OpticFawn said: I actually appreciate your tests fr, look forward to your files as this as a good baseline to apply to the upcoming game imo Expand It's always nice to have one's work appreciated, but I carry on regardless.
As I've said before, being beholden to praise or criticism paralyzes and retards the creator. We see that with the state of FM itself. They have no balls, and likely no genuine interest in the game either.
Someone said in this thread earlier 'I liked X file, but now you've gone off the rails with Y file!'. The deeper one gets, the less likely others are able to understand or sympathize.
I know I'm upsetting a fair few people by constantly coming up with new conclusions that contradict or confound the previous ones at times, not least because they may take it as painting them a fool for believing them once, but this is the process genuine progress necessitates. It's keithb who is always right from the get-go, not me.
Rain said: I understand what you're saying, but you also can't assume your player will be playing the same teams every time there will be different players playing depending on who the managers decide to play anyway. Expand You are saying either one of two things: That my method is unreliable because the opposition teams are in flux, or otherwise that deducing weightings more precisely is pointless because in reality the opposition teams are always in flux.
Look, my view is that I don't need teams to be frozen, because I actually think having teams in flux replicates the realism better. I don't need to be able to work out precisely how good a particular player is. If I can just improve the reasonable certainty from +/- 15% to +/- 8%, then that's what worth doing in my opinion. Of course, data from controlled testing is often useful, but it's not the be all or end all.
Rain said: Also we know what is important so why are you still trying to get this to an exact science 3 years later? Expand Because it is more enjoyable than playing the game itself. The difficulty makes it an enjoyable challenge, and it's one of the few things I know of that hasn't been done to death by a million people.
You say 'we know what is important', but it has become apparent that the assertion pace/acc is everything isn't quite true anymore. I'm still discovering new things that surprise me. And judging by quite a few of the recent replies in this thread, many don't have a clue what's important in the game and what's not.
Now you could say 'well, we've worked out the game well enough to win everything with ease', and that would have been true already back in 2019 - before EBFM, HarvestGreen, FM Arena, etc. - because we had Knap tactics. If other people want to persist with the game vanilla or whatnot, good for them - the problem I have with it is when people tell me what to do with my time based on that attitude.
Yarema said: And that was my whole point. That other attributes aren't completely useless and shouldn't be valued at 0. Expand They are, as demonstrated by the Messi vs Furuhashi example I showed you in detail, but you have seemingly chosen to overlook.
I have almost completed another example which is even better (Luis Suarez vs Le Fonde). Suarez has green in almost everything technical/mental, but has similar low physicals to Le Fonde. Suarez is finishing consistently 7th to 10th, making him the worst player tested so far.
I don't know why you want to die on this hill. I was talking about sta, pace, acc which are attributes that matter. A majority of attributes do not matter at all, and should be valued at 0, even when other attributes are identical, due to the CA cost.
Rain said: If you really want to test strikers and rank them should probably just give all the teams in the league including your own the exact same attributes besides the player you are testing. That way you will have consistent data and confirm your player is always playing in the same circumstances. Expand HarvestGreen did that and got results that don't align with the real rankings. You could say my rankings aren't real because the rest of my team is artificial, but it would be closer to the reality than HarvestGreen's because all of the opposition is real.
The reason for the discrepancy between HarvestGreen's data and the rankings in reality, is that he's testing against opposition that is either all 10 or 20. If I put an ST with 12 jump reach in a league where every player has 1 jump reach, guess how many aerial attempts he's going to win. But each Premier League team in reality has players with at least ~15 jump reach.
Nonetheless I'm currently plugging in HarvestGreen's 10 > 18 data into FMST26 as weights to see how it fares compared to the weights I ended up with so far, because the more accurate the starting set of weights, the less likely we get a set of weights that 'functions' properly most of the time but isn't actually using the right weights that will get us to near 100% accuracy. One problem that appears with HarvestGreen's weights is that Maeda and Furuhashi are underrated.
What I'm doing at the moment is using HarvestGreen's 10 > 18 & 6 > 18 data as a baseline, then changing just the minimum of what needs to be changed to make things line up.
One example of a radical change, and what I think is going on:
In my weightings attempt before this, I found stamina ~90 (a bit higher than acc/pace) is necessary. HarvestGreen says stamina 10 > 18 is 21.6% of acc 10 > 18. I'm saying it's ~110%.
From my own testing, I know the following is ideal: 20 acc/pace 13 sta | 18 acc/pace 14 sta | 14 acc/pace 15 sta. Something like that anyway, you get the picture. So we see sta is 1/3 of acc, or it's higher at 1 to 1 or a bit more.
Acc & Pace are 2 attributes it's conserving, so sta @ 1/3 acc is actually 2/3 of acc or pace alone. So we know it's in a range of ~67%-200%. Hence I think why something like 110% sta weighting of acc works and HarvestGreen's 21.6% doesn't.
But why did HarvestGreen get 21.6%? Well in his test his players have 10 in every attribute except the one tested, so he's doing 10 acc/pace 18 sta vs 10 acc/pace/sta. That translates to roughly ~9.7 acc/pace vs. 8.3 acc/pace. That's +17% for the 18 sta team.
I've just started having a look at this tool called FMST 26, but it looks like the solution to my problems.
In spite of the name, it works for both FM24 and FM26.
It reads direct from memory, so we've got access to all the hiddens now for FM24, and it was instant to load for me too.
It can measure and sort by match ratings and goals.
And the cherry on top is that not only does it allow for a custom set of weightings, but it seems like it's possible to do more sophisticated mathematical and logical operations with the weights too. If I'm correct in that, then that opens up the possibility of accounting for non-linear attributes and attribute interactions further down the track.
It does lack a few things Genie Scout has or had (player attribute compare, team overall % comparison), but it has the things I need most.
I've had no choice but to continue working with Genie Scout until now, because it's the only tool that can weight the hiddens for FM24. It doesn't calculate things properly, but I decided to just do the best job I could do with it.
So now I can see what my weightings are actually ranking the players accurately, and I see that Maeda is finally 3rd after Haaland and Mbappe as he should be, which is a good sign. I now have to readjust the weightings slightly, but now I can be certain they're being calculated accurately.
