BulldozerJokic said: @GeorgeFloydOverdosed Great write-up! One question tho, do you think GK impact on the overall result is worth four training sessions per week? Or do those sessions also greatly boost important outfield player attributes? Expand This was one of the surprising realizations I had reading EBFM's excel file, that the GK sessions are actually also very good for the outfield.
match practice = 0.05 drib DC, 1.15 con DC, 1.05 con DL/DR, 0.4 acc DC, 0.4 acc DL/DR, 0.65 pac DC, 0.6 pac DL/DR (2.05 physical) (4.3 key attributes) (12 CA). +0.2 agil/+0.6 aer/+0.3 ref for GK.
Handling does a lot for all positions too, not just defenders.
When I changed GK attributes in my 1 CA testing, it did make a difference, so GK has some significance. I'd say a fair bit more than I intuitively thought. Not sure exactly how much, but I'd guesstimate their worth is at least that of 1 of the 11 players.
What I didn't really mention in my post btw is that my thinking is that these sessions like handling that boost physicals moderately (~0.3) are perhaps particularly worthwhile because with a bunch of them maybe you could replace a whole quickness/resistance/physical session which gives you +1.2 but lowers a whole bunch of technicals and mentals.
Yarema said: These tests really need a bigger number of data points. Doing one 3 season test can maybe point in the right direction but far from conclusive, you can easily get deceived by random chance. Even EBFM's training videos I'd say are underpowered and probably harvestgreen's too (to my knowledge).
Don't get me wrong they are a great start and eye opening in some cases but people make way too strong conclusions from them. None of those tests can differentiate between 0,35 and 0,31 with enough accuracy. Expand Yes, I would agree with you there.
Though, as I think about it now.. given it seems to show patterns accurately enough (i.e. aerial defense gives 1.25 con DC, 1.25 con DL/DR, 1.2 con DM, 0.8 con MC, 0.5 con AMC, 0.65 con AML/AMR, 0.5 con ST) maybe it is actually non-random. Maybe the 'random' variation between different players is simply their professionalism, matches played, CA-PA diff, etc. downscaling a set max amount, and then you have different boosts for different positions giving the appearance of each player progressing seemingly randomly compared to other players.
I suppose it's something I should try and test.
Yarema said: Also you can't add up different benefits with various sessions as it was already tested by EBFM for example. Expand I haven't got that far yet
I see now he has an excel file with combo sessions, with positional breakdown. I'll examine it to try and deduce what's going on
I've been having a bit of a thinkle about training
In trying to create the best performing 1 CA players, I found that all that seems solid is pace, acc, jump, dribbling, concentration and perhaps flair. And only pace & acc is needed in all positions. For GK, it's clear that agility + aerial reach + reflexes are the key three.
I also know that decisions, technique, tackling, marking, and some others are often highly weighted but make no significant difference to results.
I was looking at EBFM's more precise breakdown for training session effects by position, albeit it is for FM23. But I've created a schedule based on its info, and it seems to still work the same.
The schedule I've created is:
1 x Quickness 1 x Goalkeeping 1 x Resistance 1 x Shot Stopping 1 x Match Practice 1 x Handling 1 x Match Tactics 1 x Aerial Defence 1 x Chance Conversion 1 x Play From The Back 1 x Distribution Quickness focus (Agility on GK)
This is a significant departure from HarvestGreen's recommendations.
I've analyzed each session, and have chosen ones that favor pace & acc, whilst retaining/boosting as much as possible drib in relevant positions + concentration in DC+DL/DR + minimize decisions & technique gain (to free up CA) + agil/aerial/reflexes for GK + some other lesser considerations.
In my initial 3-year test of it, I'm seeing the best improving players getting +5 to +6 pace/acc, +4 agil/aerial on GK, and lesser but positive movement in the other attributes I mentioned. Requires refinement, but seems to be working close to as intended. I'm testing using just the default Bournemouth starting players, put the knap tactic on, and just let things run for 3 years. I did a test of Quickness + Match Practice + 2 x Attack + Quickness focus to compare. My schedule was competitive with it, but fell a bit short. GKs improved significantly better though I think.
Here is my critique of HarvestGreen's training:
Combinations are assessed according to acc+pace+jump and overall team CA boost/cost.
If you go with minimal CA cost for max acc+pace+jump, you have to do a lot of rest which usually results in match sharpness becoming unsalvagable which significantly affects win rate. A lot of the other attributes decline, yet some of these attributes must matter to some extent as acc+pace+jump alone fails to win.
Additionally, match practice possibly impairs team performance and may be unnecessary. And there are still a lot of rest periods.
If you go for high acc+pace+jump and high CA boost, a lot of that CA could be junk attributes and it doesn't tell you how much of other important attributes such as dribbling & concentration and whether they are put in the right positions or not. For instance, it seems dribbling is good on DL/DR, but unnecessary on ST.
I'll compare 2 training sessions (using EBFM's FM23 excel file) to illustrate the issues:
Attacking - +0.74 drib, boosts to many stats with no declines, +11.46 CA. Aerial defence - +0.47 drib, boosts to many stats with no declines, +10.3 CA.
Attacking seems like the logical choice here. However if you look at the differences:
Attacking - +0.8 drib AMC (position not used in knap tactic), +0.2 drib ST (unimportant), +0.05 drib DL/DR (important). +0.38 decisions, +0.26 technique (costly near useless attributes that should be minimized).
Aerial defence - +0.35 con DC (important), +0.3 con DL/DR (important). +0.32 ant (semi-important?), +0.31 cmp (semi-important?). +0.13 acc, +0.18 pace. For 1.16 less CA.
Now Aerial defence is the no-brainer choice. Perhaps moreso if you consider that CA growth is limited to only a dozen or so CA a year on average.
This is where I get confused though. Maybe someone can help me understand. I understand that HarvestGreen is showing us that as rest boosts physicals by default, training is essentially an allocation process. But if I do quickness (+4.61 CA) + match practice (+12.01 CA) + attacking (+11.46 CA) does that all get compressed proportionally into say 12 CA?
Zippo said: One of its purposes is to limit how low the value of the "Condition" attribute can drop.
For example, if a player has "20" for Natural Fitness attribute then his Condition attribute can NOT drop lower than "7,800" value. Expand This simply isn't true. I just played a game with my 1 CA 20 NAT players to confirm, my ST went down to 61% in-match, 67% post-match. Perhaps it is true if match sharpness is also 100%. And if memory serves me correctly, natural fitness does not even affect condition fall rate, it is stamina that does.
Zippo said: No, a task to achieve 10,000 value for the Condition is almost impossible because the rate at which the Condition attribute is restored after 9,000 level is very low.
So you should look for 9,000 - 9,300 range. Expand As your stats show, the performance loss would be significant. It's not impossible to maintain ~100% condition, it's quite easy. Another reason is that you will also otherwise suffer exponentially greater injuries, as EBFM found injury rates of:
And those injuries create a vicious cycle of low match fitness & condition leading to more injuries. I would guesstimate injuries double overall if starting at 90-93% condition.
Zippo said: How much does the Condition attribute decrease after a match? It depends on many factors. The rate differs for positions and roles. Also, the Intensity of a tactic greatly affects the rate at which the Conditions are consumed. Expand Is that really so? According to EBFM's data, the difference between the most intense and least intense settings were 1-2% condition difference by end of match.
Zippo said: Also, the Team Cohesion and Match Sharpness drops a bit when players are "Resting" but it isn’t a big issue. Expand Match sharpness is very important, more important than condition. It's performance impact is greater, and it's harder to maintain. So using rest is usually a bad idea, from my analysis of it.
These results might not be directly comparable, but you say that starting at 88% condition results in 16% worse performance. EBFM found that match sharpness 100% > 90% reduces win rate by 33.3% and injuries increased by ~25%.
I know I come across as a bit of a knob here with all of this, but it's partly that I can't resist knocking down falsehoods, and partly that I wrote a post on here about this very topic recently so I'm a bit of a zealot about it. I know I state a lot of things that turn out to be erroneous myself. This is just part of the process of discovery.
Jolt said: A couple of quick questions to harvestgreen22 or someone that knows.
The training sessions generally show lowering "Condition"
But from my observations, condition always improves after each day with just training, and only lowerswhen playing matches (or getting injured, but that's beside the point). So my question, has it ever been tested whether these trainings actually: 1. Lower condition? OR 2. Training sessions with heavier condition reduction don't allow the players to recover condition as much as training sessions with lighter condition reduction? Expand
I did a quick test:
match practice doesn't decrease condition 1 x match practice recovers condition same rate as all rest 1 x match practice doesn't boost match sharpness (identical to all rest)
2 x match practice + quickness on one day, is identical to full rest in regards to condition recovery & match sharpness
What you might not know is that condition recovery doesn't have a randomness factor, unlike many other FM mechanics, so there shouldn't be any need for a ton of tests.
Seems EBFM found this back in 2022, and additionally found that match practice actually negatively impacts match performances, even though it has no condition impact.
This is strange. Maybe the 'injury risk' and 'fatigue' labels are also falsehoods and the whole training meta needs a rethink if one is not looking solely for attribute gain.
BaZuKa said: To see if I understood the post correctly, this striker is very strong because he has acceleration, pace, flair, and work rate. Expand
It would be at least most of the reason why.
You have to be a little careful extrapolating the 1 CA results to higher CAs. Even at 1 CA, it's clear other attributes do matter, just not much. For instance, I can say for certain that if you give ST higher finishing, he will do most of the goals instead of the wingers (that doesn't mean a necessarily better overall team performance though). But also it's possible that consideration of certain attributes only kicks in at say 120+ PA. This was the case in Championship Manager based on what I've read, and whilst that mechanic is probably gone now, I think there's a fair chance something similar still exists.
Another thing btw is that apparently the match rating is skewed towards technical ability, and biased against physical ability, even though physical ability beats the technical ability. Source. This is particularly interesting to me, because it helps to explain why Orion's coefficients doesn't produce a winning 1 CA team, even though it gets most of the way there - Orion's coefficients are basically comparing the match ratings, and if its biased towards lesser-performing technical ability, this would explain it.
