mackforever said: Not in the version I downloaded. D-Line is set on Higher. Expand
I have just downloaded here from this page to test and defensive line is standard.
JamesFox100 said: first time I have ever achieved invicible season in the EPL is with this tactic, its something I have been trying to achieved for 3 years in FM! Expand
Gratz! I really think there is lots of potential in this tactic, this was just a raw version to see if the concept holds. I have just started testing some improvements on it and I believe it will get better.
End of season and we are the champions! It was an extremely easy season on the league, as well as Papa's John Cup (U23 weren't as strong as I feared). Considering how well my team did against opponents of League One, I think next season will be as easy as this one.
Nothing very special happened in the past ten matches. I got 9 days between matches and was able to give vacation to most players before the final of Papa's John Cup, and also got enough time to train preparation for all set pieces.
My wage budget for next season is €1.5M (per year), or triple of current. That means I can finally hire more scouts and buy from other teams, instead of relying only on free transfers. My goal for next season is to be promoted and maybe upgrade my training facilities, since board keep refusing it time and time again.
fm arena user said: And if You looking for CM and he has all this atributes high but laks Vision, Decision and Passing You still buy such player or player who is high with atributes above but laks speed Acceleration and stamina? Do You know what i mean? Are this atributes so overpowered they more important than obvious CM atributes? Expand
I always go for the faster ones. However, my opinion matters very little when we have two tables that tested players attributes, which are definitely a better input than my personal feeling.
Mark said: OK so I had a quick play with the ratings files. The first observation is that you need to balance the ratings files. I learned this from last years attempt. What do I mean by this, well have a look at the screenshot below. There should not be major discrepancies between positions ie you should be getting most positions covered in the top 50 or so top players.
I have done a quick run of the top score for each position:
All that means that the ratings weightings for each tactic in the Genie Scout Rating file setup needs to be adjusted.
Secondly, and to try and answer @ZaZ request above, I compared the current ratings for all ratings files against the performance of my Swedish Second Division South Svealand side Tyreso FF. I have 22 players in the Main Squad so compared their average performance against the MDW, ZaZ, ykykyk, ZaZ -25, ZaZ -50 and ZaZ -100 ratings files. I have only used ratings for the position selected.
Here are the results:
Disclaimer: I have used players out of the positions identified in the results, but in the main they have played the positions indicated. In addition I have used 2 main tactics, Dragons Breath that uses an MC instead of AMC, and ZaZ Blue DM, but given the ratings files don't all have MC ratings I just used AMC.
To answer @ZaZ question the ZaZ-100 ended up equal with MDW in picking 8 of the best 11 positional players. It does need balancing to make it more usable though. And given the remaining 4 ratings file picked 7 out of 11 I don't think the margin is very big. All are definitely usable.
Now if we look at average % difference from the best rated team the order is MDW, ykykyk, ZaZ -50, ZaZ -25, ZaZ, ZaZ -100.
Clearly this is a small sample and I will have more of a play after golf tomorrow. But I think if you wanted to use one of the ratings files above, I would go with the ZaZ -50. Expand
Thanks for the thorough analysis, Mark. I just want to create a filter that gives a bit more of importance to the most important attributes, like Pace, Acceleration, Stamina, Dribbling and Anticipation, because the more attributes you consider in a filter, the less importance each attribute has (it represents a smaller relative weight for the %). Doing that, it allows you to find good players for cheaper prince, since cost of players is directly linked to CA and reputation.
fm arena user said: What is in Your opinion 4 most important atributes for each position? in very simple words without any exel spreadshets please. I followed You thread last year and search players with Your guide, Pace, Accelertion, Flair, Decisions As my clubs DNA. Would this attributes be as important as last year? Expand
Pace, Acceleration, Stamina and Dribbling/Anticipation.