I suppose I'll furnish an example, to demonstrate how dirtiness doesn't just solve Maeda, it hits a number of birds with one stone.
If you look at Lautaro Martinez, it's very hard to explain why he finishes 3.75. He has lower pace/acc than Mbappe, lower key mentals than Lewandowski, but if you get familiar with all the pros and cons, the conclusion is that he should be better.
But it's the comparison between Martinez and Maeda that really makes this clear:
Looking at this you've got to ask, how in the hell does Maeda do better than Martinez. It's one reason I've already taken 8 samples on Martinez. If you emphasize pace/acc to an extreme level, that drops the bomb on Kane & Lewandowski in particular. But with the dirtiness factor, it suddenly works out - not just for Maeda vs Martinez, but also Martinez vs Mbappe, Lewandowski, etc.
Still a fair while to go on getting a robust set of weightings, but I wanted to share something I've discovered so far.
Dirtiness seems to be far more important than I ever gave it credit for, and injury proneness isn't far off either.
How do I conclude this?
1) High weight of dirtiness is the only thing I can see that allows Maeda to as highly ranked as his position. More highly valuing any other attribute he has (of which there are few he has high), such as work rate or determination, would mess up the rank of other players. Maeda has very low dirtiness of '2'.
2) High weight of dirtiness also just so happens to fix up a lot of the other ranking problems. So, more circumstantial evidence it's real.
3) I compare not just the rankings, but the relative %, and the only one that was particularly off was Dybala. So I looked into Dybala and noticed the unique thing he does have is high injury proneness. I had not added injury proneness as a weight so far, so I added it, and it largely fixed up Dybala's %. Then I looked at my test saves of Dybala to compare.. indeed what I found that when Dybala finished 3rd he had minimal injuries, whereas when he finished 12, there were a bunch of moderate and major injuries. I added 4 of each player to the team, so I didn't think this would become an issue, but it still does. That's not a problem for the validity at all in my opinion, because in reality you're not going to have more than 4 first team STs anyway - if anything you'd have less of that caliber, which only emphasizes injury proneness more if anything.
4) I already reasoned long ago that dirtiness is worse than injury proneness, because with dirtiness you get a red card and lose a player, whereas with injury proneness you get an injury but the player gets immediately replaced with one that is only ~10% worse. On the other hand, injury proneness also leads to extended periods of low match fitness, which has a strong effect on performance, and so injury proneness is more serious than just the in-match effect.
All that said, it could just be that dirtiness and/or injury proneness are just acting as 'functional' weightings at the moment, that could be disproved later by certain other players. I'd say the probability of this is only say ~20% though, mainly because it seems impossible to explain Maeda doing so well without it.
Yarema said: Now add 40 CA to them in nonessential attributes and sim again. If possible a fair assessment, not all CA into the absolute worst attributes you can think of.
Also the not engineered distribution is sort of true but not really. Using the meta schedules to boost physicals to levels that were obviously not intended is kind of engineering them, although they do happen "naturally" as opposed to straight up editing them. Expand Here you go lad
Not exactly the same in what I would consider the meaningful attributes, but hey, what two players in reality are? And I think it's fair to say that one would assume Messi's +10 dribbling would offset Furuhashi's +8 work rate.
In a team of controlled players but also in a real league, simulated for multiple seasons on full detail.
Training was left as default, no meta training or quickness focus. Positions were also pre-edited to be the same: Only 20 in ST.
Hiddens are close to the same, Messi perhaps has the edge, but judge for yourself:
90% | Erling Haaland - 1st, 1st, 4th, 1st, 2nd, 1st = 1.666 position 85% | Kylian Mbappe - 2nd, 3rd, 3rd, 3rd, 5th, 2nd, 1st, 1st = 2.5 position 83% | Daizen Maeda - 2nd, 3rd, 4th, 4th, 2nd, 3rd, 3rd = 3 position 84% | Kyogo Furuhashi - 4th, 2nd, 3rd, 5th, 3rd, 2nd, 2nd = 3 position 84% | Harry Kane - 5th, 2nd, 3rd, 3rd, 1st, 7th = 3.5 position 84% | Robert Lewandowski - 2nd, 2nd, 2nd, 3rd, 5th, 2nd, 8th, 5th, 3rd = 3.555 position 82% | Dusan Vlahovic - 1st, 4th, 4th, 5th, 1st, 6th, 4th = 3.571 position 89% | Lautaro Martinez - 7th, 3rd, 1st, 5th, 2nd, 5th, 2nd, 5th = 3.75 position 81% | Victor Osimhen - 3rd, 3rd, 7th, 3rd, 3rd, 4th, 3rd, 4th = 3.833 position 79% | Lionel Messi - 4th, 2nd, 1st, 5th, 9th, 3rd = 4 position 86% | Viktor Gyokeres - 6th, 4th, 3rd, 4th, 5th, 5th, 2nd, 5th = 4.25 position 77% | Donyell Malen - 2nd, 2nd, 7th, 5th, 5th, 7th = 4.666 position 74% | Robert Glatzel - 5th, 7th, 6th, 3rd, 12th, 6th = 6.5 position 73% | Paulo Dybala - 5th, 5th, 6th, 10th, 3rd, 12th = 6.833 position 64% | Adam Le Fondre - 8th, 11th (sacked), 5th, 7th (sacked), 6th (sacked), 8th (sacked) = 7.5 position
DR
85% | Giovanni Di Lorenzo - 3rd, 3rd, 2nd, 9th, 2nd, 5th, 5th = 4.143 position 71% | Kingsley Schindler - 4th, 11th, 6th, 11th = 8 position
The percentage is what I get based on the current weightings I have, adjusted where the top player in the position is 90%. Currently it needs to be updated to account for Gyokeres and Martinez.
The same set of weights for ST are used for DR.