Also I'm not 100% confident on flair yet, I should test it more, but I did a little testing, and so far it seemed it was beneficial in spite of what HarvestGreen found.
I'd heard it in a few places that someone had won the Premier League with a team of 1 CA/PA players. As it turned out, it was done back in FM21 with a strikerless tactic, exploiting the low weighting cost of DMs. I had a go at recreating it, mainly as a bit of fun, but also to confirm it for myself as I've heard people deny it's possible. And it needs an update for FM24.
Mid-table results were easy with just 20 pace/acc/drib (not sure if I included jump), but it required careful tweaks — closer to Orion’s data than HarvestGreen’s — to push into the top four. Winning is barely possible and inconsistent, but the experiment taught me a lot about how attributes, combinations, fitness, morale, and form affect outcomes. I wanted to wait until I'd worked out a way to win consistently, but I'm a bit burnt out by it now - at least the main point is proven, that you can actually win with just 1 CA/PA players, and I may try and improve it more later on. Coming 2nd I can do consistently. I may also retest without the 210cm height & player traits, I actually doubt those impact the results much.
The winning result:
Scroll to bottom for attribute stats by position
Setup:
• Use Knap’s EF 424 IF HP V2 P101 AC tactic and Blue Routines set pieces. • No strikerless exploit. • Allowed: friendlies, training, captains; no reloads. • Squad: 22 players, all 1 CA.
Findings:
- It is possible to win with 1 CA players without strikerless exploit - Concentration ~16+ on CBs+FBs (no other combo) was key to winning the league - Pace+Acc alone = 0 points, 0-7 loss typical - Jump+Acc alone = a draw or two, 1-7 loss typical - Drib+Pace+Acc alone = mid-table - HarvestGreen data skew = bottom to mid-table - Orion data skew = mid-table to top 4 - Orion data skew + pace/acc 16 or less = bottom - Orion data skew + pace/acc 20 = mid-table - Orion data skew + pace/acc ~17-20 (mix) = top 4 - 20 in key attributes + maxxing out other lowest weighted ones = high mid-table - Lowest weighted attributes (i.e. finishing '1' weight on CBs) had negligible effect (i.e. no extra goals or overall team dominance) - Most attributes make no statistically significant difference, very low values will even dominate many in-game stats - (Only) wingers 1 pace/acc = high mid-table - (Only) one slow ST target man = high mid-table - League position more consistent than points (i.e. one case, almost always finish 2nd but was 66-91 points) - Winning formula finishes ~4th majority of the time, 2nd place was more consistently 2nd - Fitness management, particularly of match sharpness, is important - difference between ~mid-table and ~4th say - Morale & form somewhat important, difference between ~4th and ~2nd say - Squad rotation (even if perfectly done) or too much rest detrimental to results (maybe this is why AI managers rarely rotate, and why the game says 'I want to play you but X is in the form of his life right now' when players pipe up for more playing time) - Altering knap tactic (i.e. 'very attacking'/'defensive', 'work into box', etc.) was worse than default - 1 consistency doesn't matter, since consistency reduces CA but 1 is already the minimum; seemed to be the case in testing, but hard to verify 100% - Biggest premier league win I got was 9-0
What stayed the same/correlated with top results:
GK - agility 15-20, acc 5-6, pace 5-8, sta 3-6??, composure 6-10?, work rate 6-7, aerial 16-20, communication 3-8?, reflexes 10-13 DL/DR - acc 20, pace 20, jump 20, off the ball 20? (unlikely), vision 5-7?, work rate 7-11, dribbling 20, finishing 20 DC - acc 15-20, pace 18-20, jump 17-20, strength 4-6, flair 20?, vision 15-20?, dribbling 20 DM - acc 19-20, pace 20, jump 20, bravery 12-20?, flair 20?, vision 3-4?, work rate 7-10? AML/AMR - acc 15-18, pace 16-18, jump 20, flair 20, positioning 18-20, work rate 7 ST - pace 20, acc 18-20, flair 20, work rate 7
Question mark indicates I think it's unlikely to be essential.
I tested certain attributes (at '20', with CA boost added so no decrease of other attributes) specifically to see if they mattered (i.e. changed league position), most didn't. Some examples:
ST - finishing, agility, dribbling, work rate, decisions, stamina, strength, off the ball - not any of these DM - strength, vision, dribbling - none of these DC - agility, balance, strength, tackling, marking, heading - these do win/make close 2nd, but only slightly
I haven't tested all attributes and their combos exhaustively, so there may be some things left, but I tested a lot of them.
You can deduce a lot of the above yourself from the following images (I've kept the roles highlighted to show how misleading they are):
GK:
DL/DR:
DC:
DM:
AML/AMR:
ST:
Note: If you're wondering *why* I've changed certain attributes and in the way I do, it's that I'm following the money. The reason finishing went from '5' to '3' for strikers isn't because I never tested above 5, but because I probably did a dozen different combos with finishing above '5' that didn't work out. I think there are knock-on effects to other positions as well (i.e. if your winger has better finishing, he'll probably do most of the goals), so I'm kind of forced to just take the approach of 'if it works, it works' and I move on to try and improve another position until I can't squeeze anymore out of it and then move on to the next one.
It is certainly complex rather than simple. It is mainly the training & weighting system that is broken.
The two most important player attributes are speed & work rate. This reflects reality accurately. The problem is when you can use HarvestGreen22's training regimes to turn literally any player into an elite player (pace/acc 8 > 20), and when you can give a full back 20 pace/acc/jump/nat/long/fin/drib/det/agg/off the ball/work rate and he'll be 1 CA with a handful of points left to spare.
I think you are correct in your hypothesis about how attributes are processed by the ME. Of course until we can see the game's code, we have to make our best judgement based on results and what SI staff have said over the years.
So this is what I understand to be the case:
The ME uses typically 2-3 attributes (sometimes up to 5) for each game event (i.e. passing + work rate = pass attempt)
An ATTEMPT ('choice' ) is made FIRST A COMPLETION ('execution' ) is made SECOND
i.e. passing + work rate = pass attempt passing + decisions = pass completed
As you hypothesize, for many/most players, the only way to win is not to play at all. Instead of going for a slim chance of a slim chance of a pass leading to a slim chance of a goal, if you go with passing '1', then decisions can also be '1' since it almost never comes into play now, and you can now instead rely on dribbling '20' with pace/acc '20' to run up the park and score in ~17% of instances with [dribbling (20) + balance (7) = 67%(?) evasion chance] x [finishing (7) + composure (7) + decisions (1) = 15%(?) scoring chance].
Something I'm observing is that '20' pace/acc will win games where even '17' pace/acc would see you get demolished. I suspect this is because you simply run past/around the opposition's players, so the intercept & block chances are left out of the calculations. This is showing up in my 1 CA player testing, but I think it would still hold partially true for normal gameplay where you have weaker technicals/mentals/other physicals than your opponent.
Another thing I have observed is that while changing full backs to have drib/long/finish 20 (all players still only 1 CA) resulted in a lot more premier league wins, the full backs themselves rarely scored goals, or even assists. And looking at the match highlights, it doesn't appear to me to be that say fullbacks are making big forward runs up the side leading to later goals. This is still a mystery to me right now.
I do suspect that tactics give favor to certain attributes, and that meta tactics favor/exploit high pace/acc/drib, but I also think that only meta tactics tend to work, especially since your AI opponents (at least, the better ones) are also going to be utilizing these same meta tactics against you. I have noticed however that you can use a low pace/acc target man with no ill effect.
I think there may be something in thinking about the fact that attributes aren't denoting actual ability (i.e. you can even see in the 3d engine, that a dribbling '1' player can still dribble), but rather the only factors in decision outcomes. So a dribbling '1' player from Bedford Town dribbles just as good as Messi would in the game; it's only when Messi tries to tackle him that there is a difference. I think this partly explains why pace is so important, but technicals could be '1' and still do ok. Pace is active all the time, a slow player is always obvious and hampered in the 3D engine. But often a '1' dribbler looks just like a '20' dribbler (particularly when he has high pace to lengthen it out). Therefore the whole 'a player is only as good as his worst minimum' logic shouldn't be applied.
A similar potential misconception is that an attribute is simultaneously tendency and success rate. It seems true, but only to an extent: passing '20' gives you both more attempts & success, however a player with '1' passing still makes passes anyway, so probably '1' passing & '20' pace does more actual passes than '20' passing and '1' pace. So really '20' pass you could look at as say +5% tendency, +20% success, rather than +95% tendency +95% success.
Here's two other quirks I feel are worth mentioning:
If you go by what the staff-written guides say, strength alone can be used in lieu of tackling + marking + aggression. Strength is pretty highly weighted, but then again, it isn't subject to the consistency CA decrease like tackling/marking is, and it also says that strength reduces fouls as they can muscle the player off the ball instead of sliding in. For ST, tackling/marking is 1 weight, strength is 6, and a foul probably won't result in a penalty against you, so low strength probably makes sense here. But for DC, marking is 8 weight, tackling 5, and strength 6, and sliding in probably results in more conceded penalties. I'm yet to test this though.
If you think about it, you can set consistency to 1 for 1 CA players and it will make no difference, as consistency reduces CA for technicals to use. My early testing of this finds it does indeed make no difference. I suppose you could utilize this in a very low league team.
How much do the latest findings indicate the importance of pace/acc/dribbling and overall physical attributes?
