Royaltea Posted August 10, 2022 Share Posted August 10, 2022 (edited) Introduction I'm interested in exploring the various effects of different aspects of training over player's attributes, to that end I've concocted the following test settings. Test Settings I've set up my B team as Brest (French League 1) to test various training schedules over a period of an entire season using the following method and wanted to share the results with you. Using the in-game editor I've edited the B team to have the following: · Attributes: Each player visible attributes to be a 10. · Hidden Attributes: Each player hidden attributes to be a 10. · Player Positions: Players that could play multiple positions turned into pure position by setting 1 to all other positions but the natural position (20). · Age: Each player's age changed to 18 (Changed date of birth to 1/1/2003). · Nationality: Changed each player's nationality to French (to avoid any Adaptability effects). · Weaker Foot: Removed any value in weaker foot (All the player have one very strong foot – 20, very weak foot – 1). · Injuries: I reset any existing injuries but made sure the players have similar sharpness and fatigue. · Current Ability: I've edited the Current Ability to be the RCA (Recommended Current Ability) advised by the in-game editor (ranged between 63 for a Striker and 68 for a Centre-back). · Potential Ability: I've edited the Potential Ability of each player to be x2 the Current Ability. · Training Facilities: Brest have Good Training Facilities (12) and Good Youth Facilities (12). · Coaches: All training categories have 2.5 stars (Fairly Good) at least. · Balanced squad: 3 GK, 2 DR, 4 DC, 3 DL, 2 DM, 3 MC, 2 AMC, 2 AMR, AML, 2 ST. A balanced squad in terms of positions. · Transfers/Loans/Contracts: Players were set unavailable for loan, they were also given the same salary ($60k p/a) for the same contract duration (5 years). How do I test a training schedule? I take over the B team general and individual team training, set up a schedule for the team to train over a period of a season and go on holiday. Because the B team have no games scheduled (they do play, it just doesn't show on calendar) this training setup don't get spoiled by the occasional cup match reschedule. Tested Training Strategies Training Strategy Times Tested Avg. Overall Growth in Attributes Full Rest 1 -104 Delegate to Assistant Manager 1 +146 Test #1: 5 Endurance, 5 Overall, 6 Recovery 1 +74 Test #2: 3 Quickness, 3 Physical, 5 Defensive Engaged, 3 Attacking Patient 1 +188 Test #1: 5 ENDURANCE, 4 OVERALL Expectations: I am expecting major improvements in Stamina, Natural Fitness and Work Rate alongside mild improvements in all attributes depend on positions. Results: Most improved attributes were: Strength (+19), Stamina (+16), Jumping Reach (+12), Balance (+9), Anticipation (+9), Work Rate (+8) Most declined attributes were: Off the Ball (-7), Finishing (-6), Marking (-6), Bravery (-6) Overall growth in attributes: (+74) Name Age Position CA PA Prof Inj Pr Acc Agi Bal Jum Nat Pac Sta Str Agg Ant Bra Cmp Cnt Dec Det Fla Ldr OtB Pos Tea Vis Wor Cor Cro Dri Fin Fir Fre Hea Lon L Th Mar Pas Pen Tck Tec Sébastien Cibois 19 GK 73 142 1 1 1 Isaïa Hubert 19 GK 71 142 1 1 1 Loïck Massé 19 GK 70 140 1 1 1 1 Kylian Le Her 19 D (R) 73 136 1 1 1 1 1 -1 -1 1 Mathéo Da Silva 19 D (R) 74 136 1 1 1 1 1 -1 1 -1 Brieg Le Bars 19 D (C) 74 136 1 1 1 1 1 1 -1 -1 Noé Sow 19 D (C) 71 136 -1 1 1 1 1 -1 -1 -1 -1 1 Josué Escartin 19 D (C) 75 138 1 1 1 -1 1 1 -1 -1 -1 Ali Dicko 19 D (C) 71 136 1 1 -1 -1 -1 Kévin Galisson 19 D (L) 72 136 1 1 1 1 -1 Tom Flamanc 19 D (L) 71 136 1 1 1 -1 Téva Lossec 19 D (L) 72 136 1 1 1 1 1 -1 Kilian Liri 19 DM 68 132 1 -1 Pierre Jouan 19 DM 69 132 1 1 1 1 -1 1 1 Rémi Maugain 19 M (C) 70 128 2 1 1 1 1 1 1 Kurtis Chadet 19 M (C) 67 128 1 1 1 1 1 Abdourahmane Thior 19 M (C) 69 128 1 1 1 1 -3 1 1 1 1 Djibril Konté 19 AM (R) 69 130 1 1 -2 1 1 -1 1 -1 -1 Axel Camblan 19 AM (R) 66 130 1 1 -1 1 -1 -1 Loïck