Chriswin4
Purity said: With the exact same attributes in Nandal's save I got 68.7% for odriozola and 74.6% for wemba.
I am not saying my ratings are better I only used the attributes tested by FM arena and made it a rating and not all attributes were tested. With this rating there is pretty much only physical that matters. It considers adama traore one of the best players in the world for example...


How does Adama Traore play in the game?? I'm actually starting to think the attributes don't necessarily make individual players better, but the more pace you have, the more chances your team makes and the better your results, ergo the better that player does by association.
Purity said: I quickly did a rating system for genie scout after the FM arena attributes testing, I have great success with it but you can tweak it if you want to change the jumping reach for exemple. With this rating system Wemba is 74 and odri 69 if you wanna know. https://www.dropbox.com/scl/fi/2aefavu1ha5j1bi13e7zj/FM-arena.grf?rlkey=qqcqoqt4o2t0t06b37j11mefz&dl=0

Now that makes me giddy, although are your two players different from the ones in Nandal's save?
NandaldiaN said: 😂😂

If everyone agrees, then I think the meta stats still need more importance in the ratings, cause 3 points I think are a huge difference.

Wembangomo la gazzella calabrese.


https://www.transfernow.net/dl/202312311BBLSZif

Download this and try it, tell me what you think. In theory it should bump Wemby up nearer Odri in Genie Scout. I messed with the ratings for FB, WB, Fast Striker and Winger.
I would also choose Wemby, simply because of his pace. However, my ratings shouldn't have someone so much further behind Odri, I may need to tweak the importance of pace and acceleration for each position, maybe put it up to 75.
Good Afternoon,

You may remember me making a post about finding a metric to evaluate players to one solid number, in the thread 'We need a metric for player eval' and I've come to the conclusion that this would take a lot of testing, as was pointed out to me on here.

Anyway, I went to lab and with help from KeithB on here, I've devised my own Genie Scout rating file based on what I think the Match Engine values on each position. I've totally ignored sweeper because it's not 1990. This might not value the players in your game as highly as others, but in theory, it should put people who are the likely top performers in the ME in a certain position to the top of the rankings list.

I would love it if you guys could give this a download and test it on your saves, looking at whether it places the best performers in your worlds in the correct place on Genie Scout. If it works then I'll try and get the weights into a python for ease.

Also, one thing I may have missed, on each position tab in Genie Scout there is a weight value, I didn't fully grasp what this did so left it at 100 for each position. Let me know if this is something that should change. Also be sure to please post some screenshots of players in-game and what they are rated by my Genie Scout rating system.

Just download the file, place it in the Ratings Folder in your FM Genie Scout 2024 directory, then load these ratings in Genie once you've loaded your save game in FM.

Thanks a lot guys,
Chris

https://www.transfernow.net/dl/20231230yqxZvZha
Floppyaams said: just look at the attribute testing table here. All the tactics that don't improve results are the non-meta stats. But that does not mean they are useless. Just make sure you have a player with 10-12 on those stats. And I absolutely think meta stats can compensate for the difference in star rating (there is a lot of wrong with star rating). There was an old post on reddit where this guy won the PL with a team full of 20 pace, 20 acceleration and 20 jumping but 1s on everything else. His team even won the FA cup against Liverpool on penalty shootout despite having 1s across the board in pen and finishing.

Thank you mate!! I’m going to have a look at this
Floppyaams said: The second guy is the clear winner imo. Great physicals and better technicals (that matter for an AF) compared to the other two.

How do you make your calls on the non meta stats? I’ve been looking at player roles and trying to make sure the blue highlighted ones are above 10
Yeah I much prefer attribute evaluation than statistics. Players not being controlled by a human manager don’t necessarily play on the match engine, so you can’t judge if they’ll be good when you buy them
How would you guys evaluate/judge these guys as Advanced Forwards? Are there any traits that work this year? I generally do, first time shots, blast the ball and beat offside trap?





Floppyaams said: One thing I would look out for is to make sure that the non-meta stats are at the "acceptable range". Many people conclude that the non-meta stats are completely useless, because of the attribute testing here. What the tests actually show is that non-meta stats don't provide additional benefits after a certain point. So make sure that the non-meta attributes are at that point (the attribute testing control group is the base I use). I have managed to win the Bundesliga and the UCL 3 years in a row with a team that pays sub 1 mil per week in wages by looking for players with high meta stats and good enough non-meta stats.

I think this is where I've gone wrong, I've gone too far in the search for cheapness. What are the non-meta stats that I should be looking out for? What should they be over to try and win the CL with under 1m in wages per week.
NandaldiaN said: I have to say that sometimes i followed the python scores and bought players that have better score than the ones that my team already have and they dont perform well.. it happened to me in more than one save

I think it depends on the player traits and RNG. I noticed in the two wingers I posted earlier that one performed much better because they had a strong right foot and cut in from the left. I'm not quite sure what I've done this time, I feel like I've gone too far, or maybe it takes a team a while to gel.
Floppyaams said: Low CA players with high "meta" stats such as pace, acc, dribble and jumping is the way to go.

Have rebuilt my Dortmund team to have the second lowest wage bill at 700k a week. I have to say, I think I may have messed up as my team doesn’t seem great, despite having players that are high rated in the meta stats and fit the attribute profiles we use on python.
Tried this out on the new patch with some new instructions for the left sided forward.

Formation:
PL Table:
Player Stats:
Schedule:
Falbravv said: I made it with TOULOUSE in FRANCE, EUROPA CUP, and next year, CL. 5 CL in 7 years, but without the rules here. But for 2 first years it was near that.

All that with my own Genie ratings.


Interesting, that's a good achievement. I think we've all been able to do something similar, but I'm looking to maximise how cheaply it can be done.
I’m going to try and take a new job in my save and not spend over £5m on a player and don’t pay more than 50k per week salary. Reckon you can build a team with those rules and win the European Cup?
Do you guys believe there is a difference in how human players play on the match engine and how cpu players play in their simulated matches? It just seems there is no correlation between
Attributes to on pitch performances for non-human players.
Steelwood said: 1. We can't see their pros and cons on this screen
2. Tewelde is a better dribbler
3. IWs tend to perform better on the opposite side to their feet. Bilous is left-footed and Tewelde is right-footed

If you are playing IWs, I think you would perform better with Bilous on the right


Here they are with the pros and cons, forgot I'd taken those off to see stats.



keithb said: would need to see pics of both their attributes. Of course you will still have players who get better ratings that look worse on paper, in genie scout and with the python code. But overall the methods work

Bilous the under performing winger -

Tewelde the star -

Both playing as Inverted Winger Support on the Left
dzek said: Another way to decide on the selection filters for the players you buy is to look at the average attributes for the league you play in and create your own filters based on them. Keep in mind though that these will change from time to time so you will need to adjust them accordingly every 6 months I believe.

Go to Squad Planner > Report > Comparison

Hope this helped :)


That is a very helpful suggestion! I’m looking for a way to see how attributes influence match engine performances. Like looking at the chalkboard stats in game, the best performing players rarely have the attributes I believe to be necessary according to FM Arena testing.
keithb said: josh acheampong is a player who fits this imo. I think he peaked at 12.8 for me but was consistently one of my best players at rb. Was there better out there of course, but his speed and stamina were enough him to be a 7+ every season. His technicals were pretty poor.

See I’m having a great year with Faniel Tewelde being my best performing winger, but his key stats for that position aren’t as good as my other winger? I used genie scout and they both have the same consistency, it seems like I’m maybe missing something that influences match engine success. Python is getting me to the pond, but I can’t figure out how to drink.