Let me explain what I mean:
Player X: 18 Dribbling, 18 Pace, 3 Flair, 2 Technique Player Y: 15 Dribbling, 15 Pace, 6 Flair, 12 Technique
In this scenario, Player X would clearly perform better than Player Y, right? Expand
My impression is:
Orion's attribute coefficients are most accurate to use for each position, more effective than HarvestGreen22's attribute ratios
HOWEVER Pace/acc/jump 20 will override this (possibly some other attributes in certain positions too)
Pace/acc 15 will not be sufficient for the override. But this doesn't mean they still aren't highly valuable attributes, I'm just talking about having it override other attributes to the point where you could have a DC with 1 tackling, marking & positioning and it wouldn't matter, the 20 pace/acc player would still be better than the more balanced pace/acc 15 one
Details from old SI staff-written manuals, such as [forward runs - stamina, anticipation, off the ball] seem generally correct
HOWEVER such rules/combos can and should be broken to shift valuable high weighting attributes to valuable low weighting attributes. I.e. [DC - Positioning 2, Dribbling 1] should be changed to [DC - Positioning 1, Dribbling 20] (same CA cost) even though both Orion's attributes and the guides tell you to get positioning
My theory is that putting high weights on key attributes for the relevant position carries an implicit necessary corollary: those attributes are so heavily weighted simply to dissuade you from making 20 marking/tackling/strength DC beasts out of thin air, not because marking has 8x performance and your DC will actually only be hampered by ~1x performance if you decrease marking to 1. You take that 7x leftover and make 20 dribbling/finishing/long shot beasts instead and utilize them with a knap forward-throwing tactic
I've come to these conclusions by trying to get a 1 CA team to win the premier league. Here's a brief summary of how it's progressed:
Base my 1 CA players off my combined HarvestGreen22/Orion/etc templates - Near bottom of table. Base it more on Orion's attribute coefficients - Mid table finish. Try Orion's + decrease pace/acc to ~15 to put some 7s and 8s on some key mentals/technicals - Back to bottom of table. Do the opposite, try to get pace/acc/jump 20 for most players while preserving what I can of what Orion & manual says - High mid table finish (6th, 66 points). Reduce positioning to 1, get dribbling 20 instead, etc. - 4th, 74 points.
In your example, player X performs better. But there are some nuances. Flair and technique are generally unimportant, even detrimental in certain positions. However let me show something that astonished me:
Player X (ST): 20 pace, 19 acc, 1 jump, 1 strength, 6 finish Player Y (ST): 14 pace, 9 acc, 19 jump, 8 strength, 3 finish
You would assume that Player X is far superior. However I used these 2 players as STs in my 1 CA testing, and they got me to that 6th-66 points result. But what's more, Player Y scored 7.08 rating with 29 goals in 55 appearances, while Player X scored 6.42 rating with 15 goals in 57 appearances.
Now you could theorize and say, well that's because set piece headers are the only way you're going to score goals with 1 CA, or that perhaps the wingers are doing the work throwing it to the target man, or something else. But my point would be is that your attribute choice, even pace/acc and down to such low levels, is genuinely subject to your tactical setup in a certain minority of cases. Note that if either player were actually *bad*, the team wouldn't have ended up in 6th spot.
That does not mean we are back to where we started. I suspect ST target man is one of 2 or 3 roles that can work with low pace/acc. On the other end of the spectrum, for DC, 20 pace/acc is simply essential.
I suppose a general rule of thumb you could use is physicals over mentals, and mentals over technicals. If you choose strength 16 over 16 dribbling or 16 composure, it's almost a sideways trade and this is one of the worst you could do with this rule, whereas trading pace/acc for anything is going to be iffy or catastrophic.
I've been having a rethink on my ideal player templates, trying to improve them. In testing certain changes to attributes one-by-one I've had some interesting results.
I'll explain my process a bit first. Let's take a more nuanced attribute, 'dribbling'.
First I look at other people's findings:
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.
And then there's the FM Arena testing of course.
Orion uses a different method, he derives his attribute values from the match ratings from the actual leagues as they exist in the game, and he does it for each position. For the DR position, dribbling is ranked 6th. For DC, it is not in the top 8.
But for DR and DC, I plan to give them perhaps ~14 and ~16 dribbling respectively in my updated ideal templates. This is for several reasons:
1. With the Knap tactic I'm using, DC has 'dribble more' instruction. Normally this would be an automatic ~2 boost for me. 2. Dribbling has a very low 1.0 weighting for DC. After doing the math on it's relative value once weighted, it's clear this is worth setting very high. 3. I've read a comment that perhaps the match engine favors dribbling in general, not necessarily because dribbling is good in itself, but because simply continuing to run with the ball prevents other calculations coming into play. For instance, if a player decides to pass instead, then a decision/technique/etc error may occur, and all these calculations cause time delays. 4. According to the SI staff hints from long ago, dribbling combines well with high pace & acceleration. It intriguingly adds 'don't have too many dribblers as they'll tend to ignore any passing tactic you might have'. 5. Someone (HarvestGreen22?) theorized that attributes are simultaneously ability and tendency. Given that I expect my DC and DR to still do a fair bit of passing, that means dribbling needs to very high to take precedence. 6. A dampening factor must be applied, which is the actual availability of the attribute value. In a save I've loaded up, there are 489 DCs who are 100+ PA w/ 12+ dribbling. There are only 18 who have 14+ dribbling. Technicals decline or stagnate with HarvestGreen22's optimal training. Let's say you could take 1 of the 120 players (or newgen) with 13 drib and boost it by 1-2 with training, that's 14-15 max dribbling you're looking at. For DR it's a bit better, a small handful already start with 16. If it were not for this reason, I would get 20 drib for both positions (this is what I'm trying to communicate through my 'high' ideal templates).
In this case it's quite clear that regardless of whether the speculations are true or not, it's well established that dribbling is high performance and/or low cost. You can safely assume it without even running tests.
But sometimes there are contradictions, or nuances that haven't been explored enough, and these I perform tests to clarify.
I seem to be finding that technique and flair aren't always bad as HarvestGreen22 found them to be. Technique seems best at ~6 for most positions, flair I have around ~12. I get at least about ~10% performance loss if I drop either to 1, and 20 technique comes at high CA cost but does no better. I'm not sure what is going on here exactly. Perhaps it is that technique increases the number and quality of so-called 'tricky passes' - too much would detract from dribbling, too little would result in exponentially more lost possession, '6' takes advantage of only the most assured & promising tricky passes to nab a few extra goals that make up for any lost possession - particularly in top leagues, where the easier chances are very few and far between. Pure speculation of course.
Some others I wonder about are strength, agility and stamina. I figure that knap tactics can be weaponized against you in just the same way by certain AI managers, and that attributes have direct counters. So the counter to your AML with 20 speed (acc then pace) making a 10m charge towards goal cutting in from the left flank may be 15 agility (rapid direction change to sideways left) + 10 acc/pace to reach your AML in 3m the time it takes him to do the 10m. So if your DC has 20 agil/acc/pace, good luck to your opponent.
Yet we also see that HarvestGreen22's results show high agil/strength does relatively little. But then again, in his test his opponent has DCs with 10 agil/strength. In a tussle or direction change, they'll neither totally dominate you nor lose every contest. And what is clear is that 10 - that is, equaling them - does the job. Now in the premier league the best DCs you'll be coming up against have agil/pace/acc/strength around ~15. My theory is that it's probably best to at least equal that for optimal play, as your chance of winning a tussle probably exponentially decreases the wider the gap between the DC's strength and your striker's strength.. so agil/strength 15 on a striker. And if you put strength down to 6, because strength doesn't seem to overall matter much, are you sure you know what you're getting yourself into, because HarvestGreen never put a 6 strength striker up against a 15 strength DC.
So that is why I am in two minds about these attributes in particular. What I've found so far is that 6 str/11 agil is significantly worse than 15 str/agil, but it's not that bad (~7% worse maybe). For comparison, I found difference in off the ball to be fairly negligible.
Note that in the SI staff notes, it says of agility 'agility is an attribute which is most necessary for players in top flight football...combined with high acceleration and dribbling a high agility rating can make for a great attacking footballer'. And if you look at top goalscorers, they seem to have high agil/str - Haaland has 17 agil/17 str.
I have also tried to replicate the success of the Chinese forum member who won the Premier League with 1 CA players a few years ago. I've managed to nab a few wins and draws, but overall no cigar. This tells me that there are indeed certain non-meta attributes that you need to have certain minimums in for success, which makes me feel more justified in being wary of 6 strength and 1 technique.
Yarema said: Pretty sure you can get a new position to 20/20 unless they changed something in FM26, quite easily actually. New position might trigger some attribute relocation though due to different CA weights. Expand
tam1236 said: You can - tested many times. Especially at youngsters and with close positions. Of course it depends on versatility hidden attribute.
In theory professionalism is better - somebody on yt tested that there is no chance to make a good player with low prof. and sometimes it happens when a determination is low. But irl players with really low profess. are very, very few and very, very far between in FM base and in newgen. Balotelli is a lonely knight. And low determination you can see quite often Expand
I'll take your word for it on the position proficiency. I haven't done any testing on it, I've just always had the assumption from way back that 18 is the limit and I don't retrain players to new positions in my own saves.
I'll probably do a quick test of it to make sure later and update my main post later to reflect the fact that 20 proficiency can be trained.
In regards to pro vs det, aren't both just random generated for newgens now? To me, newgens are the main source of high PA signings. If I search in genie scout for players >140 PA w/ det 10< I get 472 players, 598 for pro 10<. So I still think it's clear if you had to choose one to search for, it would be high professionalism.
BLS122 said: Should you move all of your players into the “attacking” group or leave some in the defending one? Expand
I didn't look into the training units at all myself, but according to EBFM training units do have an effect
In the schedule I favor, it doesn't seem any of the units are split between attacking and defending, not even the 'attacking' units, so I would assume it makes no difference what group the players are in.
Robbo84FM said: So you would recommend not using the x7 Recovery? Have you done anything with developing your youth as FM26 doesn't allow us to actually manage the youth training, i have put all my 4+ star potential youths in my first team and made them available for the u18/u21 they are developing well as it's the only way to customize their training. Expand
I'm doing FM24, not FM26
Right now, I wouldn't recommend x7 recovery. But it doesn't matter much, we're probably talking like ~5% or less difference. I recommend Quick + Match Practice + Attackx2 + Double Intensity + Quick focus (Agility for GK) + remainder rest. 'No pitch or gym work' for first 3 condition icons, only 4-5 pre-season friendlies.
ZaZ said: As a matter of fact, Rest sessions are only triggered if you have three sessions in the same day, so if you do only two rest sessions and something else, it is equivalent to doing two Recoveries (and that something else). Expand
Is this so? I didn't know this
I am skeptical. If this is true, then why do recovery sessions seem to impact training results as compared to rest? Is it only because of the 3 period rest days? How can recovery also make such a large condition/jadedness difference over time if its just those infrequent 3 period rest days?