I'm no statistician, but ChatGPT says that the reasonable margin of error for a player's percentage so far is about +/- 7.8%. This is inclusive of the natural seasonal variance as well as the inaccuracy of the weightings. It means that you can be pretty certain that best player is going to be at least 82%, or for a practical example, if you were selecting the top natural ST for England, it would narrow it down to 14 players (Kane 84.27% > Danny Ings 76.90%).
tam1236 said: Actually, it does. Rather not in league but in cups: the greater gap in rep between clubs the greater the chance players from lower rep will be kind of intimidated/scared and play well below their potential, even with opponents on equal level. And (sometimes it helps) if you are a coach of such a lower rep team "nobody bet on us and this is our chance" speech is just obligatory. On the other hand, when in national teams (I know it's luckily abandoned in FM26) the reputation, in this case "ranking", is crucial. Firstly - You won't win with a team of much better ranking - ok, there is a minimal chance, but only a minimal - no matter how good players You have. And secondly, even more important, a level of physioterapists you get depends on your ranking. And tournaments like World Cup are scripted by SI in such a way that without very good or even best physioterapists You play with deadmen from 3-5 match - there's nothing you can do about it Expand Didn't know/think about that
But it brings to mind Important Matches, which I guess in a way is an indirect effect of club rep. We know that as you progress in a cup, the final stages - even the quarter finals and whatnot - have increasing reputation and therefore Important Matches comes into play. But how would distinguish that from the intimidation? Maybe intimidation is just a cosmetic symptom of that. On the other hand, when it comes to loan growth, it's 70% Division Rep 30% Club Rep, so I guess it's likely the effect on Important Matches or Pressure or whatever it is works the same way.
I didn't know this nation ranking thing either.
In regards to physiotherapists, I think it was EBFM who found that it doesn't matter what the quality of your physio is, just so long as you've got one. The benefits beyond that I think are just more accurate assessment of how long a player will be out for or something like that. I don't see why it wouldn't be the same for national teams.
I swallowed my pride and tried using ChatGPT to find the right set of weights. It seemed promising at first, but I ended up finding my own estimates were already closer to the mark and ChatGPT agreed.
After adjusting my weights more accurately to the 6-sample results, I noticed that the 7th best ST is Kyogo Furuhashi - ahead of Lewandowski and Mbappe, which is very unusual and therefore a good test of whether these weights are valid or not.
Kyogo Furuhashi:
4th, 2nd, 3rd, 5th, 3rd, 2nd = 3.166
After adjusting things slightly in light of Furuhashi's result and some more samples of a few players, Mbappe is 7th, Lewandowski is 9th, and Furuhashi is 10th.
What's insightful about this result too is that unlike Maeda who has 19 pace/acc, Furuhashi has modest physicals (16 acc/15 pace) and yet shares the same piss poor technicals including 10 drib. His mentals aren't even that remarkable either.
So he's basically showing us one of the more effective pieces of evidence for the validity of the weightings in the leanness of his attribute distribution (he is just 136 CA, compared to 184 CA for Kane below him).
According to Falbraav's 'META FMA 2.0' weightings, Furuhashi is 48th. According to keithb's weightings, Furuhashi is 47th for FS and 538th for TS.
iezzex said: Do club rep affect result? Expand On 'no detail' processing, possibly.
On 'full detail', can't remember now, but I don't think so. I know for certain that player rep doesn't. I don't think club attributes affect the results either, but I suppose it does come into consideration when AI managers are deciding what team to play against you (you know, how after mid-season, suddenly games get a lot harder once they adjust to your actual team quality).
Yarema said: Way to misrepresent the data. Why don't you add the steps from 10-15 and 15-20 as well? Oh right, because they are quite similar as 1-6 Expand If you ignore 6 > 10 for dribbling, then dribbling appears just about linear as well:
You choose to see what you want to see. I choose to see what is actually there, and what I see are two different attributes, showing the same pattern of non-linearity at the 6-10 level, which reduces quite significantly the chance that it is a statistical error.
Can't be bothered digging it up, but there's more deductions you can make like this for other attributes by comparing his 6 > 18 data to his 1 > 20, 10 > 20 data and whatnot
Yarema said: Sorry what? Other than work rate, the other attributes are pretty much linear.
Obviously we aren't looking for the 6th decimal, there is margin of error. And they might even behave slightly nonlinearly but a linear function can be a very good approximation to reduce complexity. Expand Acceleration 1 > 6 = +42.7 (+8.54/extra point) Acceleration 6 > 10 = +58.6 (+14.65/extra point)
The non-linearity is not as profound as for work rate, but it is nonetheless significantly non-linear.
Additionally I would posit that the league quality itself introduces a further level of non-linearity (or perhaps the non-linearity is a reflection of it, and would be more pronounced in a realistic league). HarvestGreen used a league that from memory had players with 10 in everything. In the Premier League, you need at least ~13 acceleration just to survive, so I would expect to see a large disparity between say acceleration 6 > 10 and acceleration 10 > 14 in the Premier League.
keithb said: So again as always you have made a false claim. Nowhere has harvest green said attributes are non linear. Expand
harvestgreen22 said: Non linear Its influence is relatively significant from 1 to 8, but it becomes much smaller from 14 to 20
If he has, for instance, 16 ，18 , or even 20, it might waste a considerable amount of ability, but the effect actually only increases very little Expand
As for Mbappe, it's true I underestimated him a little as ST. I had him as 12th; in the latest best fit to the objective results I'm working on, he's currently 6th. The difference between us is not that I don't make mistakes and you do, but that I correct my mistakes, while you continue to insist that Mbappe is equal to Haaland even when reality shows otherwise time and time again.
And no, I'm not going to remove the results of your file from my list, so stop DM'ing me about it.
keithb said: So that was it. Where does he say speed, anticipation, dribbling and jumping reach for example aren't linear. Again we knew a lot of attributes don't matter at all, some don't matter after they reach a certain value. But the high scoring ones do.
I can't work out if you're stupid or if you think we're stupid? Maybe both?
As for being a liar when you released premier league 1 you didn't set any hidden attributes. Therefore genie scout wasn't producing skewed numbers. Yet you repeatedly lied that genie scout was at fault for giving Mbappe such a low score. Again this was a lie it was your appalling ratings. Expand What actually first came to my mind was HarvestGreen's excel tables, which show the non-linear effect of multiple attributes very clearly and in detail. Look at the right side:
Here we see that speed, work rate, among others, have non-linear effects.
