Player Attributes

Z

ZaZ

Member
Joined
Nov 10, 2026
Messages
29
Reaction score
34
Player Attributes
Priority: Pace, Acceleration and Stamina are by far the most effective attributes, with a small edge to Pace. Get as much as you can.
Minimum requirements: 6-12 (depending on your league) in Work Rate, Anticipation, and Strength. These hurt a lot when below 6, but do not improve as much above 12.
Special: Jumping Reach (especially for defenders). This does not help much when below 6, but it is one of the most important attributes when above 12 (only below Pace, Acceleration, and Stamina).
Good to have: 6-12 (depending on your league) in Balance, Concentration and Natural Fitness. These hurt a bit when below 6, but not improve as much when above 12.

Example of ideal player
Pace: 12-20++
Acceleration: 12-20+
Stamina: 12-20+
Work Rate: 6-12
Anticipation: 6-20
Strength: 6-20
Jump Reach: 12-20*
Balance: 6-12
Concentration: 6-12
Natural Fitness: 6-12

+ Priority attributes, to get as high as possible
* Does not hurt much below 12, but helps a lot from 12 and up, especially for defenders.

This is the impact of each attribute when increased from 1 to 6 > 12 > 18, with a simple extrapolation for the remaining values.
1794337960059.webp

Special thanks to Harvestgreen22 for the research on influence of attributes on performance, and squirrel_plays for the table.
 
Last edited:
Stamina getting that heavy makes sense now - tired legs compound into every other check late in games. Does the 6-12 minimum band for the secondary attributes still hold in the current round of tests?
 
Re-arrangement of Harvestgreen FM27 attribute testing data to show change in win-rate if changing all players in a team from all 12s to any value for any attribute.

(Presumably changing one player would cause a one-eleventh, or perhaps one-tenth if HG didn't test GKs, change in team win rate expectation)

Screenshot 2026-11-10 at 19.03.45.webp
 
  • Like
Reactions: ZaZ
Same analysis now sorted at attribute 18. So the first one is ordered by attributes to avoid being bad at, and this sorted by attributes to seek to be good at.

In calculating these tables I have made the simplifying assmption that the tested steps from 1 to 6, 7 to 12 and 12 to 18 are linear.

Screenshot 2026-11-10 at 19.07.12.webp
 
squirrel_plays said:
Re-arrangement of Harvestgreen FM27 attribute testing data to show change in win-rate if changing all players in a team from all 12s to any value for any attribute.

(Presumably changing one player would cause a one-eleventh, or perhaps one-tenth if HG didn't test GKs, change in team win rate expectation)

View attachment 184
Click to expand...
Great stuff! Do you mind if I share it in the first post? Would also be good to have one sorted by attribute at 1, from lowest to highest (ohh, wait, that is the first).
 
I had different results for attributes importance between intense and more balanced tactics, if you go for intensity, stamina and some others are even more important, but I roughly got same results. So maybe run the test once the best tactic is found to maximise it. Same as usual for goalkeepers, reflexes and agility are 95% of a it.
 
Confused about Passing, as my own research showed it is one of most important attributes this year. Only Acceleration + Pace worked better for me. I suggest your research is more accurate than my one, however i didn`t expect to see negative numbers for Passing. Appreciate if it is possible to double check.
(Also Balance and Teamwork gave me very good increase. And standard GK attributes Aerial Reach and Reflexes didn`t up GK stats, however worked well for team indicators such as xG. Lol.)
 
squirrel_plays said:
Same analysis now sorted at attribute 18. So the first one is ordered by attributes to avoid being bad at, and this sorted by attributes to seek to be good at.

In calculating these tables I have made the simplifying assmption that the tested steps from 1 to 6, 7 to 12 and 12 to 18 are linear.

View attachment 185
Click to expand...
Maybe worthwhile to isolate if midfielders and attackers at 18 Jumping Reach are helping to impact this win rate or if it's only Defenders that have a huge impact when having higher numbers of Jumping Reach.
 
It would be great if George came back to do his in-depth research on attributes for us. George, we need you!
 
Ace Of Spades said:
It would be great if George came back to do his in-depth research on attributes for us. George, we need you!
Click to expand...
I won't be doing FM27 attribute testing as I didn't find the game good enough to buy, and HarvestGreen has already done it.

However, while FM Arena was down a lot of new info has come out from the playgm forums where people have used AI to reserve engineer the mechanics, and I have been working on an extensive new guide for FM24 (mostly applicable to FM26/FM27) that will cover all of these things. One finding is that CA plays a big role in the match engine for technical & mental attributes, which carries some quite significant implications. I think few saw the CA thing, and I don't think anyone else has realized the downstream implications I will elaborate on. Currently my guide is 15 pages long and about half done.