Steelwood said: Another post talked about how important match sharpness is to results so I think it is probably better to use recovery where possible, though my thoughts on that are more skewed towards winning matches rather than training results Expand
I agree with ZaZ's comment that rest works for match sharpness so long as you manage your squad correctly through rotation and playing them occasionally in youth friendlies if necessary.
I actually think recovery has more downsides than just slightly worse training. In my own save I was using the exploit trick to use all recovery sessions instead of rest, and found it was a bit too harsh on condition and caused excess jadedness. Jadedness isn't that bad as EBFM found it's self-limiting and it's difficult to reach and easy to solve, but it does have some impact on performance. So I've been advocating for a mix of rest and recovery, but after doing some testing, you can actually get away with no recovery sessions at all. And for the slight training bonus, you would want to.
Semi-related - At the moment I'm re-analyzing attributes. For some reason a few months ago I decided quickness + attack was the best, without match practice. When I returned to it for my post here recently, I thought I made a mistake of simply forgetting match practice. But now I vaguely recall it could be that although it's a "worse" training regime, it favors 'decisions' less say, making it actually better overall. So for me, the best training regime is still in flux, I will re-assess after completing my attribute re-analysis.
tam1236 said: 1. As far as I can see good physical progress can be observed at u19 players. If You buy 20 yo player with pace/acc 8 You lose your money. What's more if You have money to buy 140 PA young players there's no need to risk with very slow (speed 8) 18-20 yo players (BTW if you don't cheat with GS You don't know CA/PA ) 2. I would say determination is in a way more important because I dont remember players with professionalism 6-7, and not talking about less (I always try to search minimum in my teams for my information) but determination 2-3 is quite often in lower leagues - and drastically slower youngster progress - there is a hope for such a player but if you are not desperate dont risk with him. 5. You can change natural position and reach a new one to level 20 . Especially if You have a good newgen playing as CB with aerial reach ~9 you should do it. Apart from everything else aerial coaching doesnt work good. 6. If (IF!) You have money - why not. 7. Avoid players with low mental - concentration, compassion, determination have a great impact and are really hard to improve. From my experience GK with low consistency is a fail. Injury proneness IS important if You dont have good subs - on the other hand a game is buggy and player with this attribute 5 can have (in 4 years time) more injuries then the one with 15 - but statistically it is important, though You dont see these numbers without GS.
And maybe I'm wrong , but I have an impression that in FM24 loaning above-average young players to lower leagues doesnt work very good because their progress is capped by the league level. Expand
Mostly agree except a few small things
Determination 3 > 15 is equivalent to professionalism 5 > 13. But also 20 Determination = 14 professionalism, and determination 6 does most of it. (source). I do say on my initial post to get 10+ determination for optimal, but because CA growth is capped each year you only really need say half the optimal, so having 13 professionalism alone is probably enough to do the job.
Also there was another reason professionalism is better, I can only vaguely remember now; I think it was that professionalism allows the player to develop even if they don't end up getting match experience.
I may be mistaken but I don't think you can train player position to 20, it can be 'natural' but it will still be 18. That said, I was doing some more testing of 18 vs 20 and it seemed to have negligible effect. I was testing it in the first place because I figured it's likely 15 > 20 isn't going to be linear. I'll have to run it a few more times to be sure there is no significant effect of 18 > 20.
Robbo84FM said: Hello i know im being completely lazy here but i just haven't got the time atm to read back through the post has anyone done any testing yet or found the best training schedule to use FM26? so far i have just been using the Quickness/Physical - MP x2 - Attacking - Recovery x7 from FM24 Expand
From what I've heard, the FM26 training file is identical to the one in FM24, and HarvestGreen22 found that attribute importance remains largely the same (long shots, finishing were boosted somewhat)
Separate to that, I wouldn't use recovery. It actually seems to affect training gains, and you can maintain match sharpness well enough with this instead (this also has the most gains in FM24 according to Piperita a few posts above): Quickness + Match Practice + Attackx2 + Double Intensity + Quickness focus (Agility for GK) + remainder rest
Comaring training regimes after 19 days with 4 friendlies
Man City default:
Before final match - condition 83-91%, match sharp 91-95% After final match - condition 69-80%, match sharp 96-100%
Pure rest:
Before final match - condition 84-91%, match sharp 66-80% After final match - condition 67-81%, match sharp 72-86%
Quick + Attack + Match Practice + Quick focus + double intensity:
Before final match - condition 84-92%, match sharp 85-92% After final match - condition 71-81%, 93-96%
So HarvestGreen22 training is slightly better for condition, but a little lacking in match sharpness. However this may be about the right balance, as this was still before end of pre-season. Pure rest training is clearly problematic.
Training & rotation is something that has to be micromanaged as you go along, but here I've kept the same starting starting 11 for about 2 months in (5 friendlies + 6 competitive) and on the HarvestGreen22 training regime:
Before final match - condition 96-100%, match sharp 96-99% (all but one are 99%) After final match - condition 69-83%, match sharp 100%
So it happens to be well suited to fitness management to begin with. But if you were to try to keep your subs & backups at 95%+ match fitness, that would no doubt change things. So I did a full rotation (using only 7 subs though) every match:
Before final match - condition 95-100%, match sharp 79-99% After final match - condition 73-99%, match sharp 83-100%
Those figures include all the subs. 11 players had 100% match sharpness after the final match. So overall, it's not bad, but definitely needs more match sharpness for subs.
Now going by Piperita's clarification on the best HarvestGreen22 training regimes, and taking into account match sharpness, the top ranked one still overall looks viable: Quick + Match Practice + Attackx2 + Double Intensity + Quick focus. Since injuries are an enemy of match sharpness, I won't change the middle condition to double intensity, but just try and manage the players better. Additionally I've promoted more players to the first team squad, to make it 18 outfield + 2 GK, and added some more pre-season friendlies.
Results:
Before final match - condition 95-100%, match sharp 62-100% After final match - condition 69-100%, match sharp 78-100%
My observation here is that there were a significantly greater number of injuries, which led to low match fitness in a number of players throughout. Resting from training was also more frequent. You probably want just enough match sharpness so that low condition doesn't result in missed training days, which means either using a lighter training regime, or not scheduling too many friendlies.
I decided to have another go, cancelling the default friendlies and not adding too many. Because why not try and see if one can hold onto the optimal training regime.
Results:
Before final match - condition 96-100%, match sharp 89-100% After final match - condition 67-100%, match sharp 92-100%
So if you manage things right, it does work pretty well. I used just 4 pre-season friendlies here. I'm sure with some minor adjustments, you could get maintain match fitness of all players at 95%+ without having to field any low match fit players in competitive games (a handful of the backup/sub players could be played in u23 matches for the first month after pre-season to get them to 95%+).
tl;dr Use Quick + Match Practice + Attackx2 + Double Intensity + Quick focus (Agility for GK) training regime every week, even on congested schedules. 'No pitch or gym work' for first 3 condition icons. Don't do too many pre-season friendlies, it's counter-productive, 4 or 5 might be best. Full rest training lowers match sharpness far too much, but you don't need to change rest to recovery sessions or increase intensity at lower condition.
It says which training regime as caption when you expand the image. That one is 'Walters_mixed = My dense but optimized regime based on theories above. 80 injuries.'
I can't remember what training schedule I used, but it doesn't matter because it was inferior to the HarvestGreen22 training regardless of whatever that image may indicate to you.
So I have concluded so far that for condition, playing 1 in 2 matches is ideal. For match sharpness, typically 2 in 3 matches is ideal, with natural fitness allowing some variation to this. But notably some players will even need weekly matches just to keep up and avoid exponential match fitness decay. I had a look at the availability of 12+ natural fitness in genie scout amongst players, and its fairly common, so I would say try to get players of 12+ natural fitness if you can, as 10 or less is quite unforgiving.
Now, the implications for squad composition. If you think about it, this is really something that hasn't been examined at all, in spite of how consequential we all know it to be - that ill-fated tendency of us to bloat our squads; what even is the actual optimal number of players for a first team squad?
In my view, the number is 20 outfield + 2 goalkeepers + the occasional promotion from reserve/youth squad or loan to cover an injury.
9 matches need sub replacements (10 x 9) 17/53 matches permit sub replacements (6 x 17) 26/53 matches permit sub replacements for high Nat players (4 x 26)
4 developing players (AML/AMR/DL/DR) get 35 matches 6 players (2 ST, 2 DM, 2 DC) get 26 matches
Remember that 25-30 matches/season is the ideal for player development. The reason why the wonderkids should typically be wingers not centre backs or strikers I would reason is that you don't want to interrupt your ST's bucketload of goals and young players aren't usually developed enough to fill the shoes of a proper DC, but also that wingers seem to tire fast and only have 1 of each position instead of 2 on the pitch.
It is definitely adjustable to some extent according to personal preference. Basically, for each wonderkid you add, make one of your starters a high Nat player so they can handle the reduced playing time.
Some extra info:
My figures are based on using the optimal HarvestGreen22 training regime, using rest not recovery. So it is safe to combine this info with those training regimes.
Stamina does NOT effect condition recovery post-match. It only affects condition usage during the match itself.
Having a physio reduces match sharpness decay significantly, but it doesn't matter how good the physio is or how many extra you have - you just need one for the benefit.
50% match sharpness or 1 natural fitness recovers condition to 66% vs 68% for 100% sharpness. 20 natural fitness recovers condition to 74%. I mention these select figures to illustrate that natural fitness is kind of important to condition recovery, and match sharpness less so, but overall condition recovery is pretty stable and predictable regardless of the player.
There is a way to kind of exploit match sharpness gain:
friendlies:
~10 minutes for 100 (1%) match sharpness gain ~25 minutes for 200 (2%) match sharpness gain ~43 minutes for 300 (3%) match sharpness gain ~60 minutes for 400 (4%) match sharpness gain ~78 minutes for 500 (5% - max) match sharpness gain
So if you want to play absolutely optimally, you could sub a player at 78 minutes in friendlies to maximize match fitness. And throwing a player on for 10 minutes gives the most efficient match sharpness boost, while you'd know not to do 7 minutes as it gives them nothing.
