FM27 match engine: what we know so far

I also did a test to learn about home-advantage and crowds. This is a follow up from a point from @Purity in another thread. I tested home advantage on 86 different grounds, with identical sides, so the only difference is where the match is played. These are reported in study 15b
  • Overall home advantage is +0.22 xG and +0.37 points per match, same as my first test.
  • Big stadiums give more than twice as much. Grounds of 40,000 or more give +0.39 xG, smaller ones +0.15 to +0.18, with no trend from 4,000 to 40,000. So it's more like a big stadium bonus than a straight line.
  • Pitch size makes no difference.
  • The same stadium also varies from match to match more than chance allows, so something about the match itself matters too. Crowd on the day is a candidate, but so are weather and the competition.
The problem is I can't separate crowd, stadium and club here: big stadiums belong to big clubs, and I haven't found an attendance field in the match engine that I can change.
 
Once again caio, it's amazing, clearly.

The difference with Purity stay in the variety of tactics in the experiment it's OK ?

I mean, it's more accurate with your test, for the same positions ?
 
On Fitness, positions and match conditions.
Match sharpness and condition are the biggest effects I've measured so far, bigger than anything you can change on the tactics screen. Home advantage is real, morale at kick-off doesn't seem to matter. I think these are already know but its good to document and test them on the same tool.
The table below summarizes the main numbers, in points per game compared with the same matches where everyone is as fresh and sharp as saved.
ChangeEffect
Match sharpness 50% vs 100%−1.03
Match sharpness 75% vs 100%−0.37
Condition 80% vs 100%−0.80
Condition 90% vs 100%−0.28
Every outfield starter at 10 in his own positions−0.10 xPts
Every position rating at 10 (his own and all others)−0.30 xPts
Goalkeeper at 10 in goal−0.53, conceding 0.7 more goals
Team cohesion, lowest to highest+0.2 to +0.3
Playing at home (identical sides)1.54 against 1.18 points
Morale 5 to 20 at kick-offno measurable effect

Here are the questions I think people care about:
Should I rotate my players?
I think the results support that. Starting players at 75% sharpness or 90% condition costs about a third of a point a game, and at 50% sharpness or 80% condition close to a full point. Nothing I tested on the tactics screen comes close to that. These are reported in study 5.

How bad is playing someone out of position?
In my first tests, it seemed that position ratings at 10 cost 0.85 points a game, but a follow-up showed most of that comes from the ratings of the positions players are not playing. With only their own positions at 10, all outfield starters together cost about 0.1 a game. The exception is the goalkeeper, his position rating is basically his shot-stopping. These are reported in study 17

Does an unfamiliar out-of-possession position cost more than an in-possession one?
Neither costs anything I could measure, for the three or four players whose positions change between phases. These are reported in study 17

How big is home advantage?
About the size on real football, which I think is a good thing. With identical squads and tactics, the home side gets +0.21 xG, +0.28 goals and +1.3 shots a game. It varies by ground (different matches - one of twelve gave about +0.5 xG and one gave none) but not with the club's level. These are reported in study 15

Does morale matter?
Not at kick-off. Morale 5, 10 and 20 against the usual 15 are all within ±0.04 a game. It may still matter over a season through other things (training, form, player happiness), which the tool can't see. These are reported in study 15

Note that sharpness, condition and cohesion were tested on real matches from my own save, the rest on neutral ground. I am moving towards the tests on neutral ground since I don't want they to rely on my real matches going forward.
''How bad is playing someone out of position?
In my first tests, it seemed that position ratings at 10 cost 0.85 points a game, but a follow-up showed most of that comes from the ratings of the positions players are not playing. With only their own positions at 10, all outfield starters together cost about 0.1 a game. The exception is the goalkeeper, his position rating is basically his shot-stopping. These are reported in study 17''
I read the study and is this to say players who aren't competent at multiple other positions are worse e.g. a winger who can't play cam costs more points per game than a winger who is also competent centrally?
 
Once again caio, it's amazing, clearly.

The difference with Purity stay in the variety of tactics in the experiment it's OK ?