Piquionne 19 AM (C) 67 122 1 1 1 1 1 1 1 1 -1 Aymeric Faurand-Tournaire 19 AM (C) 65 122 1 1 1 1 1 Rafiki Saïd 19 AM (L) 66 130 1 1 1 1 -1 Samuel Genty 19 AM (L) 70 130 1 1 1 1 1 Lucas Gélébart 19 ST (C) 67 126 1 1 1 1 1 -1 -1 Chris Basakila Bazungula 19 ST (C) 66 126 1 1 -2 -1 -1 Sum 2 -1 2 2 9 12 7 1 16 19 0 9 -6 1 4 5 0 2 0 -7 1 0 2 8 0 -3 0 -6 1 0 0 1 0 -6 1 0 -2 1 In the table above you can see the difference between attribute at the start of the season (12/7/21) and end of the season (22/05/22). A positive value means the attribute increased, a negative value means the attribute decreased (due to injuries or training effect). There is also a sum at the bottom. Discussion, conclusion and follow-up: I was surprised to see such a massive improvement in strength, jumping reach and balance as they were not specifically trained. My hypothesis is that 1. The players naturally grow stronger and taller because of young age. 2. There is some correlation between Endurance and Resistance sessions. It's also interesting to see that the defending unit declined in terms of attacking attributes (such as off the ball -7, finishing -6), and the attacking unit declined in terms of defensive attributes (marking -6). Also interesting to see two players got an increase in professionalism despite not having any mentors. One player injury proneness decreased by 1. However, due to injuries Overall Stamina increased only by 16 and Bravery got an overall decrease of 6. Follow-up test would be changing Endurance session to Resistance and compare the results. Test #2: 3 Quickness, 3 Physical, 5 Defensive Engaged, 3 Attacking Patient Expectations: I am expecting major improvements in Pace, Agility and Acceleration (Physical, Quickness), as well as Marking, Tackling, Aggression, Anticipation, Decisions, Positioning (Defending Engaged, Attacking Patient). Results: Name Age Position CA PA Prof Inj Pr Acc Agi Bal Jum Nat Pac Sta Str Agg Ant Bra Cmp Cnt Dec Det Fla Ldr OtB Pos Tea Vis Wor Cor Cro Dri Fin Fir Fre Hea Lon L Th Mar Pas Pen Tck Tec Sébastien Cibois 19 GK 79 142 1 1 1 1 Isaïa Hubert 19 GK 73 142 1 1 Loïck Massé 19 GK 71 140 1 Kylian Le Her 19 D (R) 74 136 1 1 1 1 1 1 1 1 1 Mathéo Da Silva 19 D (R) 73 136 1 1 1 1 1 1 1 Brieg Le Bars 19 D (C) 71 136 1 1 Noé Sow 19 D (C) 76 136 1 1 1 1 1 1 1 1 1 Josué Escartin 19 D (C) 73 138 1 1 1 Ali Dicko 19 D (C) 78 136 1 1 1 1 1 1 1 1 -1 1 1 Kévin Galisson 19 D (L) 73 136 1 1 1 1 Tom Flamanc 19 D (L) 76 136 1 1 1 1 1 1 1 1 1 1 Téva Lossec 19 D (L) 73 136 1 1 1 1 1 1 Kilian Liri 19 DM 74 132 1 1 1 1 1 1 1 1 1 Pierre Jouan 19 DM 74 132 1 1 1 1 1 1 1 1 1 1 1 1 Rémi Maugain 19 M (C) 67 128 -2 1 1 1 1 1 1 Kurtis Chadet 19 M (C) 71 128 1 1 1 1 1 1 1 1 1 1 1 1 Abdourahmane Thior 19 M (C) 68 128 1 1 1 1 1 1 1 Djibril Konté 19 AM (R) 75 130 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 Axel Camblan 19 AM (R) 71 130 1 1 1 1 1 1 1 1 1 1 1 Loïck Piquionne 19 AM (C) 68 122 1 1 1 1 1 1 1 1 1 1 1 Aymeric Faurand-Tournaire 19 AM (C) 66 122 1 1 1 1 1 1 1 1 1 Rafiki Saïd 19 AM (L) 74 130 1 1 1 1 1 1 1 1 1 1 1 Samuel Genty 19 AM (L) 70 130 1 1 1 1 1 1 Lucas Gélébart 19 ST (C) 70 126 2 1 1 1 1 1 1 1 1 1 Chris Basakila Bazungula 19 ST (C) 68 126 1 1 1 1 1 3 -2 9 18 15 17 11 11 19 15 0 8 0 4 6 6 0 6 0 1 4 0 5 3 0 2 6 0 3 0 0 5 2 4 2 0 4 2 In the table above you can see the difference between attribute at the start of the season (12/7/21) and end of the season (22/05/22). A positive value means the attribute increased, a negative value means the attribute decreased (due to injuries or training effect). There is also a sum at the bottom. I've also used colors this to emphasize improvements. Most improved attributes were: Stamina (+19), Agility (+18), Jumping Reach (+17), Strength (+15), Nature Fitness (+11), Pace (+11), Acceleration (+9) Most declined attributes were: Notable attributes: Aggression (0) Overall growth in attributes: (+188) Discussion, conclusion and follow-up: This training schedule seem to have a much greater impact