For competitive as I mentioned before, it's 9% max instead of 5% max. However, EBFM did find that sometimes players would inexplicably get 10-11% max for a reason he could not identify, but I would just treat this as a kind of bonus.
9% match sharpness gain per 90min competitive match @ 50-65% sharpness 8% match sharpness gain per 90min competitive match @ 70% sharpness 6% match sharpness gain per 90min competitive match @ 80% sharpness 2% match sharpness gain per 90min competitive match @ 90% sharpness
I stated in a post that jadedness doesn't matter much, but match sharpness is crucial.
I feel like this topic deserves it's own thread and I've done some original research for it. And I think that in a way it's one of the last remaining actual challenges of the game to make it fun. If you don't use gegenpress tactic or pick/train high pace/acc then you lose a lot, simple as that. But even if you work out the mechanics of fitness management as I am trying to do, it remains a challenging and rewarding aspect of the game. In fact, just knowing how much it effects win rate to me brings back some significant enjoyment to the game.
My overall impression is that counter-intuitively to my mind, it's better to sacrifice condition for match sharpness. Maybe this is why the AI often/always plays players with 87% condition without any rotation. It's because low match sharp players suck in terms of performance, low condition not so much (injury risk is bigger downside here). The other thing about it is that you can deal pretty easily with the problems of low condition - rotate them, pick low injury proneness players, rest them. Match fitness is much trickier, there is only one modest offset for it (natural fitness), it also has a high injury downside, you have to pay the price of lower performance to get their match fitness up if you neglect it, and as I'll show below even friendlies have a limit to recovering match fitness so there's no get out of jail free card really.
I've taken this graph from EBFM's video on match sharpness. What it's telling you is that the difference between 90% and 100% match sharpness is 33% difference in win rate.
For comparison, the difference of pace/acc 10 > 18 is 23.5% win rate according to HarvestGreen22.
To boot, match sharpness significantly impacts injury rate, even at 90%:
So how do we optimize for 100% match fitness of all players every match? And also, how to handle condition in concert with match fitness? There's a lot of variables to consider, it's fascinating to contemplate, but I will try and lay things out in the most straightforward way:
Starting at 100% condition, 90 minutes will reduce to 75% condition (typically).
Condition & match fitness update at midnight (00:00), starting the midnight that may be just a few hours after the match.
It takes 10 days for full recovery to 100% condition.
In FM24, Man City (English Premier League) has the following schedule in a save I looked at:
67 games (excluding national games):
38 league 13 cup 16 continental
Total recovery days between matches:
53 x 7 days 4 x 3 days 5 x 2 days 1 x 16 days 2 x 13 days 1 x 11 days 1 x 9 days
So there are 9 matches where condition recovery will be insufficient, and another 53 which would be sub-optimal. With a weaker team with a less filled up schedule, such as Burnley, I found it could be as low as 6 + 35. In both cases, we would conclude a player should be ideally be rested every 2nd match.
But before we continue, let's look at the other half of the picture, match fitness:
after 3 weeks, match fitness dropped to 98% (recoverable in 1 game) (20 nat) after 4 weeks, match fitness dropped to 94% (20 nat)
after 6 days, match fitness dropped by 1% (10-12 nat) after 8 days, match fitness dropped by 2% (10 nat only) after 9 days, match fitness dropped by 3% (10 nat only) after 11 days, match fitness dropped by 4% (10 nat only) after 12 days, match fitness dropped by 5% (10 nat only) after 13 days, match fitness dropped by 6% (10 nat only) after 14 days, match fitness dropped by 7% (10 nat only) after 15 days, match fitness dropped by 8% (10 nat only) (99% > 91%) after 18 days, match fitness dropped by 11% (10 nat only) after 21 days, match fitness dropped by 15% (10 nat only) after 4 weeks, match fitness dropped by 22% (10 nat only) (99% > 77%)
Friendlies gain match fitness at 55% rate of competitive. As you can see, this is adequate for when the player has low match fitness, but once it hits around ~90% you run into the following complication: You can't have a friendly the day straight after the last one, so you are likely giving at least 2 days rest, yet only gaining 1% match fitness with the friendly.. so potentially you are losing 1% match fitness to later gain 1% match fitness. Hence, at 90%+ match fitness, competitive matches for gaining are recommended, and that means you have to consistently give players playing time - you'll note that the decay rate increases as a player becomes less match fit. Edit: On reflection and doing some testing, this part about friendlies hitting diminishing returns is in error. Unfortunately, I may have to re-evaluate certain conclusions I've drawn based on this presumption, which I'll revisit tomorrow.
If you do the math with the league schedule previously mentioned:
10 nat player needs to play every 8 days 12 nat player needs to play every 11 days (2 in 3 matches) 20 nat player needs to play every 21 days (1 in 3 matches)
So while for condition, a player should ideally play no more frequently than 1 in 2 matches, for match sharpness, a player typically needs to play either every match or 2 in 3 matches in most cases.
I accidentally pressed submit on this thread, so I will add more detail and my theory on squad composition in a followup post soon.
I did a test of dribbling specifically. HarvestGreen22 tested dribbling individually, and found it's the most important mental/technical attribute, with +8.2% win rate 12 > 18 drib; +33.5 goals 10 > 20 drib. But we don't know if that's still the case if the player already has high physicals, plus it's worth verifying the claim, so I did two types of tests:
8 vs 20 drib, with 20 pace/acc/jump + 8 other attributes
8 drib:
+52, 83 +77, 90 +67, 92 (came 2nd) +83, 92
20 drib:
+107, 97 +99, 95 +84, 97 +104, 100
Overall result: +30% goals, +10% points
-------------
140 PA 'ideal' templates with 8 drib and freed up CA reassigned randomly to other attributes vs 10-18 drib in my templates. I consider this test a more realistic indication of what you will get if you favor or disfavor dribbling.
8 drib:
+127, 104 +110, 106 +119, 102
10-18 drib:
+121, 107 +146, 107
Overall result: +13% goals, +3% points
--------------
So dribbling is quite significant, even with high physicals.
Note that my tests for this were poorly controlled due to laziness, but I feel it's adequate as a general indication. So it might be +20% or +40% rather than +30% say, but it's clearly a sizable difference. Maybe in the future I will do some more proper testing, so I can present it alongside results of other attribute combos or whatnot.
This was one of the surprising realizations I had reading EBFM's excel file, that the GK sessions are actually also very good for the outfield.
handling = 0.05 drib DC, 0.05 drib DL/DR, 1.15 con DC, 1.1 con DL/DR, 0.45 acc DC, 0.4 acc DL/DR, 0.55 pac DC, 0.4 pac DL/DR --- (1.8 physical) (4.15 key attributes) (11.4 CA). +0.9 agil/+0.8 aer/+0.6 ref for GK.
match practice = 0.05 drib DC, 1.15 con DC, 1.05 con DL/DR, 0.4 acc DC, 0.4 acc DL/DR, 0.65 pac DC, 0.6 pac DL/DR (2.05 physical) (4.3 key attributes) (12 CA). +0.2 agil/+0.6 aer/+0.3 ref for GK.
Handling does a lot for all positions too, not just defenders.
When I changed GK attributes in my 1 CA testing, it did make a difference, so GK has some significance. I'd say a fair bit more than I intuitively thought. Not sure exactly how much, but I'd guesstimate their worth is at least that of 1 of the 11 players.
What I didn't really mention in my post btw is that my thinking is that these sessions like handling that boost physicals moderately (~0.3) are perhaps particularly worthwhile because with a bunch of them maybe you could replace a whole quickness/resistance/physical session which gives you +1.2 but lowers a whole bunch of technicals and mentals.
Yarema said: These tests really need a bigger number of data points. Doing one 3 season test can maybe point in the right direction but far from conclusive, you can easily get deceived by random chance. Even EBFM's training videos I'd say are underpowered and probably harvestgreen's too (to my knowledge).
Don't get me wrong they are a great start and eye opening in some cases but people make way too strong conclusions from them. None of those tests can differentiate between 0,35 and 0,31 with enough accuracy.
Yes, I would agree with you there.
Though, as I think about it now.. given it seems to show patterns accurately enough (i.e. aerial defense gives 1.25 con DC, 1.25 con DL/DR, 1.2 con DM, 0.8 con MC, 0.5 con AMC, 0.65 con AML/AMR, 0.5 con ST) maybe it is actually non-random. Maybe the 'random' variation between different players is simply their professionalism, matches played, CA-PA diff, etc. downscaling a set max amount, and then you have different boosts for different positions giving the appearance of each player progressing seemingly randomly compared to other players.
I suppose it's something I should try and test.
Yarema said: Also you can't add up different benefits with various sessions as it was already tested by EBFM for example.
I haven't got that far yet
I see now he has an excel file with combo sessions, with positional breakdown. I'll examine it to try and deduce what's going on
In trying to create the best performing 1 CA players, I found that all that seems solid is pace, acc, jump, dribbling, concentration and perhaps flair. And only pace & acc is needed in all positions. For GK, it's clear that agility + aerial reach + reflexes are the key three.
I also know that decisions, technique, tackling, marking, and some others are often highly weighted but make no significant difference to results.
I was looking at EBFM's more precise breakdown for training session effects by position, albeit it is for FM23. But I've created a schedule based on its info, and it seems to still work the same.
The schedule I've created is:
1 x Quickness
1 x Goalkeeping
1 x Resistance
1 x Shot Stopping
1 x Match Practice
1 x Handling
1 x Match Tactics
1 x Aerial Defence
1 x Chance Conversion
1 x Play From The Back
1 x Distribution
Quickness focus (Agility on GK)
This is a significant departure from HarvestGreen's recommendations.
I've analyzed each session, and have chosen ones that favor pace & acc, whilst retaining/boosting as much as possible drib in relevant positions + concentration in DC+DL/DR + minimize decisions & technique gain (to free up CA) + agil/aerial/reflexes for GK + some other lesser considerations.
In my initial 3-year test of it, I'm seeing the best improving players getting +5 to +6 pace/acc, +4 agil/aerial on GK, and lesser but positive movement in the other attributes I mentioned. Requires refinement, but seems to be working close to as intended. I'm testing using just the default Bournemouth starting players, put the knap tactic on, and just let things run for 3 years. I did a test of Quickness + Match Practice + 2 x Attack + Quickness focus to compare. My schedule was competitive with it, but fell a bit short. GKs improved significantly better though I think.