Dribbling follows more of an inverted U-shaped curve, so in practice is mostly linear, but not always. I've always said that myself:
GeorgeFloydOverdosed said: HarvestGreen22, using a very artificial setup method (all attributes 10 bar the one being tested) and assessing without regard for position, claims dribbling is 4th most impactful attribute and follows a largely but not entirely linear benefit through 1-20. Expand
You've got the GS narrative backwards. I created the Premier League 1.0 weightings (without releasing them), then was informed by someone else that there was a problem with GS, and deduced it was the hidden attributes causing the majority of the issues because GS doesn't calculate things correctly with them. The weights work as intended in other programs (i.e. FM26 scoring system), including the hiddens (i.e. FM SuperScout). When I released Premier League 1.0 for GS, I removed the hiddens so it could work as best it could, since we can't really do away with GS entirely yet.
Objective testing has shown now that Mbappe is worse than 34-year-old Lewandowski and 36-year-old Messi gives him a run for his money. Mbappe is no where near Haaland on any measure - team position, goals, rating. And your own rating file turned out be worse than FM Genie Scout's default on key measures. Your file shows that apparently Dybala is as good as Lewandowski or Kane 🤡
So as I posted in another thread, I've ran some controlled tests of specific players and found they finish in this ranking order (ranked by team position):
Erling Haaland = 1.666 Robert Lewandowski = 2.666 Kylian Mbappe = 3 Daizen Maeda = 3 Harry Kane = 3.5 Dusan Vlahovic = 3.571 Victor Osimhen = 3.833 Lionel Messi = 4 Donyell Malen = 4.666 Robert Glatzel = 6.5 Paulo Dybala = 6.833 Adam Le Fondre = 7.5
6 full season samples each, all relevant competitions on full detail, and this isn't your run-of-the-mill plop Haaland at Arsenal and see how he goes - the rest of the team are all identical non-exceptional players in order to minimize variation while remaining realistic and tested to finish ~5 on average by themselves (we replace the STs with 4 duplicates of the player we want to test)
If you use Falbraav's file which puts the weightings into Genie Scout, it doesn't align with the real results at all:
If we remove the hiddens, which are known to be problematic in GS, not much changes. Lewandowski remains behind Mbappe & Osimhen; Maeda remains near the bottom.
The rankings remain incorrect if we plug Falbraav's GS figures into FM26 scoring system (to rule out GS being the problem):
While I was doing this and looking at OP's table, I realized what could be a big part of the problem. He's using points to measure success rather than team position. Points are very unreliable, because they can be 57 for one season sample and 83 the next - what doesn't change much is the team position, so that team will likely finish 2nd both times even though the point tally is all over the place. You would still get some valid correlations obviously, i.e. no pace/acc = no points at all.
I suggest to the OP if he intends to have another crack at it, to assess by the team position and not the team points.
And another problem with points vs team position could be that it neglects defensive advantages. It could explain why positioning does so terribly, for perhaps better defense favors draws as opposed to victories and losses - but a draw is only worth 1 point; a win 3. We want victories anyway, so this doesn't matter all too much either way.
He said it here, about Work Rate. Not sure about the other attributes, but I remember reading that and I've had a filter set at 6+ Work Rate ever since. Lowering the Work rate 1 , it have very serious consequences (goal difference -110),
and it should be ensured that a player has at least 6 and preferably 10 Work rate. A player with a Work rate is not desirable. But continuing to increase Work rate (from 10->20) is low effect. The difference between Work rate 10 and Work rate 20 is relative small.
The 6-point attribute seems to be the threshold for some attributes.
Core attributes are these: Special : Work rate (need to reach 10, higher is useless) Expand
keithb said: Well if thats all he said then he didn't say it did he. But no surprise if Bozo is quoting that. He's a liar. Expand Amazing.
Yarema said: If an attribute point has a value of 10 from 4 to 5 and 1 from 14 to 15, how exactly do you set a fixed value for scoring in GS? That is the fundamental issue. You need linear scaling at least in the vast majority of values. You can't just say oh it's an average of lets say 3 because it'll be right only in a narrow bracket or maybe even never.
What you can do for attributes that have minimal or no value above a certain point is to set it's value to 0 (or whatever minimal value the calculations show) with a caveat that it only applies above X and that you need to filter for it. I am not disputing that some attributes might act that way, but that assigning them some value (by feel?) is misleading. And again FM-Arena tests showed that going from 18 to 19 to 20 pace is more or less just as valuable, it keeps it's scaling. Can you win with 18? Sure, but 20 is still better. Expand With respect, my post was my answer to that very question.
You could argue the method is inadequate, but you can't say it achieves nothing at all, and I'd challenge you with the following question: If not this method, then what? Assume attributes are linear and get assessments that are even more wrong?
It's true that filters can be part of the solution, and I've suggested to people before to use filters in combination with GS weightings if they can, but obviously this isn't a very good solution. Filtering is too blunt a tool for one thing; it's better to use weighted weightings instead usually.
I don't believe in FM Arena testing much, because they obviously contradict the findings of HarvestGreen, Orion, and myself. Even aside from the results, I didn't like some of the methodology that I could see, in the same way that I don't like how HarvestGreen uses an artificial league where everybody has the same attributes, instead of using a real league like the Premier League.
We see something similar reveal itself in this guy's testing. I would characterize it as a kind of blase 'let's apply a python script to this' mentality. Sometimes, the kind of info that gets can be very useful, but I think more typically it leads to skipping over the necessary nuances that only someone who has waded deep into it with hours of trial and error and reflection and critical thought can discern.
For example, he says his results show that attributes are linear. Ok, so why hasn't he reflected and deduced why this contradicts HarvestGreen's findings? The most likely answer is that he simply hasn't put in the time in doing that. By his own account the data was from an '8 hour overnight session', and I don't know how long he spent on it overall, but my impression is that this is probably something he's done over a week or two.
I'm not trying to throw shade, and it's no crime to not spend your life on analyzing FM, I'm just saying it's an approach that has certain downsides. My downsides are that I'm not a programmer, so I can't use such a methodical and fast approach, and because I'm always coming up with different ideas, I don't say 'this is the result I got, that's it from me'.