My own findings and methods have been largely eclipsed by the new AI methods. I'm still absorbing the findings from LightningFlik's program, but now we also have a bunch of people from the chinese forum coming out with stuff using AI as well. And then there's the GK testing I want to continue, as well as training growth testing, and finally FM24 outfield player testing and weights for FMST26. So there's just an endless amount of things to look at now, and since testing a few attributes can take up an entire day, I have to be selective.

I feel the most important things to do are:

1) Summarize the new AI-derived findings
2) Finish off the GK attribute testing
3) Create a new set of outfield player weights for FMST26 using the updated player comparison data

I had a fiddle around with the tactics in FM27 for a few hours during the beta, planning to become the Knap of FM27. Best I could do was 16th with Hull. And others have explored things like tactical effectiveness in more depth faster than I could. It seems that exploit tactics might just not be a thing in FM27, while the game itself isn't good enough IMO - not even with the reversions to FM24 in aspects, so I think I'll be leaving FM27 to others while I continue at a leisurely pace with stuff around FM24.

The FM27 database was backported to FM24, which is something I was dreaming of doing. It's not perfect (I saw a duplicated player when I took a quick look), but my impressions from a quick glance are that it looks decent enough to use. With a bit more polish, maybe it'll take away the one big downside of FM24, which having to play with the outdated database.
 
GeorgeFloydOverdosed said:
I won't be doing FM27 attribute testing as I didn't find the game good enough to buy, and HarvestGreen has already done it.

However, while FM Arena was down a lot of new info has come out from the playgm forums where people have used AI to reserve engineer the mechanics, and I have been working on an extensive new guide for FM24 (mostly applicable to FM26/FM27) that will cover all of these things. One finding is that CA plays a big role in the match engine for technical & mental attributes, which carries some quite significant implications. I think few saw the CA thing, and I don't think anyone else has realized the downstream implications I will elaborate on. Currently my guide is 15 pages long and about half done.

My own findings and methods have been largely eclipsed by the new AI methods. I'm still absorbing the findings from LightningFlik's program, but now we also have a bunch of people from the chinese forum coming out with stuff using AI as well. And then there's the GK testing I want to continue, as well as training growth testing, and finally FM24 outfield player testing and weights for FMST26. So there's just an endless amount of things to look at now, and since testing a few attributes can take up an entire day, I have to be selective.

I feel the most important things to do are:

1) Summarize the new AI-derived findings
2) Finish off the GK attribute testing
3) Create a new set of outfield player weights for FMST26 using the updated player comparison data

I had a fiddle around with the tactics in FM27 for a few hours during the beta, planning to become the Knap of FM27. Best I could do was 16th with Hull. And others have explored things like tactical effectiveness in more depth faster than I could. It seems that exploit tactics might just not be a thing in FM27, while the game itself isn't good enough IMO - not even with the reversions to FM24 in aspects, so I think I'll be leaving FM27 to others while I continue at a leisurely pace with stuff around FM24.

The FM27 database was backported to FM24, which is something I was dreaming of doing. It's not perfect (I saw a duplicated player when I took a quick look), but my impressions from a quick glance are that it looks decent enough to use. With a bit more polish, maybe it'll take away the one big downside of FM24, which having to play with the outdated database.
Click to expand...
That's genuinely exciting news, George! The discovery about CA playing a significant role in how technical and mental attributes work sounds massive, especially considering the potential implications for player evaluation and the META.

It's a shame you won't be testing FM27 directly, but I completely understand your reasoning. I have to admit, I'd be genuinely sad if you ever stopped your research altogether. I've been following your work closely for a long time, and I always look forward to your findings and discoveries.

Your upcoming guide and the updated FMST26 weights sound incredibly interesting, especially the implications that others might have overlooked. And if you're planning to release updated weights, possibly a 2.2 version, you can definitely count on me being there to follow it and test it out!

Also, if you happen to have any links to research conducted by others, whether AI-assisted or not, I'd be extremely interested! I'm particularly looking for updated attribute weights based on the FM27 META, so any reliable findings, tests, or resources you could point me towards would be greatly appreciated.

Your research has always been incredibly valuable to the community, and I'm definitely one of those who really appreciate everything you do. Thank you for all the time and effort you put into it, George!
 
You must log in or register to reply here.
Back
Top