nepejke Posted December 16, 2022 Share Posted December 16, 2022 (edited) Hello! Inspired by fm-arena testing attributes tables, I wanted to do my own testing. I pickuped team(Premier division England) and make few seasons holiday simulations(after season end I load save and start it again from the same date). Same tactic, same players, only one difference. In first example I decreased both Pace and Acceleration by 4, and increased Marking and Tackling by 4, then freeze the players by editor. In second example I decreased both Marking and Tackling by 4, and increased Pace and Acceleration by 4, then freeze the players by editor. Here is results: 1) first CB 6.99 season rating, second CB 6.66 season rating. 2)first CB 7.15 season rating, second CB 6.85 season rating. Sad to see all of that in the game, where all, especially here forum defenders like Xaw, or Davencid talking about realism. I understand, Pace and Acceleration usefull stats for every role, but then make this stats essential for every role and every player (green light on stats view), because today green stats for BPD Marking and Tackling useless compare to this. CB's with Tackling and Marking 9 and 10, playing better than Tackling 17 and 18 Marking CB's in Premier league ? Cmon Final player stats: 1 test: Marking 18 Tackling 17 Pace 11 Acceleration 10 2 test: Marking 10 Tackling 9 Pace 19 Acceleration 18 PS. here fm-arena attributes testing link, where you can find so much more disbalanced attributes like thesehttps://fm-arena.com/table/18-fm23-attribute-testing/ Edited December 16, 2022 by nepejke 1 Link to post Share on other sites More sharing options...
nepejke Posted December 16, 2022 Author Share Posted December 16, 2022 (edited) I wanted to see in this game that green attributes for players role has most significant impact on player perfomance, but now looks like pace and acceleration has much more impact on player performance even when this stats not mustly needed for the role. Edited December 16, 2022 by nepejke Link to post Share on other sites More sharing options...
Popular Post XaW Posted December 16, 2022 Popular Post Share Posted December 16, 2022 While I think you are not approaching this at a fair level (considering you are labeling me and @Daveincid as "forum defenders" without tagging us to let us know you did, not even sure what this means, but it feel derogatory somehow), this is clearly a flawed test. What attributes are more important cannot just be done like this. Here are my counter points without doing much thoughtwork: You come into this with a confirmation bias looking to be proved right. That mindset will just skew any test in the direction of your wish. You have run it ONCE for each setup. For this to come anywhere close to useful, you need to run it loads of times, establish margins of errors, etc. The level of statistical knowledge, or at least application, is sorely lacking here. Even if you did it and found something, you have only tested it FOR ONE TACTICAL SETUP. Pace will be FAR more important with a high line than a low block for example You only tested it with ONE TEAM. What attributes are most important can also relate to the skill of the player around the defenders. A very good DM can counter some attributes that are lacking in defenders. You fail to control pretty much any variable at all, other than what you changed. Players with great decisions and positioning would be far better at handling low pace than those who are also lacking in those attributes. This is not taken into account at all. So what I am saying is, your tests are not very good and you haven't proven anything. I also think you are entering this in bad faith. If you REALLY wanted to make a difference, you would have at the very least posted this in the bug tracker with a few saves showing you findings. You chose to post it in here with almost no detail and no proof other than a link to another site that also has flawed testing with many of the same issues I outlined above. This feels more like a cheap shot than a deep statistical analysis of the game. If you did that, I'd be the first to congratulate you on your findings and had hoped SI would take a deep look at it, but this is far from it. 10 Link to post Share on other sites More sharing options...
Daveincid Posted December 16, 2022 Share Posted December 16, 2022 1 Link to post Share on other sites More sharing options...
nepejke Posted December 16, 2022 Author Share Posted December 16, 2022 I expected something like this from you. Unfortunately I dont have enough time for further testing, but I am pretty sure this things works like I said. May be some1 who have more time will prove it to you. Link to post Share on other sites More sharing options...
XaW Posted December 16, 2022 Share Posted December 16, 2022 Well, that just shows clearly that you are not looking to contribute, but rather the opposite. /Thread 1 Link to post Share on other sites More sharing options...
Recommended Posts