Here is my critique of HarvestGreen's training:
Combinations are assessed according to acc+pace+jump and overall team CA boost/cost.
If you go with minimal CA cost for max acc+pace+jump, you have to do a lot of rest which usually results in match sharpness becoming unsalvagable which significantly affects win rate. A lot of the other attributes decline, yet some of these attributes must matter to some extent as acc+pace+jump alone fails to win.
Additionally, match practice possibly impairs team performance and may be unnecessary. And there are still a lot of rest periods.
If you go for high acc+pace+jump and high CA boost, a lot of that CA could be junk attributes and it doesn't tell you how much of other important attributes such as dribbling & concentration and whether they are put in the right positions or not. For instance, it seems dribbling is good on DL/DR, but unnecessary on ST.
I'll compare 2 training sessions (using EBFM's FM23 excel file) to illustrate the issues:
Attacking - +0.74 drib, boosts to many stats with no declines, +11.46 CA.
Aerial defence - +0.47 drib, boosts to many stats with no declines, +10.3 CA.
Attacking seems like the logical choice here. However if you look at the differences:
Attacking - +0.8 drib AMC (position not used in knap tactic), +0.2 drib ST (unimportant), +0.05 drib DL/DR (important). +0.38 decisions, +0.26 technique (costly near useless attributes that should be minimized).
Aerial defence - +0.35 con DC (important), +0.3 con DL/DR (important). +0.32 ant (semi-important?), +0.31 cmp (semi-important?). +0.13 acc, +0.18 pace. For 1.16 less CA.
Now Aerial defence is the no-brainer choice. Perhaps moreso if you consider that CA growth is limited to only a dozen or so CA a year on average.
This is where I get confused though. Maybe someone can help me understand. I understand that HarvestGreen is showing us that as rest boosts physicals by default, training is essentially an allocation process. But if I do quickness (+4.61 CA) + match practice (+12.01 CA) + attacking (+11.46 CA) does that all get compressed proportionally into say 12 CA?
For example, if a player has "20" for Natural Fitness attribute then his Condition attribute can NOT drop lower than "7,800" value.
This simply isn't true. I just played a game with my 1 CA 20 NAT players to confirm, my ST went down to 61% in-match, 67% post-match. Perhaps it is true if match sharpness is also 100%. And if memory serves me correctly, natural fitness does not even affect condition fall rate, it is stamina that does.
Zippo said: No, a task to achieve 10,000 value for the Condition is almost impossible because the rate at which the Condition attribute is restored after 9,000 level is very low.
So you should look for 9,000 - 9,300 range.
As your stats show, the performance loss would be significant. It's not impossible to maintain ~100% condition, it's quite easy. Another reason is that you will also otherwise suffer exponentially greater injuries, as EBFM found injury rates of:
(starting) 100% condition = 8 in-match injuries
(starting) 80% condition = 20 in-match injuries
(starting) 60% condition = 87 in-match injuries
And those injuries create a vicious cycle of low match fitness & condition leading to more injuries. I would guesstimate injuries double overall if starting at 90-93% condition.
Zippo said: How much does the Condition attribute decrease after a match?
It depends on many factors. The rate differs for positions and roles. Also, the Intensity of a tactic greatly affects the rate at which the Conditions are consumed.
Is that really so? According to EBFM's data, the difference between the most intense and least intense settings were 1-2% condition difference by end of match.
Zippo said: Also, the Team Cohesion and Match Sharpness drops a bit when players are "Resting" but it isn’t a big issue.
Match sharpness is very important, more important than condition. It's performance impact is greater, and it's harder to maintain. So using rest is usually a bad idea, from my analysis of it.
These results might not be directly comparable, but you say that starting at 88% condition results in 16% worse performance. EBFM found that match sharpness 100% > 90% reduces win rate by 33.3% and injuries increased by ~25%.
I know I come across as a bit of a knob here with all of this, but it's partly that I can't resist knocking down falsehoods, and partly that I wrote a post on here about this very topic recently so I'm a bit of a zealot about it. I know I state a lot of things that turn out to be erroneous myself. This is just part of the process of discovery.
The training sessions generally show lowering "Condition"
But from my observations, condition always improves after each day with just training, and only lowerswhen playing matches (or getting injured, but that's beside the point). So my question, has it ever been tested whether these trainings actually: 1. Lower condition? OR 2. Training sessions with heavier condition reduction don't allow the players to recover condition as much as training sessions with lighter condition reduction?
I did a quick test:
match practice doesn't decrease condition
1 x match practice recovers condition same rate as all rest
1 x match practice doesn't boost match sharpness (identical to all rest)
2 x match practice + quickness on one day, is identical to full rest in regards to condition recovery & match sharpness
What you might not know is that condition recovery doesn't have a randomness factor, unlike many other FM mechanics, so there shouldn't be any need for a ton of tests.
Seems EBFM found this back in 2022, and additionally found that match practice actually negatively impacts match performances, even though it has no condition impact.
This is strange. Maybe the 'injury risk' and 'fatigue' labels are also falsehoods and the whole training meta needs a rethink if one is not looking solely for attribute gain.
It would be at least most of the reason why.
You have to be a little careful extrapolating the 1 CA results to higher CAs. Even at 1 CA, it's clear other attributes do matter, just not much. For instance, I can say for certain that if you give ST higher finishing, he will do most of the goals instead of the wingers (that doesn't mean a necessarily better overall team performance though). But also it's possible that consideration of certain attributes only kicks in at say 120+ PA. This was the case in Championship Manager based on what I've read, and whilst that mechanic is probably gone now, I think there's a fair chance something similar still exists.
Another thing btw is that apparently the match rating is skewed towards technical ability, and biased against physical ability, even though physical ability beats the technical ability. Source. This is particularly interesting to me, because it helps to explain why Orion's coefficients doesn't produce a winning 1 CA team, even though it gets most of the way there - Orion's coefficients are basically comparing the match ratings, and if its biased towards lesser-performing technical ability, this would explain it.
Also I'm not 100% confident on flair yet, I should test it more, but I did a little testing, and so far it seemed it was beneficial in spite of what HarvestGreen found.
Mid-table results were easy with just 20 pace/acc/drib (not sure if I included jump), but it required careful tweaks — closer to Orion’s data than HarvestGreen’s — to push into the top four. Winning is barely possible and inconsistent, but the experiment taught me a lot about how attributes, combinations, fitness, morale, and form affect outcomes. I wanted to wait until I'd worked out a way to win consistently, but I'm a bit burnt out by it now - at least the main point is proven, that you can actually win with just 1 CA/PA players, and I may try and improve it more later on. Coming 2nd I can do consistently. I may also retest without the 210cm height & player traits, I actually doubt those impact the results much.
The winning result:
Scroll to bottom for attribute stats by position
Setup:
• Use Knap’s EF 424 IF HP V2 P101 AC tactic and Blue Routines set pieces.
• No strikerless exploit.
• Allowed: friendlies, training, captains; no reloads.
• Squad: 22 players, all 1 CA.
Findings:
- It is possible to win with 1 CA players without strikerless exploit
- Concentration ~16+ on CBs+FBs (no other combo) was key to winning the league
- Pace+Acc alone = 0 points, 0-7 loss typical
- Jump+Acc alone = a draw or two, 1-7 loss typical
- Drib+Pace+Acc alone = mid-table
- HarvestGreen data skew = bottom to mid-table
- Orion data skew = mid-table to top 4
- Orion data skew + pace/acc 16 or less = bottom
- Orion data skew + pace/acc 20 = mid-table
- Orion data skew + pace/acc ~17-20 (mix) = top 4
- 20 in key attributes + maxxing out other lowest weighted ones = high mid-table
- Lowest weighted attributes (i.e. finishing '1' weight on CBs) had negligible effect (i.e. no extra goals or overall team dominance)
- Most attributes make no statistically significant difference, very low values will even dominate many in-game stats
- (Only) wingers 1 pace/acc = high mid-table
- (Only) one slow ST target man = high mid-table
- League position more consistent than points (i.e. one case, almost always finish 2nd but was 66-91 points)
- Winning formula finishes ~4th majority of the time, 2nd place was more consistently 2nd
- Fitness management, particularly of match sharpness, is important - difference between ~mid-table and ~4th say
- Morale & form somewhat important, difference between ~4th and ~2nd say
- Squad rotation (even if perfectly done) or too much rest detrimental to results (maybe this is why AI managers rarely rotate, and why the game says 'I want to play you but X is in the form of his life right now' when players pipe up for more playing time)
- Altering knap tactic (i.e. 'very attacking'/'defensive', 'work into box', etc.) was worse than default
- 1 consistency doesn't matter, since consistency reduces CA but 1 is already the minimum; seemed to be the case in testing, but hard to verify 100%
- Biggest premier league win I got was 9-0
What stayed the same/correlated with top results:
GK - agility 15-20, acc 5-6, pace 5-8, sta 3-6??, composure 6-10?, work rate 6-7, aerial 16-20, communication 3-8?, reflexes 10-13
DL/DR - acc 20, pace 20, jump 20, off the ball 20? (unlikely), vision 5-7?, work rate 7-11, dribbling 20, finishing 20
DC - acc 15-20, pace 18-20, jump 17-20, strength 4-6, flair 20?, vision 15-20?, dribbling 20
DM - acc 19-20, pace 20, jump 20, bravery 12-20?, flair 20?, vision 3-4?, work rate 7-10?
AML/AMR - acc 15-18, pace 16-18, jump 20, flair 20, positioning 18-20, work rate 7
ST - pace 20, acc 18-20, flair 20, work rate 7
Question mark indicates I think it's unlikely to be essential.
I tested certain attributes (at '20', with CA boost added so no decrease of other attributes) specifically to see if they mattered (i.e. changed league position), most didn't. Some examples:
ST - finishing, agility, dribbling, work rate, decisions, stamina, strength, off the ball - not any of these
DM - strength, vision, dribbling - none of these
DC - agility, balance, strength, tackling, marking, heading - these do win/make close 2nd, but only slightly
I haven't tested all attributes and their combos exhaustively, so there may be some things left, but I tested a lot of them.