Now you say that assigning values by feel is misleading. Well let's look at a concrete example:
I brute force test players until I know who ranks where.
Starting with informed guesses (i.e. start pace at '100' not '0', and passing at '0' not '100' ), I adjust the weightings in Genie Scout until the player rankings all correspond exactly to their actual performance ranking.
Because some or many attributes are not linear, there may be a 1% error here, a 3% error there, and even a handful of ranks out of the thousands out of alignment, but for the most part, it does its job.
I could continue the example further, but I think the point is already clear - where exactly is the error or impossibility in it?
It's always nice to have one's work appreciated, but I carry on regardless.
As I've said before, being beholden to praise or criticism paralyzes and retards the creator. We see that with the state of FM itself. They have no balls, and likely no genuine interest in the game either.
Someone said in this thread earlier 'I liked X file, but now you've gone off the rails with Y file!'. The deeper one gets, the less likely others are able to understand or sympathize.
I know I'm upsetting a fair few people by constantly coming up with new conclusions that contradict or confound the previous ones at times, not least because they may take it as painting them a fool for believing them once, but this is the process genuine progress necessitates. It's keithb who is always right from the get-go, not me.
You are saying either one of two things: That my method is unreliable because the opposition teams are in flux, or otherwise that deducing weightings more precisely is pointless because in reality the opposition teams are always in flux.
Look, my view is that I don't need teams to be frozen, because I actually think having teams in flux replicates the realism better. I don't need to be able to work out precisely how good a particular player is. If I can just improve the reasonable certainty from +/- 15% to +/- 8%, then that's what worth doing in my opinion. Of course, data from controlled testing is often useful, but it's not the be all or end all.
Rain said: Also we know what is important so why are you still trying to get this to an exact science 3 years later?
Because it is more enjoyable than playing the game itself. The difficulty makes it an enjoyable challenge, and it's one of the few things I know of that hasn't been done to death by a million people.
You say 'we know what is important', but it has become apparent that the assertion pace/acc is everything isn't quite true anymore. I'm still discovering new things that surprise me. And judging by quite a few of the recent replies in this thread, many don't have a clue what's important in the game and what's not.
Now you could say 'well, we've worked out the game well enough to win everything with ease', and that would have been true already back in 2019 - before EBFM, HarvestGreen, FM Arena, etc. - because we had Knap tactics. If other people want to persist with the game vanilla or whatnot, good for them - the problem I have with it is when people tell me what to do with my time based on that attitude.
Adam Le Fondre - 8th, 11th (sacked), 5th, 7th (sacked), 6th (sacked), 8th (sacked) = 7.5 position
Luis Suarez - 7th, 9th, 10th, 8th, 8th = 8.4 position
Note that Suarez isn't (significantly) superior in the key mentals of anticipation, concentration, work rate.
The argument that useless technicals/mentals should be counted even a little is in the toilet.
They are, as demonstrated by the Messi vs Furuhashi example I showed you in detail, but you have seemingly chosen to overlook.
I have almost completed another example which is even better (Luis Suarez vs Le Fonde). Suarez has green in almost everything technical/mental, but has similar low physicals to Le Fonde. Suarez is finishing consistently 7th to 10th, making him the worst player tested so far.
I don't know why you want to die on this hill. I was talking about sta, pace, acc which are attributes that matter. A majority of attributes do not matter at all, and should be valued at 0, even when other attributes are identical, due to the CA cost.
HarvestGreen did that and got results that don't align with the real rankings. You could say my rankings aren't real because the rest of my team is artificial, but it would be closer to the reality than HarvestGreen's because all of the opposition is real.
The reason for the discrepancy between HarvestGreen's data and the rankings in reality, is that he's testing against opposition that is either all 10 or 20. If I put an ST with 12 jump reach in a league where every player has 1 jump reach, guess how many aerial attempts he's going to win. But each Premier League team in reality has players with at least ~15 jump reach.
Nonetheless I'm currently plugging in HarvestGreen's 10 > 18 data into FMST26 as weights to see how it fares compared to the weights I ended up with so far, because the more accurate the starting set of weights, the less likely we get a set of weights that 'functions' properly most of the time but isn't actually using the right weights that will get us to near 100% accuracy. One problem that appears with HarvestGreen's weights is that Maeda and Furuhashi are underrated.
What I'm doing at the moment is using HarvestGreen's 10 > 18 & 6 > 18 data as a baseline, then changing just the minimum of what needs to be changed to make things line up.
One example of a radical change, and what I think is going on:
In my weightings attempt before this, I found stamina ~90 (a bit higher than acc/pace) is necessary.
HarvestGreen says stamina 10 > 18 is 21.6% of acc 10 > 18. I'm saying it's ~110%.
From my own testing, I know the following is ideal: 20 acc/pace 13 sta | 18 acc/pace 14 sta | 14 acc/pace 15 sta. Something like that anyway, you get the picture. So we see sta is 1/3 of acc, or it's higher at 1 to 1 or a bit more.
Acc & Pace are 2 attributes it's conserving, so sta @ 1/3 acc is actually 2/3 of acc or pace alone. So we know it's in a range of ~67%-200%. Hence I think why something like 110% sta weighting of acc works and HarvestGreen's 21.6% doesn't.
But why did HarvestGreen get 21.6%? Well in his test his players have 10 in every attribute except the one tested, so he's doing 10 acc/pace 18 sta vs 10 acc/pace/sta. That translates to roughly ~9.7 acc/pace vs. 8.3 acc/pace. That's +17% for the 18 sta team.
In spite of the name, it works for both FM24 and FM26.
It reads direct from memory, so we've got access to all the hiddens now for FM24, and it was instant to load for me too.
It can measure and sort by match ratings and goals.
And the cherry on top is that not only does it allow for a custom set of weightings, but it seems like it's possible to do more sophisticated mathematical and logical operations with the weights too. If I'm correct in that, then that opens up the possibility of accounting for non-linear attributes and attribute interactions further down the track.