ykykyk05251 said: HI, Thank you for your interest in our research results. It is necessary to make some supplementary explanations to the research results so as not to be misleading. Stamina, in our system, we keep the players full in every game, and all players play the entire game. In the actual game, for those substitute players who occasionally play, stamina should not be as important as the results indicate. In a dense schedule, the stamina of the key players may be more important than the research results. Of course, it is related to the team's rotation strategy. Strength and Aggression, in order to simplify the entire training process, we actually discarded the injury game. Higher aggression will make the players more likely to get injured. This negative factor is not included in our experiment. In addition, our research has found that strength has the effect of avoiding serious injuries to a certain extent, so the importance of strength should be higher than that on the table. Punching, in fact, this attribute shows some negative effects in our research, and the weight in the table should be a negative number. However, due to a bug in our system, there are some problems with the processing of negative attributes, so accurate weight values are not obtained. 0 is manually set. Expand
Thanks for the explanation!
I believe the first point (setting condition to max before matches), should be a goal of any manager with proper rotation and resting after matches. It is achievable during most of the season except a couple of tight weeks, so it's not a big deal and I believe the conclusion about stamina still stands in real gameplay. About the second point (removing injuries), it can somehow be managed with a balanced squad. Removing injuries and setting condition to maximum are ways to simulate the human behaviour of rotating squad, because assistant manager is pretty bad at that.
Anyway, thanks for the hard work on the experiment! It gave us some nice input that will help lots of managers find the best players to overachieve in their games.
@Mark, I made three poor man's version and I would like to know your opinion in which one gives the best results. Can you test the three and tell me which one you like the most? Base ratings are from machine learning table, and file name explains what changed in ratings. All of them try to focus on most important attributes in some way.
Version 1: all ratings minus 25 (removes lower ratings and increases % gap of remaining) Version 2: all ratings under 50 minuts 25 (removes lower ratings and keeps gap of top) Version 3: all ratings minus difference to 100 (removes lower ratings and increases absolute gap of remaining) * Gap means the difference in rating between two attributes. For example, 25 to 40 have a gap of 15.
Mark said: I think this needs a little work to balance out the ratings. As it sits it favours Strikers and AMs too much and there are no GKs anywhere to be seen. The weightings have only been developed by position and not across all positions so need adjusting in terms of weight for each position. I would also like to derive the positions not covered.
I will have a play and update it next week. I have a very busy weekend unfortunately. Expand
I'll do my poor man's version too, with only the most important attributes, so it puts more importance on those attributes and gives good cheaper players.
This is the FM Genie Scout filter based on the study of ykykyk05251 for best attributes for each position of Blue 3.0 RPM (aka Blue DM). The study used machine learning to get the weights.
Keep in mind that two attributes of the table have ranges from -100 to 0 in FM Genie Scout instead of 0 to 100: eccentricity and punching tendency. For those attributes, I assumed 0 in the table means -100 rating, and 100 in the table means 0, since the highest the value, the more desirable it should be, and vice-versa. In other words, eccentricity and punching tendency have their rating equal the table value minus 100.
Mark said: @Zippo if you look at my spreadsheet conversion I don't think what you are saying is correct Expand
You are actually right, I didn't notice numbers were different! I actually went to the chinese website and asked someone to translate, and they provided me that translation. I'll check which one is correct and update it.
P.S.: Meanwhile, I'll update it with @Mark's table, which seems to be correct. After checking properly, DC and WB had the same table, so it was probably an innocent mistake there.
Now at 36 of 46 matches in the league. Title is already won, it's just a matter of time. Unfortunatelly, I got kicked out from FA Cup, but I'm in the semi-finals of Papa's John Cup. Since I survived long enough in the cups, finances are going really well, and the board allowed me to improve youth facilities, recruitment, coaching and to hire one extra physio (with 17 in physio). Unfortunatelly, they keep denying me training facilities, so I hope they propose it themselves once we get promoted.
I have decided to keep the same squad without hiring anyone new. To be fair, I just got lazy to scout new players and do trials. When you are in the lower leagues, it's just better to use trials instead, 30 players at a time. Next season I will probably start using scouts to help revealing player attributes, but right now it's too inefficient.