I mean, it's more accurate with your test, for the same positions ?
I think our results are pretty close. The actual numbers will change depending on the match, the squad, and the tactics being used in the tests. I'm redoing my analyses on a Catenaccio pre set, and the results are pretty much the same. Catenaccio's centre-backs gain less from Pace, Acceleration and Jumping Reach than Gegenpress's, probably because they defend deep with three of them. Its central attacking midfielders gain more from Pace, but otherwise they are very similar.
 
''How bad is playing someone out of position?
In my first tests, it seemed that position ratings at 10 cost 0.85 points a game, but a follow-up showed most of that comes from the ratings of the positions players are not playing. With only their own positions at 10, all outfield starters together cost about 0.1 a game. The exception is the goalkeeper, his position rating is basically his shot-stopping. These are reported in study 17''
I read the study and is this to say players who aren't competent at multiple other positions are worse e.g. a winger who can't play cam costs more points per game than a winger who is also competent centrally?
Kind of, but it's small per player. In the test I capped every position rating at 10 for all ten outfield starters at once. That cost about 0.30 xPts a game, against 0.10 when I only capped their own positions. So the other positions mattered, but that's ten players bad at everything at the same time, which never happens. A single winger who can't play AMC probably costs very little.My guess is the engine uses the rating for the area a player is in, so it matters more for a winger who cuts inside. I'd use it as a tie-breaker between similar players. I can test the single-player version if people are interested.
 
Thank you for your work.

What about player's traits? Are there an impact on the result? Is attributes weight changes when a traits is active? Or traits must have some required attributes to work? Maybe they're are worse if required attributes aren't matched? And it can depends of the position!
Then, it's seems very hard to test...
I can't imagine some attributes are totally useless. My guess is playmaking attributes maybe works only well with a passing traits?
Thanks! Not tested yet, but the engine work I'm reporting in another thread gives a first answer. Traits are switches inside the player's decisions: "Tries Killer Balls Often" changes which teammates he looks for, "Runs With Ball Often" is checked when he decides to dribble, and so on. So a trait changes what a player tries, not how well he does it. The attributes still decide whether it works.That makes your idea plausible: a passing trait could make Passing and Vision matter more, because the player attempts more of the passes where they count. The test is doable since the tool can set traits: trait on/off, with high and low Passing/Vision, by position. I can try adding it to the list of things to check.
 
amazing work. I appreciate what youve done here!

Im interested in your team instruction findings. Have you tried a clean slate style, while using batched instructions that define styles of play, such as pairing each tempo and directness, then testing those results with each level of patience?
Thanks! Not yet. All my instruction tests changed one instruction at a time on a tactic that already works (Stormbringer from Knap since I was using it on my own save), and on that basis most single instructions did nothing measurable. Starting from a clean slate and testing whole styles makes a lot more sense, since instructions probably work together more than alone. I want to do that when I have free cores (the other tests and the game (my save) are using them). But it would be cool to learn what works with what from scratch or come up with a new tactic using a gradient method.
 
I also did a test to learn about home-advantage and crowds. This is a follow up from a point from @Purity in another thread. I tested home advantage on 86 different grounds, with identical sides, so the only difference is where the match is played. These are reported in study 15b
  • Overall home advantage is +0.22 xG and +0.37 points per match, same as my first test.
  • Big stadiums give more than twice as much. Grounds of 40,000 or more give +0.39 xG, smaller ones +0.15 to +0.18, with no trend from 4,000 to 40,000. So it's more like a big stadium bonus than a straight line.
  • Pitch size makes no difference.
  • The same stadium also varies from match to match more than chance allows, so something about the match itself matters too. Crowd on the day is a candidate, but so are weather and the competition.
The problem is I can't separate crowd, stadium and club here: big stadiums belong to big clubs, and I haven't found an attendance field in the match engine that I can change.
I really wish FM Arena hadn't been wiped recently; I experimented with this on FM 24 and laid out all my findings in its own thread. I made a tool which let me play a fixture over and over and I used that to take a real in-game fixture, then swap out teams 1 and 2 (so they're playing at a neutral venue), then messing with max capacity vs actual capacity. The main conclusion was that the more fans you had rooting for you (home or away), the bigger your boost.
 