It does lack a few things Genie Scout has or had (player attribute compare, team overall % comparison), but it has the things I need most.
I've had no choice but to continue working with Genie Scout until now, because it's the only tool that can weight the hiddens for FM24. It doesn't calculate things properly, but I decided to just do the best job I could do with it.
So now I can see what my weightings are actually ranking the players accurately, and I see that Maeda is finally 3rd after Haaland and Mbappe as he should be, which is a good sign. I now have to readjust the weightings slightly, but now I can be certain they're being calculated accurately.
If you look at Lautaro Martinez, it's very hard to explain why he finishes 3.75. He has lower pace/acc than Mbappe, lower key mentals than Lewandowski, but if you get familiar with all the pros and cons, the conclusion is that he should be better.
But it's the comparison between Martinez and Maeda that really makes this clear:
Looking at this you've got to ask, how in the hell does Maeda do better than Martinez. It's one reason I've already taken 8 samples on Martinez. If you emphasize pace/acc to an extreme level, that drops the bomb on Kane & Lewandowski in particular. But with the dirtiness factor, it suddenly works out - not just for Maeda vs Martinez, but also Martinez vs Mbappe, Lewandowski, etc.
Dirtiness seems to be far more important than I ever gave it credit for, and injury proneness isn't far off either.
How do I conclude this?
1) High weight of dirtiness is the only thing I can see that allows Maeda to as highly ranked as his position. More highly valuing any other attribute he has (of which there are few he has high), such as work rate or determination, would mess up the rank of other players. Maeda has very low dirtiness of '2'.
2) High weight of dirtiness also just so happens to fix up a lot of the other ranking problems. So, more circumstantial evidence it's real.
3) I compare not just the rankings, but the relative %, and the only one that was particularly off was Dybala. So I looked into Dybala and noticed the unique thing he does have is high injury proneness. I had not added injury proneness as a weight so far, so I added it, and it largely fixed up Dybala's %. Then I looked at my test saves of Dybala to compare.. indeed what I found that when Dybala finished 3rd he had minimal injuries, whereas when he finished 12, there were a bunch of moderate and major injuries. I added 4 of each player to the team, so I didn't think this would become an issue, but it still does. That's not a problem for the validity at all in my opinion, because in reality you're not going to have more than 4 first team STs anyway - if anything you'd have less of that caliber, which only emphasizes injury proneness more if anything.
4) I already reasoned long ago that dirtiness is worse than injury proneness, because with dirtiness you get a red card and lose a player, whereas with injury proneness you get an injury but the player gets immediately replaced with one that is only ~10% worse. On the other hand, injury proneness also leads to extended periods of low match fitness, which has a strong effect on performance, and so injury proneness is more serious than just the in-match effect.
All that said, it could just be that dirtiness and/or injury proneness are just acting as 'functional' weightings at the moment, that could be disproved later by certain other players. I'd say the probability of this is only say ~20% though, mainly because it seems impossible to explain Maeda doing so well without it.
Also the not engineered distribution is sort of true but not really. Using the meta schedules to boost physicals to levels that were obviously not intended is kind of engineering them, although they do happen "naturally" as opposed to straight up editing them.
Here you go lad
Kyogo Furuhashi - 4th, 2nd, 3rd, 5th, 3rd, 2nd, 2nd = 3 position
Lionel Messi - 4th, 2nd, 1st, 5th, 9th, 3rd = 4 position
Furuhashi vs Messi - 136 CA vs 185 CA
Not exactly the same in what I would consider the meaningful attributes, but hey, what two players in reality are? And I think it's fair to say that one would assume Messi's +10 dribbling would offset Furuhashi's +8 work rate.
In a team of controlled players but also in a real league, simulated for multiple seasons on full detail.
Training was left as default, no meta training or quickness focus. Positions were also pre-edited to be the same: Only 20 in ST.
Hiddens are close to the same, Messi perhaps has the edge, but judge for yourself:
ST
90% | Erling Haaland - 1st, 1st, 4th, 1st, 2nd, 1st = 1.666 position
85% | Kylian Mbappe - 2nd, 3rd, 3rd, 3rd, 5th, 2nd, 1st, 1st = 2.5 position
83% | Daizen Maeda - 2nd, 3rd, 4th, 4th, 2nd, 3rd, 3rd = 3 position
84% | Kyogo Furuhashi - 4th, 2nd, 3rd, 5th, 3rd, 2nd, 2nd = 3 position
84% | Harry Kane - 5th, 2nd, 3rd, 3rd, 1st, 7th = 3.5 position
84% | Robert Lewandowski - 2nd, 2nd, 2nd, 3rd, 5th, 2nd, 8th, 5th, 3rd = 3.555 position
82% | Dusan Vlahovic - 1st, 4th, 4th, 5th, 1st, 6th, 4th = 3.571 position
89% | Lautaro Martinez - 7th, 3rd, 1st, 5th, 2nd, 5th, 2nd, 5th = 3.75 position
81% | Victor Osimhen - 3rd, 3rd, 7th, 3rd, 3rd, 4th, 3rd, 4th = 3.833 position
79% | Lionel Messi - 4th, 2nd, 1st, 5th, 9th, 3rd = 4 position
86% | Viktor Gyokeres - 6th, 4th, 3rd, 4th, 5th, 5th, 2nd, 5th = 4.25 position
77% | Donyell Malen - 2nd, 2nd, 7th, 5th, 5th, 7th = 4.666 position
74% | Robert Glatzel - 5th, 7th, 6th, 3rd, 12th, 6th = 6.5 position
73% | Paulo Dybala - 5th, 5th, 6th, 10th, 3rd, 12th = 6.833 position
64% | Adam Le Fondre - 8th, 11th (sacked), 5th, 7th (sacked), 6th (sacked), 8th (sacked) = 7.5 position
DR
85% | Giovanni Di Lorenzo - 3rd, 3rd, 2nd, 9th, 2nd, 5th, 5th = 4.143 position
71% | Kingsley Schindler - 4th, 11th, 6th, 11th = 8 position
The percentage is what I get based on the current weightings I have, adjusted where the top player in the position is 90%. Currently it needs to be updated to account for Gyokeres and Martinez.