Now, I'm gonna finish my experiments with Blue with W-Su > DW-Su, since I often change tackle harder to ease off tackles when my players get a yellow card. That is not possible for DW-Su, so I'm testing all flank roles to see which one does better (including mezzala and carillero), and W-Su is currently doing the best.
I'm also considering if I should use FM Genie Scout to test the filters, or if I should avoid it so I keep being in the blind for hidden attributes. I prefer being in the blind so I focus more on the current match, because when I see players hidden attributes I often play thinking too much about the next seasons.
Zippo said: If you ask my opinion then I would not blindly follow it.
For example, if look at the central defenders spreadsheet then you'll see that "Long Throws" and "Crossing" has more weight than "Heading" but we know that Central Defenders don't take throw-ins and don't do crosses but they take a lot of headers so I don't understand how "Long Throws" and "Crossing" can be more important than Heading for Central Defenders... that doesn't make sense at all Expand
I understand your point, and that happens because they used machine learning to define the scores. It shouldn't be blindly followed, but it definitely has some use. Maybe it can be adjusted by common sense.
P.S.: I might make a poor man's version of that filter considering only green attributes (50+), which would lead to good players for cheaper cost.
I'll post below the study of ykykyk05251, from a basketball manager development team, about the most important attributes for each position in Blue 3.0 RPM (aka Blue DM). Keep in mind that not everything might translate to other tactics, but it's only natural that a good part of it actually does. Great thanks to ahstzl1989 for translating the entire text from chinese. If anyone wants to see the original post, visit this thread from a chinese FM forum.
Basic description. 1, we are a basketball manager game development team, the development process reference study a lot of FM settings, including the game engine. We have learned a lot about the game from the bursting shed, and the purpose of this post is to give back to the community
2, in order to analyze the mechanism of FM's game engine, as well as the degree of science, we designed a system for measuring the degree of influence of each player's attributes on the final victory or defeat in FM.
3. fm-arena.com gives a preliminary test of the relationship between player attributes and wins and losses in FM2021 and FM2022, which is of some reference value and inspiration for our work However, as it is only the non-professional work of amateurs, from the point of view of rigour, there are the following problems 1) It only gives the results of which attributes have a greater and lesser impact on winning and losing for all players, but in reality, the key attributes are obviously different for different positions, and the results of the test clearly show that only those attributes that are important for all positions will be more important, while those attributes that are important only for certain positions will be less important in the test. For example, its test results show that shooting has almost no effect on winning, while explosive power has a big effect on winning and losing. 2) Its test sample is insufficient, its test for each attribute was only carried out for about 900 matches simulated, but for a normally randomly distributed sequence, in general it needs to be randomized at 10,000 times before it converges relatively well to the mean. (3) The effect of attributes on wins and losses is non-linear, and the test only deducts 4 points from the attribute to investigate whether it has an effect on wins and losses, but sometimes, just because 4 points have no effect does not mean that 8 points also have no effect, and it is also possible that 2 points have an effect that is close to 4 points. 4) There is a correlation between the impact of attributes on wins and losses, and its test of only changing one attribute at a time to test the impact on wins and losses can be interfered with by the correlation. For example, a breakthrough can lie on the ball to change direction and accelerate past a player, relying on physicality but not on discography, or it can rely on discography. If a player is good at both physicality and discography, then reducing his discography will not significantly affect his breakthrough effectiveness, as he can accelerate past people with a lie-back change of direction instead of discography. 5) Different tactics require different attributes for each position of the player, and it is not rigorous to talk about which attributes have an impact on winning or losing away from specific tactics.