on the players than test #1. Massive improvements in all areas of players physicality. Surprisingly stamina improved more with this schedule than the "pure" Endurance from the previous test. Also no decline whatsoever (is that even a good thing? I'd rather have strikers that can't mark). I think that the most baffling result is that defending unit's aggression attribute hasn't changed despite having dedicated 7 (!) training sessions (Defending Engaged), and 3 training sessions (Attacking Patient) as secondary impact for an entire season. Because of that I had to recheck the sessions descriptions: Follow-up test would be to see if aggression is even trainable, using 21 Defensive Engaged schedules. Any comments, ideas, suggestions and critique are welcomed. Edited August 11, 2022 by Royaltea updated metrics Link to post Share on other sites More sharing options...
herne79 Posted August 10, 2022 Share Posted August 10, 2022 I know SI are always interested in studies such as this and once finished you could create a thread / post your data and results into the Bugs forum (I know it’s not a bug but that’s the place to upload such information). My only observation for now concerns the design ethos and intent behind the Training Module - it’s designed to be realistic. Because of that, once we start introducing unrealistic inputs such as changing things in the editor or creating synthetic training schedules, we run the risk of receiving unrealistic outputs. So if some results are surprising, whilst it may show an area of concern within the module it’s probably more to do with the complexities of the module and the realistic basis. Because of that, trying to draw conclusions is perhaps coming at it from the wrong angle - instead output the results and then, as said above, feed them into SI to analyse and let them draw conclusions 👍. Link to post Share on other sites More sharing options...
Royaltea Posted August 10, 2022 Author Share Posted August 10, 2022 (edited) 1 hour ago, herne79 said: I know SI are always interested in studies such as this and once finished you could create a thread / post your data and results into the Bugs forum (I know it’s not a bug but that’s the place to upload such information). My only observation for now concerns the design ethos and intent behind the Training Module - it’s designed to be realistic. Because of that, once we start introducing unrealistic inputs such as changing things in the editor or creating synthetic training schedules, we run the risk of receiving unrealistic outputs. So if some results are surprising, whilst it may show an area of concern within the module it’s probably more to do with the complexities of the module and the realistic basis. Because of that, trying to draw conclusions is perhaps coming at it from the wrong angle - instead output the results and then, as said above, feed them into SI to analyse and let them draw conclusions 👍. Thanks for the reply. The decision to sort of "reset" the players using the editor was done to ease the testing, isolate variables, and being able to change one variable and see the effects on others. In addition, I don't think the edits are so far fetched from a realistic team of 17-18 years old players advancing to the B team. As far synthetic training schedules I'm not sure what you mean, both training schedules seem sensible to me. Test #3: Aggression in Normal Environment I used the save without any edits. I left training to my assistant manager and went on holiday for entire season. The default training schedule for the B team is Pre-Season: 'Tactical - Early, Mid, Late' then cycling through the 'Training Style' category with no particular order. 