The same set of weights for ST are used for DR.
I'm no statistician, but ChatGPT says that the reasonable margin of error for a player's percentage so far is about +/- 7.8%. This is inclusive of the natural seasonal variance as well as the inaccuracy of the weightings. It means that you can be pretty certain that best player is going to be at least 82%, or for a practical example, if you were selecting the top natural ST for England, it would narrow it down to 14 players (Kane 84.27% > Danny Ings 76.90%).
On the other hand, when in national teams (I know it's luckily abandoned in FM26) the reputation, in this case "ranking", is crucial. Firstly - You won't win with a team of much better ranking - ok, there is a minimal chance, but only a minimal - no matter how good players You have. And secondly, even more important, a level of physioterapists you get depends on your ranking. And tournaments like World Cup are scripted by SI in such a way that without very good or even best physioterapists You play with deadmen from 3-5 match - there's nothing you can do about it
Didn't know/think about that
But it brings to mind Important Matches, which I guess in a way is an indirect effect of club rep. We know that as you progress in a cup, the final stages - even the quarter finals and whatnot - have increasing reputation and therefore Important Matches comes into play. But how would distinguish that from the intimidation? Maybe intimidation is just a cosmetic symptom of that. On the other hand, when it comes to loan growth, it's 70% Division Rep 30% Club Rep, so I guess it's likely the effect on Important Matches or Pressure or whatever it is works the same way.
I didn't know this nation ranking thing either.
In regards to physiotherapists, I think it was EBFM who found that it doesn't matter what the quality of your physio is, just so long as you've got one. The benefits beyond that I think are just more accurate assessment of how long a player will be out for or something like that. I don't see why it wouldn't be the same for national teams.
After adjusting my weights more accurately to the 6-sample results, I noticed that the 7th best ST is Kyogo Furuhashi - ahead of Lewandowski and Mbappe, which is very unusual and therefore a good test of whether these weights are valid or not.
Kyogo Furuhashi:
4th, 2nd, 3rd, 5th, 3rd, 2nd = 3.166
After adjusting things slightly in light of Furuhashi's result and some more samples of a few players, Mbappe is 7th, Lewandowski is 9th, and Furuhashi is 10th.
What's insightful about this result too is that unlike Maeda who has 19 pace/acc, Furuhashi has modest physicals (16 acc/15 pace) and yet shares the same piss poor technicals including 10 drib. His mentals aren't even that remarkable either.
So he's basically showing us one of the more effective pieces of evidence for the validity of the weightings in the leanness of his attribute distribution (he is just 136 CA, compared to 184 CA for Kane below him).
According to Falbraav's 'META FMA 2.0' weightings, Furuhashi is 48th.
According to keithb's weightings, Furuhashi is 47th for FS and 538th for TS.
On 'no detail' processing, possibly.
On 'full detail', can't remember now, but I don't think so. I know for certain that player rep doesn't. I don't think club attributes affect the results either, but I suppose it does come into consideration when AI managers are deciding what team to play against you (you know, how after mid-season, suddenly games get a lot harder once they adjust to your actual team quality).
If you ignore 6 > 10 for dribbling, then dribbling appears just about linear as well:
Dribbling
1 > 6 = +15.1 (+3.02/extra point)
6 > 10 = +2.7 (+0.675/extra point)
10 > 15 = +19.5 (+3.9/extra point)
15 > 20 = +14.0 (+2.8/extra point)
You choose to see what you want to see. I choose to see what is actually there, and what I see are two different attributes, showing the same pattern of non-linearity at the 6-10 level, which reduces quite significantly the chance that it is a statistical error.
Can't be bothered digging it up, but there's more deductions you can make like this for other attributes by comparing his 6 > 18 data to his 1 > 20, 10 > 20 data and whatnot
Obviously we aren't looking for the 6th decimal, there is margin of error. And they might even behave slightly nonlinearly but a linear function can be a very good approximation to reduce complexity.
Acceleration 1 > 6 = +42.7 (+8.54/extra point)
Acceleration 6 > 10 = +58.6 (+14.65/extra point)
The non-linearity is not as profound as for work rate, but it is nonetheless significantly non-linear.
Additionally I would posit that the league quality itself introduces a further level of non-linearity (or perhaps the non-linearity is a reflection of it, and would be more pronounced in a realistic league). HarvestGreen used a league that from memory had players with 10 in everything. In the Premier League, you need at least ~13 acceleration just to survive, so I would expect to see a large disparity between say acceleration 6 > 10 and acceleration 10 > 14 in the Premier League.
harvestgreen22 said: Non linear
Its influence is relatively significant from 1 to 8, but it becomes much smaller from 14 to 20
If he has, for instance, 16 ，18 , or even 20, it might waste a considerable amount of ability, but the effect actually only increases very little
As for Mbappe, it's true I underestimated him a little as ST. I had him as 12th; in the latest best fit to the objective results I'm working on, he's currently 6th. The difference between us is not that I don't make mistakes and you do, but that I correct my mistakes, while you continue to insist that Mbappe is equal to Haaland even when reality shows otherwise time and time again.
And no, I'm not going to remove the results of your file from my list, so stop DM'ing me about it.
I can't work out if you're stupid or if you think we're stupid? Maybe both?
As for being a liar when you released premier league 1 you didn't set any hidden attributes. Therefore genie scout wasn't producing skewed numbers. Yet you repeatedly lied that genie scout was at fault for giving Mbappe such a low score. Again this was a lie it was your appalling ratings.
What actually first came to my mind was HarvestGreen's excel tables, which show the non-linear effect of multiple attributes very clearly and in detail. Look at the right side:
Here we see that speed, work rate, among others, have non-linear effects.
Dribbling follows more of an inverted U-shaped curve, so in practice is mostly linear, but not always. I've always said that myself:
GeorgeFloydOverdosed said: HarvestGreen22, using a very artificial setup method (all attributes 10 bar the one being tested) and assessing without regard for position, claims dribbling is 4th most impactful attribute and follows a largely but not entirely linear benefit through 1-20.