4, we use artificial intelligence to conduct attribute importance research, artificial intelligence technology used a deep neural network similar to Go Alpahgo. We fixed the tactic as the strongest tactic on fm-arena.com, ZaZ-Blue DM, for this experiment in order to analyse which attributes of players in each position have a greater impact on wins and losses under this tactic. We built an artificial intelligence system and trained the AI to design the best attribute assignment for each position's performance with the same CA for each player in the whole team. The approximate process was as follows: first, each attribute at each position was assigned the same value, then the neural network would try to change the value of certain attributes at certain positions (keeping CA constant) to see if the change had a positive or negative impact on wins and losses, and thus iterate over the multi-layer network to know what the AI thought was the best combination of 11 players. We obtained the convergence results by training roughly 140 machines for 3 weeks and simulating 40 million games.
5. To verify whether the best 11-player attribute allocation given by the AI is really the best, we conducted another result validation test Three groups of teams were designed, one with evenly distributed player attributes, another with the best 11-player attribute assignment given manually by experienced players based on their gaming experience, and the last group given by the AI. Put into a test league for 100,000 games, the AI's solution was significantly better than the human players' solution and far superior to the even distribution.
Conclusion. For the ZaZ-Blue DM tactic, for each position we obtained the following levels of importance for the attributes (the values were normalized by 5 for ease of viewing, which should be sufficient accuracy for the game).
Significance of use. 1, players can get the key attributes of each position according to the above table, 100 is the most critical attribute, while 1 is the least critical attribute, so as to guide the selection of materials. 2. The values in the above table can also be used as attribute weights to calculate the weighted average value of the attributes and then calculate the "Tactical True CA" Tactical True CA = Weighted average value of attributes * 20 - 121 If the Tactical True CA is higher than the Player CA it means that the player is a good fit for the ZaZ-Blue DM tactics, the higher the Tactical True CA the better the fit. This is used for player selection For the convenience of future players playing in this way, the table we have given above is arranged according to the order in which the attributes are displayed in the player's interface, even if you don't write the program, you can also get the "Tactical True CA" in excel after quickly and manually entering the values of the attributes in three rows 3. For guidance on training, the attributes take up CA, so in order to train the players with the highest tactical true CA, we can train the attributes with the highest cost effectiveness The cost effectiveness of each attribute for each position can be measured by the attribute Tactical True CA Weight / Attribute CA Weight, the higher the value, the higher the attribute will make the player's attribute Tactical True CA higher if it increases the same CA.
Main limitations. The current experiments are costly and non-replicable, engine versions are changed, tactics are changed, and the AI needs to be retrained without migration learning, so our next phase will focus on migration learning, where the results obtained from iterative training in the case of a certain version of a certain tactic are used as the basis for a new engine and a new tactic, and rapid iterations are made to obtain results in a new environment.
A FM Genie Scout rating based on the findings of the study can be downloaded here.
I have updated the important attributes in first post. I also created a separated thread for it in the general section, which @Zippo coult kindly pin if possible.
Mark said: I have gone to the site and done my best to interpret the Chinese into English. I am thinking we should have a go at doing a Genie Scout filter based on their results. First I will provide a summary of their results of attribute importance by position for ZaZ Blue. Then I will open a separate post with the converted text. There are quite few things that don't seem to convert easily. But here goes:
Expand
Thanks, Mark! Gonna check it and make the filter to add to first post. I will make sure to reference the authors.
I have just downloaded here from this page to test and defensive line is standard.
JamesFox100 said: first time I have ever achieved invicible season in the EPL is with this tactic, its something I have been trying to achieved for 3 years in FM!
Gratz! I really think there is lots of potential in this tactic, this was just a raw version to see if the concept holds. I have just started testing some improvements on it and I believe it will get better.
Nothing very special happened in the past ten matches. I got 9 days between matches and was able to give vacation to most players before the final of Papa's John Cup, and also got enough time to train preparation for all set pieces.
My wage budget for next season is €1.5M (per year), or triple of current. That means I can finally hire more scouts and buy from other teams, instead of relying only on free transfers. My goal for next season is to be promoted and maybe upgrade my training facilities, since board keep refusing it time and time again.
https://fm-arena.com/thread/1949-fm22-positional-filters-what-are-the-best-attributes-for-each-position/page-2/#comment-11020
I always go for the faster ones. However, my opinion matters very little when we have two tables that tested players attributes, which are definitely a better input than my personal feeling.