'Training Style – Balanced', 'Training Style – Defence' and 'Training Style – Physical' all have the 'Defending' and 'Defending Engaged' training sessions which list Aggression under their impacted attributes. Results: Name Age Position Personality CA PA Prof Inj Pr Acc Agi Bal Jum Nat Pac Sta Str Agg Ant Bra Cmp Cnt Dec Det Fla Ldr OtB Pos Tea Vis Wor Cor Cro Dri Fin Fir Fre Hea Lon L Th Mar Pas Pen Tck Tec Sébastien Cibois 24 GK Fairly Professional Isaïa Hubert 19 GK Balanced 1 Kylian Le Her 20 D/WB (R) Balanced -1 1 -1 -2 -1 Brieg Le Bars 20 D (C), DM Fairly Determined 1 -1 1 -1 -1 -1 -1 -1 -1 -1 -1 -1 -1 -1 -1 -1 Noé Sow 23 D (C), DM, M (C) Balanced 1 1 Josué Escartin 19 D (C) Balanced 1 1 1 1 1 1 Ali Dicko 19 D (C) Balanced 1 1 1 1 1 1 -1 1 1 -1 -1 -1 -1 -1 -1 -1 -1 -1 -1 -1 Kévin Galisson 19 D (LC) Fairly Determined -1 1 1 1 1 Tom Flamanc 20 D (L) Balanced 2 1 1 1 -1 Téva Lossec 19 D/WB (L) Determined 1 1 1 1 1 1 1 1 Kilian Liri 19 M/AM (C) Balanced -1 1 1 1 1 1 1 1 1 1 1 Pierre Jouan 19 DM, M (C) Balanced -1 1 -1 1 -1 1 -1 -1 -1 -1 -1 -1 -1 -1 -1 Kurtis Chadet 18 M/AM (C) Balanced 1 -1 1 1 -1 1 1 -1 -1 -1 1 -1 -1 -1 Abdourahmane Thior 20 M (C) Fairly Determined 4 1 1 1 1 -1 -1 -1 -1 -1 -1 -1 -1 Djibril Konté 19 AM (RL) Balanced 3 1 -1 1 -1 -1 -1 -1 -1 -1 -1 -1 -1 -1 -1 -1 -1 -1 Axel Camblan 19 AM (R), ST (C) Balanced 2 -1 -1 -1 -1 -1 -1 -1 -1 -1 Rémi Maugain 20 AM (C), ST (C) Fairly Determined 3 1 2 1 -1 1 Aymeric Faurand-Tournaire 18 AM (C) Balanced 1 Rafiki Saïd 22 AM (RL), ST (C) Fairly Loyal Loïck Piquionne 21 ST (C) Balanced 1 Samuel Genty 20 ST (C) Balanced 2 -1 1 1 1 -1 -1 -1 Lucas Gélébart 19 ST (C) Balanced 1 0 -1 -1 4 6 3 4 0 5 7 0 1 -2 -1 1 -4 4 3 1 1 -2 -3 -2 -3 -7 -3 0 -4 -3 -5 -2 -4 -2 -3 -2 -2 -4 -3 Discussion, conclusion and follow-up: This was done to further verify that Aggression cannot be trained despite it being listed under General->Defending and Defending->Defending Engaged under normal settings. Aggression hasn't changed in a single player. Edited August 10, 2022 by Royaltea Made table more compact for readability Link to post Share on other sites More sharing options...
herne79 Posted August 11, 2022 Share Posted August 11, 2022 7 hours ago, Royaltea said: This was done to further verify that Aggression cannot be trained despite it being listed under General->Defending and Defending->Defending Engaged under normal settings. Aggression hasn't changed in a single player. Why does it mean Aggression cannot be trained? Perhaps it means that Aggression is just very slow to develop in players, sometimes to the extent where it doesn’t shift at all - and isn’t that realistic? Do real life footballers become more Aggressive as they age? I don’t have data to know one or the other but I know SI do, therefore isn’t their training module realistic? Just because Aggression is listed under a training session doesn’t mean it will automatically develop - it means it’ll behave in line with the realistic training module that SI have developed. As said above, drawing conclusions is perhaps coming at this from the wrong direction - if you think something is off with the training module (SI have never said it’s perfect) then feed the data into SI, let them study it and draw conclusions 👍. Link to post Share on other sites More sharing options...
poobington Posted August 11, 2022 Share Posted August 11, 2022 @Royaltea Have you checked out Evidence Based FM's videos on training. He does similar experiments and goes into a great deal of detail about the efficacy of the different training schedules in FM 22. This video in particular summarises all of the training schedules in FM and shows what attributes they affect. Link to post Share on other sites More sharing options...
Dotsworthy Posted August 11, 2022 Share Posted August 11, 2022 (edited) This is anecdotal, but I've never seen aggression change on any of my players in any FM game for the last few years. While SI do mention there is some stuff going on under the hood which we don't know about, I do believe that aggression is bugged. Would be interested to know how this differs from just following the assistant managers advice, whether there is significant different in attributes. Edited August 11, 2022 by Dotsworthy Link to post Share on other sites More sharing options...