You can deduce a lot of the above yourself from the following images (I've kept the roles highlighted to show how misleading they are):
GK:
DL/DR:
DC:
DM:
AML/AMR:
ST:
Note: If you're wondering *why* I've changed certain attributes and in the way I do, it's that I'm following the money. The reason finishing went from '5' to '3' for strikers isn't because I never tested above 5, but because I probably did a dozen different combos with finishing above '5' that didn't work out. I think there are knock-on effects to other positions as well (i.e. if your winger has better finishing, he'll probably do most of the goals), so I'm kind of forced to just take the approach of 'if it works, it works' and I move on to try and improve another position until I can't squeeze anymore out of it and then move on to the next one.
The two most important player attributes are speed & work rate. This reflects reality accurately. The problem is when you can use HarvestGreen22's training regimes to turn literally any player into an elite player (pace/acc 8 > 20), and when you can give a full back 20 pace/acc/jump/nat/long/fin/drib/det/agg/off the ball/work rate and he'll be 1 CA with a handful of points left to spare.
I think you are correct in your hypothesis about how attributes are processed by the ME. Of course until we can see the game's code, we have to make our best judgement based on results and what SI staff have said over the years.
So this is what I understand to be the case:
The ME uses typically 2-3 attributes (sometimes up to 5) for each game event (i.e. passing + work rate = pass attempt)
An ATTEMPT ('choice' ) is made FIRST
A COMPLETION ('execution' ) is made SECOND
i.e.
passing + work rate = pass attempt
passing + decisions = pass completed
As you hypothesize, for many/most players, the only way to win is not to play at all. Instead of going for a slim chance of a slim chance of a pass leading to a slim chance of a goal, if you go with passing '1', then decisions can also be '1' since it almost never comes into play now, and you can now instead rely on dribbling '20' with pace/acc '20' to run up the park and score in ~17% of instances with [dribbling (20) + balance (7) = 67%(?) evasion chance] x [finishing (7) + composure (7) + decisions (1) = 15%(?) scoring chance].
Something I'm observing is that '20' pace/acc will win games where even '17' pace/acc would see you get demolished. I suspect this is because you simply run past/around the opposition's players, so the intercept & block chances are left out of the calculations. This is showing up in my 1 CA player testing, but I think it would still hold partially true for normal gameplay where you have weaker technicals/mentals/other physicals than your opponent.
Another thing I have observed is that while changing full backs to have drib/long/finish 20 (all players still only 1 CA) resulted in a lot more premier league wins, the full backs themselves rarely scored goals, or even assists. And looking at the match highlights, it doesn't appear to me to be that say fullbacks are making big forward runs up the side leading to later goals. This is still a mystery to me right now.
I do suspect that tactics give favor to certain attributes, and that meta tactics favor/exploit high pace/acc/drib, but I also think that only meta tactics tend to work, especially since your AI opponents (at least, the better ones) are also going to be utilizing these same meta tactics against you. I have noticed however that you can use a low pace/acc target man with no ill effect.
I think there may be something in thinking about the fact that attributes aren't denoting actual ability (i.e. you can even see in the 3d engine, that a dribbling '1' player can still dribble), but rather the only factors in decision outcomes. So a dribbling '1' player from Bedford Town dribbles just as good as Messi would in the game; it's only when Messi tries to tackle him that there is a difference. I think this partly explains why pace is so important, but technicals could be '1' and still do ok. Pace is active all the time, a slow player is always obvious and hampered in the 3D engine. But often a '1' dribbler looks just like a '20' dribbler (particularly when he has high pace to lengthen it out). Therefore the whole 'a player is only as good as his worst minimum' logic shouldn't be applied.
A similar potential misconception is that an attribute is simultaneously tendency and success rate. It seems true, but only to an extent: passing '20' gives you both more attempts & success, however a player with '1' passing still makes passes anyway, so probably '1' passing & '20' pace does more actual passes than '20' passing and '1' pace. So really '20' pass you could look at as say +5% tendency, +20% success, rather than +95% tendency +95% success.
Here's two other quirks I feel are worth mentioning:
If you go by what the staff-written guides say, strength alone can be used in lieu of tackling + marking + aggression. Strength is pretty highly weighted, but then again, it isn't subject to the consistency CA decrease like tackling/marking is, and it also says that strength reduces fouls as they can muscle the player off the ball instead of sliding in. For ST, tackling/marking is 1 weight, strength is 6, and a foul probably won't result in a penalty against you, so low strength probably makes sense here. But for DC, marking is 8 weight, tackling 5, and strength 6, and sliding in probably results in more conceded penalties. I'm yet to test this though.
If you think about it, you can set consistency to 1 for 1 CA players and it will make no difference, as consistency reduces CA for technicals to use. My early testing of this finds it does indeed make no difference. I suppose you could utilize this in a very low league team.
How much do the latest findings indicate the importance of pace/acc/dribbling and overall physical attributes?
Let me explain what I mean:
Player X: 18 Dribbling, 18 Pace, 3 Flair, 2 Technique
Player Y: 15 Dribbling, 15 Pace, 6 Flair, 12 Technique
In this scenario, Player X would clearly perform better than Player Y, right?
My impression is:
Orion's attribute coefficients are most accurate to use for each position, more effective than HarvestGreen22's attribute ratios
HOWEVER Pace/acc/jump 20 will override this (possibly some other attributes in certain positions too)
Pace/acc 15 will not be sufficient for the override. But this doesn't mean they still aren't highly valuable attributes, I'm just talking about having it override other attributes to the point where you could have a DC with 1 tackling, marking & positioning and it wouldn't matter, the 20 pace/acc player would still be better than the more balanced pace/acc 15 one
Details from old SI staff-written manuals, such as [forward runs - stamina, anticipation, off the ball] seem generally correct
HOWEVER such rules/combos can and should be broken to shift valuable high weighting attributes to valuable low weighting attributes. I.e. [DC - Positioning 2, Dribbling 1] should be changed to [DC - Positioning 1, Dribbling 20] (same CA cost) even though both Orion's attributes and the guides tell you to get positioning
My theory is that putting high weights on key attributes for the relevant position carries an implicit necessary corollary: those attributes are so heavily weighted simply to dissuade you from making 20 marking/tackling/strength DC beasts out of thin air, not because marking has 8x performance and your DC will actually only be hampered by ~1x performance if you decrease marking to 1. You take that 7x leftover and make 20 dribbling/finishing/long shot beasts instead and utilize them with a knap forward-throwing tactic
I've come to these conclusions by trying to get a 1 CA team to win the premier league. Here's a brief summary of how it's progressed:
Base my 1 CA players off my combined HarvestGreen22/Orion/etc templates - Near bottom of table.
Base it more on Orion's attribute coefficients - Mid table finish.
Try Orion's + decrease pace/acc to ~15 to put some 7s and 8s on some key mentals/technicals - Back to bottom of table.
Do the opposite, try to get pace/acc/jump 20 for most players while preserving what I can of what Orion & manual says - High mid table finish (6th, 66 points).
Reduce positioning to 1, get dribbling 20 instead, etc. - 4th, 74 points.
In your example, player X performs better. But there are some nuances. Flair and technique are generally unimportant, even detrimental in certain positions. However let me show something that astonished me:
Player X (ST): 20 pace, 19 acc, 1 jump, 1 strength, 6 finish
Player Y (ST): 14 pace, 9 acc, 19 jump, 8 strength, 3 finish
You would assume that Player X is far superior. However I used these 2 players as STs in my 1 CA testing, and they got me to that 6th-66 points result. But what's more, Player Y scored 7.08 rating with 29 goals in 55 appearances, while Player X scored 6.42 rating with 15 goals in 57 appearances.
Now you could theorize and say, well that's because set piece headers are the only way you're going to score goals with 1 CA, or that perhaps the wingers are doing the work throwing it to the target man, or something else. But my point would be is that your attribute choice, even pace/acc and down to such low levels, is genuinely subject to your tactical setup in a certain minority of cases. Note that if either player were actually *bad*, the team wouldn't have ended up in 6th spot.
That does not mean we are back to where we started. I suspect ST target man is one of 2 or 3 roles that can work with low pace/acc. On the other end of the spectrum, for DC, 20 pace/acc is simply essential.
I suppose a general rule of thumb you could use is physicals over mentals, and mentals over technicals. If you choose strength 16 over 16 dribbling or 16 composure, it's almost a sideways trade and this is one of the worst you could do with this rule, whereas trading pace/acc for anything is going to be iffy or catastrophic.
I'll explain my process a bit first. Let's take a more nuanced attribute, 'dribbling'.
First I look at other people's findings:
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.
And then there's the FM Arena testing of course.
Orion uses a different method, he derives his attribute values from the match ratings from the actual leagues as they exist in the game, and he does it for each position. For the DR position, dribbling is ranked 6th. For DC, it is not in the top 8.
But for DR and DC, I plan to give them perhaps ~14 and ~16 dribbling respectively in my updated ideal templates. This is for several reasons:
1. With the Knap tactic I'm using, DC has 'dribble more' instruction. Normally this would be an automatic ~2 boost for me.
2. Dribbling has a very low 1.0 weighting for DC. After doing the math on it's relative value once weighted, it's clear this is worth setting very high.
3. I've read a comment that perhaps the match engine favors dribbling in general, not necessarily because dribbling is good in itself, but because simply continuing to run with the ball prevents other calculations coming into play. For instance, if a player decides to pass instead, then a decision/technique/etc error may occur, and all these calculations cause time delays.
4. According to the SI staff hints from long ago, dribbling combines well with high pace & acceleration. It intriguingly adds 'don't have too many dribblers as they'll tend to ignore any passing tactic you might have'.
5. Someone (HarvestGreen22?) theorized that attributes are simultaneously ability and tendency. Given that I expect my DC and DR to still do a fair bit of passing, that means dribbling needs to very high to take precedence.