I really wish FM Arena hadn't been wiped recently; I experimented with this on FM 24 and laid out all my findings in its own thread. I made a tool which let me play a fixture over and over and I used that to take a real in-game fixture, then swap out teams 1 and 2 (so they're playing at a neutral venue), then messing with max capacity vs actual capacity. The main conclusion was that the more fans you had rooting for you (home or away), the bigger your boost.
You can access quite a bit via waybackmachine and repost if you'd like to.

You can find a post of yours on home ground advantage here (#864) and there are some other notable posts on this page:


Some stuff has unfortunately been lost. But I was also able to retrieve some stuff from Google's AI. Notably the full content of your 'Some training and development conventional wisdom for fm24' thread. If you want I could send it to you.

I found that with attendance an away fan is worth 25% of a home fan.

I also calculated that +1 home fan increases one of your outfield player's pace by ~0.0035 in FM24, effectively. It's a rough figure, just find it fun to work out the equivalents.
 
Pace and Acceleration are worth about twice as much per point as anything else, and most technical attributes are worth close to nothing inside a match. The table below summarizes the main learnings. It reports the value of one point change in a given attribute (holding the others constant) for the whole squad on expected points (xPts) per game. The report (Fm-Match-Lab/docs/studies/01-attribute-values.md at main · caiograco289/Fm-Match-Lab) also describes the effect on shots, Xg for and against. This was done on neutral ground,


AttributePer pointShape from 1 to 20
Acceleration+0.0680.082 below 11, 0.049 above
Pace+0.0620.073 below 13, 0.038 above
Jumping Reach+0.0350.019 up to 9, 0.047 above: pays at the top
Stamina+0.034steep below 4 (0.092), then 0.026
Strength+0.009 in xPts, but about +0.036 in real pointsworks through goals, not chances (see FAQ)
Anticipation+0.024straight line
Work Rate+0.022a floor: 1 costs half a point a game, 3-6 a little, 9-20 nothing
Concentration+0.021straight line
Balance+0.017straight line
Dribbling+0.017straight line
Decisions+0.005about nothing at every level
Passing+0.003about nothing at every level, even at 1
Finishing0.000 in xPtsconverts chances: +0.3 goals a game from 1 to 20
Not in the table, from the +5 tests on real matches: Composure and Vision show a little; Technique, First Touch, Crossing, Off the Ball, Tackling, Marking and Positioning show nothing clear; Natural Fitness is exactly zero inside a match.

I also did a bunch of additional tests that are interesting. Let me try to summarize them in a questions and answer format:

Is every point worth the same? Is 16 to 20 bigger than 8 to 12?
For most attributes, yes, each point is worth about the same, with no threshold. Four exceptions: Pace and Acceleration are worth more below about 12 than above. Jumping Reach is the opposite, flat up to 9 and then the steepest gain of any attribute (12 to 18 is worth +0.34 a game). Work Rate only hurts below about 9. Stamina hurts most below 4. There are reported in study 12

Do physical attributes only win because tests use meta tactics?
No. Physical +5 (Pace, Acceleration, Stamina, Work Rate, Aggression) is worth +0.38 to +0.48 a game under all five styles tested, possession styles included. Passing attributes help direct styles (Direct Counter-Attack +0.15, Gegenpress +0.11) and do nothing for Tiki-Taka or Control Possession, whose players already complete 90-92% of passes. Stamina pays at every intensity, most with a high press. These are reported in study 9

Do attributes combine, so a group is worth more than its parts?
No. Groups of four (playmaking, ball carrying, defending, aerial) raised together are worth exactly what their members are worth one at a time. Each group's value comes from one or two members: Jumping Reach is the whole aerial group, Anticipation the whole defending group. These are in study 10

Are weights different by position? Is Pace worth more for a winger than a midfielder?
I couldn't find any statistical differences. Giving +5 to one pair of players at a time (centre-backs, full-backs, DMs, wide AMs, strikers) gives about 0.03-0.06 a game for all positions. There are some small differences but statistically it could be just from chance. I didn't test all attributes and this is something I want to revisit at some point, but so far the conclusion is that there are no clear differences. These are reported in study 13

Is one player worth one-eleventh of the whole team?
This surprised me and the answer is not always. The results show that +5 Jumping Reach on just three players gives two-thirds of the whole team's gain. The first players you improve seems to matter most. These are reported in study 13.