Royaltea Posted August 11, 2022 Author Share Posted August 11, 2022 @poobington I wasn't aware of his tests. His videos are pure gold mine. Very informative and methodical. They also revealed potential flaws in my test settings because the B team do play matches (even though they do not appear on schedule), and there's no way to know whether I'm recording training effects or match effects. There's also the issue of repeating tests multiple times and using averages instead of absolute values which I'll do in the future. Thanks for the link! @herne79 You're absolutely right. Maybe it take years for aggression to increase and this will have to be further investigated. The right wording would be that under a period of one season of numerous training schedules (including the default one) the Aggression attribute hasn't changed once. --------------------- At the end of the day, I'd like some actionable training strategy to follow, and being able compare different training strategies. Therefore, I decided to introduce a metric and give different training strategies a "score" (overall growth in attributes). This metric should be taken with a grain of salt because training affects lots of things: fatigue, sharpness, familiarity with the tactical system, injuries, morale, team's cohesion etc. Test #0: Establishing a Base Line: Full Week of Rest Results: Name Position CA PA Prof Inj Pr Acc Agi Bal Jum Nat Pac Sta Str Agg Ant Bra Cmp Cnt Dec Det Fla Ldr OtB Pos Tea Vis Wor Cor Cro Dri Fin Fir Fre Hea Lon L Th Mar Pas Pen Tck Tec Att Ch. Sébastien Cibois GK 72 142 1 1 -1 1 2 Isaïa Hubert GK 71 142 1 1 1 1 1 1 1 -1 -1 1 1 7 Loïck Massé GK 70 140 1 1 1 1 1 1 -1 1 1 1 8 Kylian Le Her D (R) 69 136 1 -1 -2 1 -1 -1 -2 -1 -1 1 -2 -8 Mathéo Da Silva D (R) 67 136 1 -1 1 -1 -2 -2 -1 -2 -1 -1 -2 -11 Brieg Le Bars D (C) 70 136 1 1 1 -1 -1 -1 -1 -1 -1 -1 1 -1 -1 -1 -6 Noé Sow D (C) 69 136 1 1 -2 -1 -1 -1 -1 -2 -1 -1 -1 -2 -11 Josué Escartin D (C) 70 138 -2 -1 -2 -2 -1 -2 -1 -1 -1 1 -1 -13 Ali Dicko D (C) 69 136 1 1 -2 -2 -1 -2 -2 -1 -1 -1 -1 -1 -12 Kévin Galisson D (L) 70 136 -1 -2 1 -1 -2 -2 -1 -1 -1 -10 Tom Flamanc D (L) 69 136 1 -1 -2 -1 -1 -2 -1 -1 1 -1 -8 Téva Lossec D (L) 70 136 2 1 -1 -1 -1 -1 -1 -1 -1 -1 -5 Kilian Liri DM 67 132 1 1 1 2 1 1 1 -1 -1 -1 -1 -1 -1 1 1 1 -1 -1 1 -1 1 1 -1 -1 1 4 Pierre Jouan DM 67 132 1 1 1 1 1 1 -1 -1 -1 -1 -1 -1 1 -1 1 1 -1 -1 -1 -1 1 -1 1 1 -1 1 -1 -1 -2 Rémi Maugain M (C) 65 128 1 1 1 1 1 1 1 -1 -1 -1 -1 -1 1 1 1 1 -1 -1 1 -1 1 1 -1 1 -1 5 Kurtis Chadet M (C) 64 128 -1 -1 -1 -2 -1 -1 -2 -1 -10 Abdourahmane Thior M (C) 66 128 1 1 1 1 -1 -1 -1 -1 -1 -1 -1 -1 -4 Djibril Konté AM (R) 68 130 1 1 1 -1 -1 -1 -1 -1 -1 1 -1 -2 -1 1 -5 Axel Camblan AM (R) 65 130 1 1 1 1 1 -1 -1 -2 -1 -2 -1 -1 -2 -2 -2 -2 1 -11 Loïck Piquionne AM (C) 62 122 1 1 1 1 1 1 1 -1 -1 -1 -1 1 -1 1 1 -1 -1 -1 1 1 -1 -1 1 -1 -1 1 Aymeric Faurand-Tournaire AM (C) 63 122 -1 -1 -2 1 -1 -1 -1 -1 -7 Rafiki Saïd AM (L) 67 130 1 1 1 1 1 1 -1
Recommended Posts