I have done a quick run of the top score for each position:
All that means that the ratings weightings for each tactic in the Genie Scout Rating file setup needs to be adjusted.
Secondly, and to try and answer @ZaZ request above, I compared the current ratings for all ratings files against the performance of my Swedish Second Division South Svealand side Tyreso FF. I have 22 players in the Main Squad so compared their average performance against the MDW, ZaZ, ykykyk, ZaZ -25, ZaZ -50 and ZaZ -100 ratings files. I have only used ratings for the position selected.
Here are the results:
Disclaimer: I have used players out of the positions identified in the results, but in the main they have played the positions indicated. In addition I have used 2 main tactics, Dragons Breath that uses an MC instead of AMC, and ZaZ Blue DM, but given the ratings files don't all have MC ratings I just used AMC.
To answer @ZaZ question the ZaZ-100 ended up equal with MDW in picking 8 of the best 11 positional players. It does need balancing to make it more usable though. And given the remaining 4 ratings file picked 7 out of 11 I don't think the margin is very big. All are definitely usable.
Now if we look at average % difference from the best rated team the order is MDW, ykykyk, ZaZ -50, ZaZ -25, ZaZ, ZaZ -100.
Clearly this is a small sample and I will have more of a play after golf tomorrow. But I think if you wanted to use one of the ratings files above, I would go with the ZaZ -50.
Thanks for the thorough analysis, Mark. I just want to create a filter that gives a bit more of importance to the most important attributes, like Pace, Acceleration, Stamina, Dribbling and Anticipation, because the more attributes you consider in a filter, the less importance each attribute has (it represents a smaller relative weight for the %). Doing that, it allows you to find good players for cheaper prince, since cost of players is directly linked to CA and reputation.
Pace, Acceleration, Stamina and Dribbling/Anticipation.
Thank you for your interest in our research results.
It is necessary to make some supplementary explanations to the research results so as not to be misleading.
Stamina, in our system, we keep the players full in every game, and all players play the entire game. In the actual game, for those substitute players who occasionally play, stamina should not be as important as the results indicate. In a dense schedule, the stamina of the key players may be more important than the research results. Of course, it is related to the team's rotation strategy.
Strength and Aggression, in order to simplify the entire training process, we actually discarded the injury game. Higher aggression will make the players more likely to get injured. This negative factor is not included in our experiment. In addition, our research has found that strength has the effect of avoiding serious injuries to a certain extent, so the importance of strength should be higher than that on the table.
Punching, in fact, this attribute shows some negative effects in our research, and the weight in the table should be a negative number. However, due to a bug in our system, there are some problems with the processing of negative attributes, so accurate weight values are not obtained. 0 is manually set.
Thanks for the explanation!
I believe the first point (setting condition to max before matches), should be a goal of any manager with proper rotation and resting after matches. It is achievable during most of the season except a couple of tight weeks, so it's not a big deal and I believe the conclusion about stamina still stands in real gameplay. About the second point (removing injuries), it can somehow be managed with a balanced squad. Removing injuries and setting condition to maximum are ways to simulate the human behaviour of rotating squad, because assistant manager is pretty bad at that.
Anyway, thanks for the hard work on the experiment! It gave us some nice input that will help lots of managers find the best players to overachieve in their games.
Version 1: all ratings minus 25 (removes lower ratings and increases % gap of remaining)
Version 2: all ratings under 50 minuts 25 (removes lower ratings and keeps gap of top)
Version 3: all ratings minus difference to 100 (removes lower ratings and increases absolute gap of remaining)
* Gap means the difference in rating between two attributes. For example, 25 to 40 have a gap of 15.
Anyone else's opinion would also be appreciated!
I will have a play and update it next week. I have a very busy weekend unfortunately.