Do attribute values change for a strong or weak team?
It doesn't seem so (at least among premier league level teams). The ranking is the same for top, middle and bottom sides, favourite or underdog (rank correlation 0.84). An underdog may gain a little less overall, mostly at the back, but that this is not statistically different. These are reported in study 14

Why do Strength and Finishing look weak in xG-based tables?
It seems the answer is that they act on conversion, which xG don't measure. From 1 to 20, Strength adds 0.5 goals a game above xG and +0.69 real points. It also hurt a lot below 6. Finishing adds about 0.3 goals. These are reported in study 12

Which hidden attributes matter? Is it only Consistency?
It doesn't seem so. Consistency matters across matches in a way that is not capture by the ME. In the ME, it barely matters and isn't even read (need to understand why) in many matches. Pressure is the one with a clear effect. A a squad at Pressure 1 loses 0.15 a game (about 6 points a season). Important Matches doesn't seem to be worth more on European games. Dirtiness and Temperament change cards but not results. These are reported on study 11

Does Current Ability play a role in the match?
No. With every attribute fixed, a squad at CA 40 does as well as one at CA 200, and Passing or Decisions +5 is worth the same at CA 60 as at 180. CA is read during the match, but it doesn't seem to change chances or results. Better players matter through their attributes. These are reported in study 20

What about goalkeepers?
This surprised me and I think I need to revist this since it goes against what we previously knew. In my tests, no single goalkeeper attribute +5 has a clear effect alone (Positioning and Concentration lead). All of them +5 together raise the save rate from 75% to 80%, about +0.26 points a game. The keeper's position rating matters far more: at 10 his save rate drops to 60%, conceding 0.7 more goals a game. xG-based tables miss keepers entirely, because saves change goals, not xG. These are reported in study 21

Why do these numbers differ from HarvestGreen's?
As far as I know we use the same engine (lqf0624's tool), with different setups. The order at the top agrees. On the Windows tool's sample match with everyone at 12, most effects shrink to about 40% of their size on real matches and some vanish. I think it has to do with the type of match it was trained on. I did most of my simulations on real matches (loaded from my save) and then modified. On these real matches, even levelled to 12, mental attributes like Anticipation and Concentration clearly count. These are reported in study 2

There are a lot of caveats and different ways to have done these. So let me know if you spot any issues or any follow ups that might be useful.
Cool research there, especially using real teams to collect data. Is it possible to create a table based on the change in actual points per match (which I consider more valuable than xP)? I would recommend something like this table here made by the squirel guy, with data from Harvestgreen, because I believe something with a similar design would be the best for the overall public.
1794742298142.webp
 
Cool research there, especially using real teams to collect data. Is it possible to create a table based on the change in actual points per match (which I consider more valuable than xP)? I would recommend something like this table here made by the squirel guy, with data from Harvestgreen, because I believe something with a similar design would be the best for the overall public.
Here you go, same format as squirrel's table so they can be compared directly: change in win rate when the whole team goes from 12 to any value in one attribute. It is a linear extrapolation from the tests between 1→6, 7→12 and 12→18.

1794744632127.webp


And here in text format so its easier to copy and paste:

Sorted by value at 18 (attributes to seek to be good at)