You've got the GS narrative backwards. I created the Premier League 1.0 weightings (without releasing them), then was informed by someone else that there was a problem with GS, and deduced it was the hidden attributes causing the majority of the issues because GS doesn't calculate things correctly with them. The weights work as intended in other programs (i.e. FM26 scoring system), including the hiddens (i.e. FM SuperScout). When I released Premier League 1.0 for GS, I removed the hiddens so it could work as best it could, since we can't really do away with GS entirely yet.
Objective testing has shown now that Mbappe is worse than 34-year-old Lewandowski and 36-year-old Messi gives him a run for his money. Mbappe is no where near Haaland on any measure - team position, goals, rating. And your own rating file turned out be worse than FM Genie Scout's default on key measures. Your file shows that apparently Dybala is as good as Lewandowski or Kane 🤡
So as I posted in another thread, I've ran some controlled tests of specific players and found they finish in this ranking order (ranked by team position):
Erling Haaland = 1.666
Robert Lewandowski = 2.666
Kylian Mbappe = 3
Daizen Maeda = 3
Harry Kane = 3.5
Dusan Vlahovic = 3.571
Victor Osimhen = 3.833
Lionel Messi = 4
Donyell Malen = 4.666
Robert Glatzel = 6.5
Paulo Dybala = 6.833
Adam Le Fondre = 7.5
6 full season samples each, all relevant competitions on full detail, and this isn't your run-of-the-mill plop Haaland at Arsenal and see how he goes - the rest of the team are all identical non-exceptional players in order to minimize variation while remaining realistic and tested to finish ~5 on average by themselves (we replace the STs with 4 duplicates of the player we want to test)
If you use Falbraav's file which puts the weightings into Genie Scout, it doesn't align with the real results at all:
If we remove the hiddens, which are known to be problematic in GS, not much changes. Lewandowski remains behind Mbappe & Osimhen; Maeda remains near the bottom.
The rankings remain incorrect if we plug Falbraav's GS figures into FM26 scoring system (to rule out GS being the problem):
Haaland = 16.8
Mbappe = 16.1
Osimhen = 15.6
Lewandowski = 14.5
Maeda = 14.3
Messi = 14.1
While I was doing this and looking at OP's table, I realized what could be a big part of the problem. He's using points to measure success rather than team position. Points are very unreliable, because they can be 57 for one season sample and 83 the next - what doesn't change much is the team position, so that team will likely finish 2nd both times even though the point tally is all over the place. You would still get some valid correlations obviously, i.e. no pace/acc = no points at all.
I suggest to the OP if he intends to have another crack at it, to assess by the team position and not the team points.
And another problem with points vs team position could be that it neglects defensive advantages. It could explain why positioning does so terribly, for perhaps better defense favors draws as opposed to victories and losses - but a draw is only worth 1 point; a win 3. We want victories anyway, so this doesn't matter all too much either way.
He said it here, about Work Rate. Not sure about the other attributes, but I remember reading that and I've had a filter set at 6+ Work Rate ever since.
Lowering the Work rate 1 ,
it have very serious consequences (goal difference -110),
and it should be ensured that a player has at least 6 and preferably 10 Work rate.
A player with a Work rate is not desirable.
But continuing to increase Work rate (from 10->20) is low effect. The difference between Work rate 10 and Work rate 20 is relative small.
The 6-point attribute seems to be the threshold for some attributes.
Core attributes are these:
Special : Work rate (need to reach 10, higher is useless)
keithb said: Well if thats all he said then he didn't say it did he. But no surprise if Bozo is quoting that. He's a liar.
Amazing.
What you can do for attributes that have minimal or no value above a certain point is to set it's value to 0 (or whatever minimal value the calculations show) with a caveat that it only applies above X and that you need to filter for it. I am not disputing that some attributes might act that way, but that assigning them some value (by feel?) is misleading. And again FM-Arena tests showed that going from 18 to 19 to 20 pace is more or less just as valuable, it keeps it's scaling. Can you win with 18? Sure, but 20 is still better.
With respect, my post was my answer to that very question.
You could argue the method is inadequate, but you can't say it achieves nothing at all, and I'd challenge you with the following question: If not this method, then what? Assume attributes are linear and get assessments that are even more wrong?
It's true that filters can be part of the solution, and I've suggested to people before to use filters in combination with GS weightings if they can, but obviously this isn't a very good solution. Filtering is too blunt a tool for one thing; it's better to use weighted weightings instead usually.
I don't believe in FM Arena testing much, because they obviously contradict the findings of HarvestGreen, Orion, and myself. Even aside from the results, I didn't like some of the methodology that I could see, in the same way that I don't like how HarvestGreen uses an artificial league where everybody has the same attributes, instead of using a real league like the Premier League.
We see something similar reveal itself in this guy's testing. I would characterize it as a kind of blase 'let's apply a python script to this' mentality. Sometimes, the kind of info that gets can be very useful, but I think more typically it leads to skipping over the necessary nuances that only someone who has waded deep into it with hours of trial and error and reflection and critical thought can discern.
For example, he says his results show that attributes are linear. Ok, so why hasn't he reflected and deduced why this contradicts HarvestGreen's findings? The most likely answer is that he simply hasn't put in the time in doing that. By his own account the data was from an '8 hour overnight session', and I don't know how long he spent on it overall, but my impression is that this is probably something he's done over a week or two.
I'm not trying to throw shade, and it's no crime to not spend your life on analyzing FM, I'm just saying it's an approach that has certain downsides. My downsides are that I'm not a programmer, so I can't use such a methodical and fast approach, and because I'm always coming up with different ideas, I don't say 'this is the result I got, that's it from me'.
Now you say that assigning values by feel is misleading. Well let's look at a concrete example:
I brute force test players until I know who ranks where.
Starting with informed guesses (i.e. start pace at '100' not '0', and passing at '0' not '100' ), I adjust the weightings in Genie Scout until the player rankings all correspond exactly to their actual performance ranking.
Because some or many attributes are not linear, there may be a 1% error here, a 3% error there, and even a handful of ranks out of the thousands out of alignment, but for the most part, it does its job.
I could continue the example further, but I think the point is already clear - where exactly is the error or impossibility in it?