the same season four times, rotating which Baroni gets which trait so that it wasn't always a crappy team getting the "shoots with power" trait or whatever. So if you're counting, that's 80 Baroni seasons, with 20 for each finishing trait. I was the Amiens coach, but on vacation mode. The rest I left to the AI. Baroni was good enough to be the first-choice starter on all teams, including PSG. In the first season, I made the mistake of not adjusting injury susceptibility -- it was around 13 or so, and injuries therefore added some noise to the simulation. In subsequent simulations, I set it to 1 for all the Baronis. It was surprising how much time was lost to injury even for the proneness = 1 Baronis: league games missed for the average injury-proned Baroni was 6.5, while for the average non-injury proned Baroni it was 3.68.

On to the results. The outcome variable was goals/90 minutes. Here are the means across the finishing traits:
Places: 0.45
Power: 0.45
Rounds: 0.48
Lobs: 0.49

If you are worried that the distribution of g90 is skewed, here are the medians:
Places: 0.44
Power: 0.47
Rounds: 0.49
Lobs: 0.47

Long story short, there is almost no difference, nowhere near statistically significant, either taken individually (T-test) or collectively (F-test). I then tested a few other hypotheses that I had, and nothing mattered. For instance, I thought that a player on a good team might face more bunkering and would have fewer one-on-ones. So I looked for whether the impact of the trait differed by media prediction. No dice.

What about movement (and curls ball)? Again, here is average g90:
No trait: 0.42
Beats offside trap: 0.44
Moves into channels: 0.53
Runs with ball: 0.48
Curls ball: 0.48

Not exactly inspiring, but "moves into channels" is in fact statistically significant compared to having no movement traits at all. This is the only trait that makes a difference statistically. I mentioned earlier that I dismissed that the PPMs was interacting with attributes. It's possible that such interactions matter, I do not dispute that. But this is a striker who is good at everything, so such interactions are unlikely to matter for this particular exercise.


Need that someone tested it in FM21 :D

0

If you're curious, here's Baroni:

0

Bogeyman said: Need that someone tested it in FM21 :D

Yeah, all that work, now it's obsolete. For my next test, the manager will have his hands in his pockets or hold them out.

0

One thing I wanted to add was that the "moves into channels" result was particularly surprising to me. My intuition was that trait would be valuable for creating goal opportunities for others more so than oneself by drawing out defenders creating space for other attacking players to exploit. If anything, I thought it would reduce goal scoring for the player with the trait by taking him further from goal. Guess not.

0

runitout said: One thing I wanted to add was that the "moves into channels" result was particularly surprising to me. My intuition was that trait would be valuable for creating goal opportunities for others more so than oneself by drawing out defenders creating space for other attacking players to exploit. If anything, I thought it would reduce goal scoring for the player with the trait by taking him further from goal. Guess not.

Almost all the attacking roles in the game have "Moves Into Channels" PI by default... I think it proves that this PI is a must have PI for attacking roles :)

0

runitout said: I did a bunch of PPM testing for strikers one time, and I thought I'd share my results. This is a little long, so tl;dr almost nothing mattered. Sorry to disappoint. Just a disclaimer -- I only did the test with one striker, and it's possible that the PPMs interact with player traits. As I'll explain, if there is such an interaction, I believe it had no effect on my test.

So I had a newgen striker that was amazing, Michael Baroni. Very composed, very fast, good movement. Basically he was good at just about everything, maybe CA 180 or even greater. But he wasn't scoring like I thought he should. It was the usual story -- lots of one-on-ones fired right into the keeper. After reading all the manuals and forums, I thought he would do well with "places shots," and trained him for that. Afterwards, I was second guessing that decision, so hence the motivation for my test.

What I did was fire up a new save in Ligue 1, which has 20 teams. I created 20 clones of Michael Baroni, and put one on every team in Ligue 1. (As an aside, it was interesting to see the salary that the game gave to each clone. It ranged from $22mm/yr for PSG down to $1mm/yr for the poor teams.)  I then randomly assigned finishing and movement traits to each clone. I tried four finishing traits -- places shots, shoots with power, lobs keeper, and rounds keeper. I also interacted those with five other traits for movement: no movement trait, beats offside trap, moves into channels, runs with ball often. I was also curious about the "curls ball" trait, so I tried that too. So that's 20 possible combinations: 4 finishing X 5 other traits (4 movement +curls ball). If I had it to do over again, I would have had one test Baroni with no finishing trait, rather than one Baroni with no movement trait. But "lobs keeper" I figured would be basically useless, so would serve as a baseline for comparison for the other finishing traits.

I then simulated the same season four times, rotating which Baroni gets which trait so that it wasn't always a crappy team getting the "shoots with power" trait or whatever. So if you're counting, that's 80 Baroni seasons, with 20 for each finishing trait. I was the Amiens coach, but on vacation mode. The rest I left to the AI. Baroni was good enough to be the first-choice starter on all teams, including PSG. In the first season, I made the mistake of not adjusting injury susceptibility -- it was around 13 or so, and injuries therefore added some noise to the simulation. In subsequent simulations, I set it to 1 for all the Baronis. It was surprising how much time was lost to injury even for the proneness = 1 Baronis: league games missed for the average injury-proned Baroni was 6.5, while for the average non-injury proned Baroni it was 3.68.

On to the results. The outcome variable was goals/90 minutes. Here are the means across the finishing traits:
Places: 0.45
Power: 0.45
Rounds: 0.48
Lobs: 0.49

If you are worried that the distribution of g90 is skewed, here are the medians:
Places: 0.44
Power: 0.47
Rounds: 0.49
Lobs: 0.47

Long story short, there is almost no difference, nowhere near statistically significant, either taken individually (T-test) or collectively (F-test). I then tested a few other hypotheses that I had, and nothing mattered. For instance, I thought that a player on a good team might face more bunkering and would have fewer one-on-ones. So I looked for whether the impact of the trait differed by media prediction. No dice.

What about movement (and curls ball)? Again, here is average g90:
No trait: 0.42
Beats offside trap: 0.44
Moves into channels: 0.53
Runs with ball: 0.48
Curls ball: 0.48

Not exactly inspiring, but "moves into channels" is in fact statistically significant compared to having no movement traits at all. This is the only trait that makes a difference statistically. I mentioned earlier that I dismissed that the PPMs was interacting with attributes. It's possible that such interactions matter, I do not dispute that. But this is a striker who is good at everything, so such interactions are unlikely to matter for this particular exercise.


:)

0

Tsubasa said: :)



Says the guy on the gaming forum. :)

I'll admit that I had as much or more fun with this than I usually do actually playing the game.

0