Attribute1→67→1212→18123456789101112131415161718
Jumping Reach6.3%1.5%16.9%-7.9%-6.6%-5.3%-4.1%-2.8%-1.5%-1.3%-1.0%-0.8%-0.5%-0.3%0.0%2.8%5.6%8.5%11.3%14.1%16.9%
Pace13.2%9.5%16.6%-22.6%-20.0%-17.4%-14.7%-12.1%-9.5%-7.9%-6.3%-4.7%-3.2%-1.6%0.0%2.8%5.5%8.3%11.1%13.8%16.6%
Acceleration13.3%17.1%11.6%-30.4%-27.7%-25.1%-22.4%-19.8%-17.1%-14.3%-11.4%-8.6%-5.7%-2.9%0.0%1.9%3.9%5.8%7.7%9.7%11.6%
Stamina17.1%0.5%8.8%-17.6%-14.2%-10.8%-7.3%-3.9%-0.5%-0.4%-0.3%-0.2%-0.2%-0.1%0.0%1.5%2.9%4.4%5.8%7.3%8.8%
Anticipation11.9%-0.7%6.9%-11.2%-8.8%-6.4%-4.1%-1.7%0.7%0.6%0.5%0.4%0.2%0.1%0.0%1.1%2.3%3.4%4.6%5.7%6.9%
Concentration6.7%4.1%5.7%-10.8%-9.5%-8.1%-6.8%-5.5%-4.1%-3.4%-2.7%-2.1%-1.4%-0.7%0.0%0.9%1.9%2.8%3.8%4.7%5.7%
Dribbling3.0%0.0%3.3%-3.0%-2.4%-1.8%-1.2%-0.6%0.0%0.0%0.0%0.0%0.0%0.0%0.0%0.6%1.1%1.7%2.2%2.8%3.3%
Strength16.3%6.0%1.2%-22.3%-19.1%-15.8%-12.5%-9.3%-6.0%-5.0%-4.0%-3.0%-2.0%-1.0%0.0%0.2%0.4%0.6%0.8%1.0%1.2%
Work Rate20.8%-1.2%1.0%-19.6%-15.4%-11.3%-7.1%-2.9%1.2%1.0%0.8%0.6%0.4%0.2%0.0%0.2%0.3%0.5%0.7%0.8%1.0%
Decisions-1.8%6.0%0.1%-4.1%-4.5%-4.9%-5.2%-5.6%-6.0%-5.0%-4.0%-3.0%-2.0%-1.0%0.0%0.0%0.0%0.1%0.1%0.1%0.1%
Passing0.5%3.5%-1.1%-4.0%-3.9%-3.8%-3.7%-3.6%-3.5%-2.9%-2.3%-1.7%-1.2%-0.6%0.0%-0.2%-0.4%-0.5%-0.7%-0.9%-1.1%
Finishing0.3%8.3%-2.0%-8.6%-8.6%-8.5%-8.5%-8.4%-8.3%-6.9%-5.6%-4.2%-2.8%-1.4%0.0%-0.3%-0.7%-1.0%-1.3%-1.7%-2.0%
Balance5.1%9.9%-2.3%-15.0%-14.0%-13.0%-12.0%-10.9%-9.9%-8.2%-6.6%-4.9%-3.3%-1.6%0.0%-0.4%-0.8%-1.1%-1.5%-1.9%-2.3%

Anyway, I think it is worth re-doing this with larger samples since these small jumps are hard to detect statistically.

Compared to squirrel's table the shape is very similar, the same attributes at the top, but our drops at the bottom are smaller (Pace at 1 is -23% here vs -40% in his). Most likely because his base is a team with every attribute at 12, while mine is a normal mid-PL squad with only that one attribute changed, so the rest of the team covers more.
 
Also, worth pointing out that @Purity research and mine copied below show differences in values across positions. I didn't run 1–20 by position. What I have is a 50k-match test where every attribute of every starter moves randomly ±5 each match, and one regression pulls out what each attribute is worth in each position. I put it on the same scale as the tables above. i.e. change in win rate for going 12→18, but for one player in that position (assuming it's linear, which is all this test can say).

Change in win rate for +6 levels (12→18) in one player of that position (Gegenpress, 50k matches; * = confirmed in both random halves of the data). "Whole team" = all 10 outfield players +6 at once, so it lines up with the 12→18 column above.