I'll do my poor man's version too, with only the most important attributes, so it puts more importance on those attributes and gives good cheaper players.
Keep in mind that two attributes of the table have ranges from -100 to 0 in FM Genie Scout instead of 0 to 100: eccentricity and punching tendency. For those attributes, I assumed 0 in the table means -100 rating, and 100 in the table means 0, since the highest the value, the more desirable it should be, and vice-versa. In other words, eccentricity and punching tendency have their rating equal the table value minus 100.
Excel file
https://docs.google.com/spreadsheets/d/1muBAU22TzHTO9_zdNYx7ubePAAUtgVQg/edit?usp=sharing&ouid=106197286223391117795&rtpof=true&sd=true
Thanks for the tables! I took the freedom to put them all together into a single image.
Just updated.
You are actually right, I didn't notice numbers were different! I actually went to the chinese website and asked someone to translate, and they provided me that translation. I'll check which one is correct and update it.
P.S.: Meanwhile, I'll update it with @Mark's table, which seems to be correct. After checking properly, DC and WB had the same table, so it was probably an innocent mistake there.
I have decided to keep the same squad without hiring anyone new. To be fair, I just got lazy to scout new players and do trials. When you are in the lower leagues, it's just better to use trials instead, 30 players at a time. Next season I will probably start using scouts to help revealing player attributes, but right now it's too inefficient.
Now, I'm gonna finish my experiments with Blue with W-Su > DW-Su, since I often change tackle harder to ease off tackles when my players get a yellow card. That is not possible for DW-Su, so I'm testing all flank roles to see which one does better (including mezzala and carillero), and W-Su is currently doing the best.
I'm also considering if I should use FM Genie Scout to test the filters, or if I should avoid it so I keep being in the blind for hidden attributes. I prefer being in the blind so I focus more on the current match, because when I see players hidden attributes I often play thinking too much about the next seasons.
Anyway, I hope all of you have a great new year!
For example, if look at the central defenders spreadsheet then you'll see that "Long Throws" and "Crossing" has more weight than "Heading" but we know that Central Defenders don't take throw-ins and don't do crosses but they take a lot of headers so I don't understand how "Long Throws" and "Crossing" can be more important than Heading for Central Defenders... that doesn't make sense at all
I understand your point, and that happens because they used machine learning to define the scores. It shouldn't be blindly followed, but it definitely has some use. Maybe it can be adjusted by common sense.
P.S.: I might make a poor man's version of that filter considering only green attributes (50+), which would lead to good players for cheaper cost.
Basic description.
1, we are a basketball manager game development team, the development process reference study a lot of FM settings, including the game engine.
We have learned a lot about the game from the bursting shed, and the purpose of this post is to give back to the community
2, in order to analyze the mechanism of FM's game engine, as well as the degree of science, we designed a system for measuring the degree of influence of each player's attributes on the final victory or defeat in FM.
3. fm-arena.com gives a preliminary test of the relationship between player attributes and wins and losses in FM2021 and FM2022, which is of some reference value and inspiration for our work
However, as it is only the non-professional work of amateurs, from the point of view of rigour, there are the following problems
1) It only gives the results of which attributes have a greater and lesser impact on winning and losing for all players, but in reality, the key attributes are obviously different for different positions, and the results of the test clearly show that only those attributes that are important for all positions will be more important, while those attributes that are important only for certain positions will be less important in the test. For example, its test results show that shooting has almost no effect on winning, while explosive power has a big effect on winning and losing.
2) Its test sample is insufficient, its test for each attribute was only carried out for about 900 matches simulated, but for a normally randomly distributed sequence, in general it needs to be randomized at 10,000 times before it converges relatively well to the mean.
(3) The effect of attributes on wins and losses is non-linear, and the test only deducts 4 points from the attribute to investigate whether it has an effect on wins and losses, but sometimes, just because 4 points have no effect does not mean that 8 points also have no effect, and it is also possible that 2 points have an effect that is close to 4 points.