AttributeWhole teamSTAMDMFBCB
Acceleration11.6%2.0%*0.6%0.9%*0.9%*1.4%*
Pace10.8%1.5%*1.0%*0.9%*1.2%*0.8%*
Jumping Reach7.4%1.1%*0.8%*0.3%0.4%1.5%*
Strength5.8%0.7%0.8%*0.5%0.5%0.4%
Stamina5.0%0.5%0.7%0.3%0.7%*0.4%
Anticipation4.9%0.5%0.2%0.5%0.6%*0.5%
Balance3.9%0.8%0.0%0.2%0.3%0.8%*
Concentration3.9%0.4%0.4%0.1%0.6%0.5%
Dribbling3.2%0.3%0.4%0.3%0.5%0.2%
Finishing2.5%0.8%*0.2%0.3%0.0%0.0%
Goalkeeper (change in win rate, keeper +6 levels)

AttributeWin rate
Agility2.8%*
Reflexes2.2%*
Acceleration0.9%
Aerial Reach0.9%

One player is worth roughly a tenth of the whole team, like squirrel guessed. One player +6 is only 1–2% win rate, so each cell here is about ±0.2%. The "Whole team" column is the more precise check on the table above: Acceleration comes out at exactly the same 11.6%; Pace and Jumping Reach are lower here (11% and 7% vs 17%) because this test only moves players ±5 around a normal squad, where those curves are flatter. The win rate tables show the shape (floors, and where it gets steep); this one shows where in the team it matters.

Two fun ones from the position table:
  • Keepers: Agility and Reflexes are the biggest single-player effects in the whole table (2–3% win rate for +6), but basically nothing in xPts. The keeper saves more shots without changing the xG he faces, so xPts can't see it at all.
  • Finishing for strikers shows up in win rate (+0.8%, confirmed) but not in xPts, which is what you'd expect if it converts chances without creating more. Jumping Reach is a striker's and a centre-back's attribute, much less a midfielder's.
 
Here you go, same format as squirrel's table so they can be compared directly: change in win rate when the whole team goes from 12 to any value in one attribute. It is a linear extrapolation from the tests between 1→6, 7→12 and 12→18.

View attachment 785

And here in text format so its easier to copy and paste:

Sorted by value at 18 (attributes to seek to be good at)

Attribute1→67→1212→18123456789101112131415161718
Jumping Reach6.3%1.5%16.9%-7.9%-6.6%-5.3%-4.1%-2.8%-1.5%-1.3%-1.0%-0.8%-0.5%-0.3%0.0%2.8%5.6%8.5%11.3%14.1%16.9%
Pace13.2%9.5%16.6%-22.6%-20.0%-17.4%-14.7%-12.1%-9.5%-7.9%-6.3%-4.7%-3.2%-1.6%0.0%2.8%5.5%8.3%11.1%13.8%16.6%
Acceleration13.3%17.1%11.6%-30.4%-27.7%-25.1%-22.4%-19.8%-17.1%-14.3%-11.4%-8.6%-5.7%-2.9%0.0%1.9%3.9%5.8%7.7%9.7%11.6%
Stamina17.1%0.5%8.8%-17.6%-14.2%-10.8%-7.3%-3.9%-0.5%-0.4%-0.3%-0.2%-0.2%-0.1%0.0%1.5%2.9%4.4%5.8%7.3%8.8%
Anticipation11.9%-0.7%6.9%-11.2%-8.8%-6.4%-4.1%-1.7%0.7%0.6%0.5%0.4%0.2%0.1%0.0%1.1%2.3%3.4%4.6%5.7%6.9%
Concentration6.7%4.1%5.7%-10.8%-9.5%-8.1%-6.8%-5.5%-4.1%-3.4%-2.7%-2.1%-1.4%-0.7%0.0%0.9%1.9%2.8%3.8%4.7%5.7%
Dribbling3.0%0.0%3.3%-3.0%-2.4%-1.8%-1.2%-0.6%0.0%0.0%0.0%0.0%0.0%0.0%0.0%0.6%1.1%1.7%2.2%2.8%3.3%
Strength16.3%6.0%1.2%-22.3%-19.1%-15.8%-12.5%-9.3%-6.0%-5.0%-4.0%-3.0%-2.0%-1.0%0.0%0.2%0.4%0.6%0.8%1.0%1.2%
Work Rate20.8%-1.2%1.0%-19.6%-15.4%-11.3%-7.1%-2.9%1.2%1.0%0.8%0.6%0.4%0.2%0.0%0.2%0.3%0.5%0.7%0.8%1.0%
Decisions-1.8%6.0%0.1%-4.1%-4.5%-4.9%-5.2%-5.6%-6.0%-5.0%-4.0%-3.0%-2.0%-1.0%0.0%0.0%0.0%0.1%0.1%0.1%0.1%
Passing0.5%3.5%-1.1%-4.0%-3.9%-3.8%-3.7%-3.6%-3.5%-2.9%-2.3%-1.7%-1.2%-0.6%0.0%-0.2%-0.4%-0.5%-0.7%-0.9%-1.1%
Finishing0.3%8.3%-2.0%-8.6%-8.6%-8.5%-8.5%-8.4%-8.3%-6.9%-5.6%-4.2%-2.8%-1.4%0.0%-0.3%-0.7%-1.0%-1.3%-1.7%-2.0%
Balance5.1%9.9%-2.3%-15.0%-14.0%-13.0%-12.0%-10.9%-9.9%-8.2%-6.6%-4.9%-3.3%-1.6%0.0%-0.4%-0.8%-1.1%-1.5%-1.9%-2.3%

Anyway, I think it is worth re-doing this with larger samples since these small jumps are hard to detect statistically.

Compared to squirrel's table the shape is very similar, the same attributes at the top, but our drops at the bottom are smaller (Pace at 1 is -23% here vs -40% in his). Most likely because his base is a team with every attribute at 12, while mine is a normal mid-PL squad with only that one attribute changed, so the rest of the team covers more.
Great, thanks! Then I will wait for you to finish it with larger samples to add to my post about attributes (if you allow), and update my recommendations. =)
 
I'm struggling to keep up with all the new information, so apologies if I've missed something about this, but I think you found that dirtiness didn't impact results, only changed the numbers of cards (so a post-match ban or substitution effect only I suppose).

I was looking at some posts I made not long before the forum went down, and one of them was a finding using simulation that dirtiness doesn't kick in until condition is low. I would have had limited samples at that point and it was for FM24 though, but if you're testing using players with 100% condition, it could be something to look at again. I've previously found in manual testing, and I've seen at least one or two other testers find the same thing, that dirtiness is pretty high up in its effect.

I just thought I'd make some comments on some other things too.

You've found that jumping reach matters more for DC and ST and that 66% of the benefit comes from boosts to just 3 players. I've previously found this to be true in my testing, though I didn't know the exact numbers. I never really said much about it I think, but I had a separate meta template for DC where jumping reach was 17, and I mentioned once or twice that it's viable on ST instead but you need at least ~2 players with high jumping reach. And the meta team I test with, has 17 jump on the DCs. This is in spite of the conclusion I came to that generally speaking, all positions share the same attribute importance. So it's definitely a thing, and not just for FM27 but FM24 too.

While the forum was down, I replaced the FM24 match engine files in simatches.fmf with the ones from FM27 with the exception of a few of them (OOP, new tactical effectiveness files, and one other folder), and it worked. I comparatively tested it in simulation and found there was only a ~1-4% difference, but it seemed to be a definite difference. So I figure that FM27 is actually using pretty much the same attribute values in the match engine as FM24, and that something else (the tactical effectiveness say) is somehow changing everything up.

In testing GK attributes with sometimes 50,000 samples per attribute I'm pretty sure I've found that some of the attributes give a (slight) boost to goals scored rather than just reducing goals conceded. Probably not the ones you've looked at so far, but just thought I'd mention it.
 
First of all, I want to thank everyone involved, on behalf of myself and everyone else who has been enlightened by this. I think you've overlooked something. There's also the issue of the compatibility of the abilities you mentioned. For example, pace alone isn't enough, but pace and acceleration together are. While having pace and acceleration at 15 with a finishing score of 10 might be sufficient, having finishing at 1 with both pace and acceleration at 20 isn't enough for a striker. I understand you're speaking based on the percentages of those abilities you mentioned, but I'm starting to think you're overlooking the connection between them. In other words, it's necessary to analyze not only which abilities need to be developed for maximum benefit, but also what the minimum values of other abilities should be. Thank you for listening to my opinion and for all your effort.
 
Back
Top