4) There is a correlation between the impact of attributes on wins and losses, and its test of only changing one attribute at a time to test the impact on wins and losses can be interfered with by the correlation. For example, a breakthrough can lie on the ball to change direction and accelerate past a player, relying on physicality but not on discography, or it can rely on discography. If a player is good at both physicality and discography, then reducing his discography will not significantly affect his breakthrough effectiveness, as he can accelerate past people with a lie-back change of direction instead of discography.
5) Different tactics require different attributes for each position of the player, and it is not rigorous to talk about which attributes have an impact on winning or losing away from specific tactics.
4, we use artificial intelligence to conduct attribute importance research, artificial intelligence technology used a deep neural network similar to Go Alpahgo.
We fixed the tactic as the strongest tactic on fm-arena.com, ZaZ-Blue DM, for this experiment in order to analyse which attributes of players in each position have a greater impact on wins and losses under this tactic.
We built an artificial intelligence system and trained the AI to design the best attribute assignment for each position's performance with the same CA for each player in the whole team.
The approximate process was as follows: first, each attribute at each position was assigned the same value, then the neural network would try to change the value of certain attributes at certain positions (keeping CA constant) to see if the change had a positive or negative impact on wins and losses, and thus iterate over the multi-layer network to know what the AI thought was the best combination of 11 players.
We obtained the convergence results by training roughly 140 machines for 3 weeks and simulating 40 million games.
5. To verify whether the best 11-player attribute allocation given by the AI is really the best, we conducted another result validation test
Three groups of teams were designed, one with evenly distributed player attributes, another with the best 11-player attribute assignment given manually by experienced players based on their gaming experience, and the last group given by the AI. Put into a test league for 100,000 games, the AI's solution was significantly better than the human players' solution and far superior to the even distribution.
Conclusion.
For the ZaZ-Blue DM tactic, for each position we obtained the following levels of importance for the attributes (the values were normalized by 5 for ease of viewing, which should be sufficient accuracy for the game).
* Translated table provided by @Mark.
Significance of use.
1, players can get the key attributes of each position according to the above table, 100 is the most critical attribute, while 1 is the least critical attribute, so as to guide the selection of materials.
2. The values in the above table can also be used as attribute weights to calculate the weighted average value of the attributes and then calculate the "Tactical True CA"
Tactical True CA = Weighted average value of attributes * 20 - 121
If the Tactical True CA is higher than the Player CA it means that the player is a good fit for the ZaZ-Blue DM tactics, the higher the Tactical True CA the better the fit.
This is used for player selection
For the convenience of future players playing in this way, the table we have given above is arranged according to the order in which the attributes are displayed in the player's interface, even if you don't write the program, you can also get the "Tactical True CA" in excel after quickly and manually entering the values of the attributes in three rows
3. For guidance on training, the attributes take up CA, so in order to train the players with the highest tactical true CA, we can train the attributes with the highest cost effectiveness
The cost effectiveness of each attribute for each position can be measured by the attribute Tactical True CA Weight / Attribute CA Weight, the higher the value, the higher the attribute will make the player's attribute Tactical True CA higher if it increases the same CA.
Main limitations.
The current experiments are costly and non-replicable, engine versions are changed, tactics are changed, and the AI needs to be retrained without migration learning, so our next phase will focus on migration learning, where the results obtained from iterative training in the case of a certain version of a certain tactic are used as the basis for a new engine and a new tactic, and rapid iterations are made to obtain results in a new environment.
A FM Genie Scout rating based on the findings of the study can be downloaded here.
Thanks, Mark! Gonna check it and make the filter to add to first post. I will make sure to reference the authors.
source: https://www.playgm.cn/thread-943500-1-1.html
Hey! It would help a lot if I could understand a single word there. Hopefully it will help others. Thanks for bringing it up, anyway!
P.S.: I would be grateful if anyone that understands chinese could translate it for us.
It's on the first post.