6. A dampening factor must be applied, which is the actual availability of the attribute value. In a save I've loaded up, there are 489 DCs who are 100+ PA w/ 12+ dribbling. There are only 18 who have 14+ dribbling. Technicals decline or stagnate with HarvestGreen22's optimal training. Let's say you could take 1 of the 120 players (or newgen) with 13 drib and boost it by 1-2 with training, that's 14-15 max dribbling you're looking at. For DR it's a bit better, a small handful already start with 16. If it were not for this reason, I would get 20 drib for both positions (this is what I'm trying to communicate through my 'high' ideal templates).
In this case it's quite clear that regardless of whether the speculations are true or not, it's well established that dribbling is high performance and/or low cost. You can safely assume it without even running tests.
But sometimes there are contradictions, or nuances that haven't been explored enough, and these I perform tests to clarify.
I seem to be finding that technique and flair aren't always bad as HarvestGreen22 found them to be. Technique seems best at ~6 for most positions, flair I have around ~12. I get at least about ~10% performance loss if I drop either to 1, and 20 technique comes at high CA cost but does no better. I'm not sure what is going on here exactly. Perhaps it is that technique increases the number and quality of so-called 'tricky passes' - too much would detract from dribbling, too little would result in exponentially more lost possession, '6' takes advantage of only the most assured & promising tricky passes to nab a few extra goals that make up for any lost possession - particularly in top leagues, where the easier chances are very few and far between. Pure speculation of course.
Some others I wonder about are strength, agility and stamina. I figure that knap tactics can be weaponized against you in just the same way by certain AI managers, and that attributes have direct counters. So the counter to your AML with 20 speed (acc then pace) making a 10m charge towards goal cutting in from the left flank may be 15 agility (rapid direction change to sideways left) + 10 acc/pace to reach your AML in 3m the time it takes him to do the 10m. So if your DC has 20 agil/acc/pace, good luck to your opponent.
Yet we also see that HarvestGreen22's results show high agil/strength does relatively little. But then again, in his test his opponent has DCs with 10 agil/strength. In a tussle or direction change, they'll neither totally dominate you nor lose every contest. And what is clear is that 10 - that is, equaling them - does the job. Now in the premier league the best DCs you'll be coming up against have agil/pace/acc/strength around ~15. My theory is that it's probably best to at least equal that for optimal play, as your chance of winning a tussle probably exponentially decreases the wider the gap between the DC's strength and your striker's strength.. so agil/strength 15 on a striker. And if you put strength down to 6, because strength doesn't seem to overall matter much, are you sure you know what you're getting yourself into, because HarvestGreen never put a 6 strength striker up against a 15 strength DC.
So that is why I am in two minds about these attributes in particular. What I've found so far is that 6 str/11 agil is significantly worse than 15 str/agil, but it's not that bad (~7% worse maybe). For comparison, I found difference in off the ball to be fairly negligible.
Note that in the SI staff notes, it says of agility 'agility is an attribute which is most necessary for players in top flight football...combined with high acceleration and dribbling a high agility rating can make for a great attacking footballer'. And if you look at top goalscorers, they seem to have high agil/str - Haaland has 17 agil/17 str.
I have also tried to replicate the success of the Chinese forum member who won the Premier League with 1 CA players a few years ago. I've managed to nab a few wins and draws, but overall no cigar. This tells me that there are indeed certain non-meta attributes that you need to have certain minimums in for success, which makes me feel more justified in being wary of 6 strength and 1 technique.
tam1236 said: You can - tested many times. Especially at youngsters and with close positions. Of course it depends on versatility hidden attribute.
In theory professionalism is better - somebody on yt tested that there is no chance to make a good player with low prof. and sometimes it happens when a determination is low. But irl players with really low profess. are very, very few and very, very far between in FM base and in newgen. Balotelli is a lonely knight. And low determination you can see quite often
I'll take your word for it on the position proficiency. I haven't done any testing on it, I've just always had the assumption from way back that 18 is the limit and I don't retrain players to new positions in my own saves.
I'll probably do a quick test of it to make sure later and update my main post later to reflect the fact that 20 proficiency can be trained.
In regards to pro vs det, aren't both just random generated for newgens now? To me, newgens are the main source of high PA signings. If I search in genie scout for players >140 PA w/ det 10< I get 472 players, 598 for pro 10<. So I still think it's clear if you had to choose one to search for, it would be high professionalism.
I didn't look into the training units at all myself, but according to EBFM training units do have an effect
In the schedule I favor, it doesn't seem any of the units are split between attacking and defending, not even the 'attacking' units, so I would assume it makes no difference what group the players are in.
I'm doing FM24, not FM26
Right now, I wouldn't recommend x7 recovery. But it doesn't matter much, we're probably talking like ~5% or less difference. I recommend Quick + Match Practice + Attackx2 + Double Intensity + Quick focus (Agility for GK) + remainder rest. 'No pitch or gym work' for first 3 condition icons, only 4-5 pre-season friendlies.
Is this so? I didn't know this
I am skeptical. If this is true, then why do recovery sessions seem to impact training results as compared to rest? Is it only because of the 3 period rest days? How can recovery also make such a large condition/jadedness difference over time if its just those infrequent 3 period rest days?
Steelwood said: Another post talked about how important match sharpness is to results so I think it is probably better to use recovery where possible, though my thoughts on that are more skewed towards winning matches rather than training results
I agree with ZaZ's comment that rest works for match sharpness so long as you manage your squad correctly through rotation and playing them occasionally in youth friendlies if necessary.
I actually think recovery has more downsides than just slightly worse training. In my own save I was using the exploit trick to use all recovery sessions instead of rest, and found it was a bit too harsh on condition and caused excess jadedness. Jadedness isn't that bad as EBFM found it's self-limiting and it's difficult to reach and easy to solve, but it does have some impact on performance. So I've been advocating for a mix of rest and recovery, but after doing some testing, you can actually get away with no recovery sessions at all. And for the slight training bonus, you would want to.
Semi-related - At the moment I'm re-analyzing attributes. For some reason a few months ago I decided quickness + attack was the best, without match practice. When I returned to it for my post here recently, I thought I made a mistake of simply forgetting match practice. But now I vaguely recall it could be that although it's a "worse" training regime, it favors 'decisions' less say, making it actually better overall. So for me, the best training regime is still in flux, I will re-assess after completing my attribute re-analysis.
What's more if You have money to buy 140 PA young players there's no need to risk with very slow (speed 8) 18-20 yo players (BTW if you don't cheat with GS You don't know CA/PA
2. I would say determination is in a way more important because I dont remember players with professionalism 6-7, and not talking about less (I always try to search minimum in my teams for my information) but determination 2-3 is quite often in lower leagues - and drastically slower youngster progress - there is a hope for such a player but if you are not desperate dont risk with him.
5. You can change natural position and reach a new one to level 20 . Especially if You have a good newgen playing as CB with aerial reach ~9 you should do it. Apart from everything else aerial coaching doesnt work good.
6. If (IF!) You have money - why not.
7. Avoid players with low mental - concentration, compassion, determination have a great impact and are really hard to improve. From my experience GK with low consistency is a fail. Injury proneness IS important if You dont have good subs - on the other hand a game is buggy and player with this attribute 5 can have (in 4 years time) more injuries then the one with 15 - but statistically it is important, though You dont see these numbers without GS.
And maybe I'm wrong , but I have an impression that in FM24 loaning above-average young players to lower leagues doesnt work very good because their progress is capped by the league level.
Mostly agree except a few small things
Determination 3 > 15 is equivalent to professionalism 5 > 13. But also 20 Determination = 14 professionalism, and determination 6 does most of it. (source). I do say on my initial post to get 10+ determination for optimal, but because CA growth is capped each year you only really need say half the optimal, so having 13 professionalism alone is probably enough to do the job.
Also there was another reason professionalism is better, I can only vaguely remember now; I think it was that professionalism allows the player to develop even if they don't end up getting match experience.
I may be mistaken but I don't think you can train player position to 20, it can be 'natural' but it will still be 18. That said, I was doing some more testing of 18 vs 20 and it seemed to have negligible effect. I was testing it in the first place because I figured it's likely 15 > 20 isn't going to be linear. I'll have to run it a few more times to be sure there is no significant effect of 18 > 20.
From what I've heard, the FM26 training file is identical to the one in FM24, and HarvestGreen22 found that attribute importance remains largely the same (long shots, finishing were boosted somewhat)
Separate to that, I wouldn't use recovery. It actually seems to affect training gains, and you can maintain match sharpness well enough with this instead (this also has the most gains in FM24 according to Piperita a few posts above): Quickness + Match Practice + Attackx2 + Double Intensity + Quickness focus (Agility for GK) + remainder rest
Man City default:
Before final match - condition 83-91%, match sharp 91-95%
After final match - condition 69-80%, match sharp 96-100%
Pure rest:
Before final match - condition 84-91%, match sharp 66-80%
After final match - condition 67-81%, match sharp 72-86%
Quick + Attack + Match Practice + Quick focus + double intensity:
Before final match - condition 84-92%, match sharp 85-92%
After final match - condition 71-81%, 93-96%
So HarvestGreen22 training is slightly better for condition, but a little lacking in match sharpness. However this may be about the right balance, as this was still before end of pre-season. Pure rest training is clearly problematic.
Training & rotation is something that has to be micromanaged as you go along, but here I've kept the same starting starting 11 for about 2 months in (5 friendlies + 6 competitive) and on the HarvestGreen22 training regime:
Before final match - condition 96-100%, match sharp 96-99% (all but one are 99%)
After final match - condition 69-83%, match sharp 100%
So it happens to be well suited to fitness management to begin with. But if you were to try to keep your subs & backups at 95%+ match fitness, that would no doubt change things. So I did a full rotation (using only 7 subs though) every match:
Before final match - condition 95-100%, match sharp 79-99%
After final match - condition 73-99%, match sharp 83-100%
Those figures include all the subs. 11 players had 100% match sharpness after the final match. So overall, it's not bad, but definitely needs more match sharpness for subs.
Now going by Piperita's clarification on the best HarvestGreen22 training regimes, and taking into account match sharpness, the top ranked one still overall looks viable: Quick + Match Practice + Attackx2 + Double Intensity + Quick focus. Since injuries are an enemy of match sharpness, I won't change the middle condition to double intensity, but just try and manage the players better. Additionally I've promoted more players to the first team squad, to make it 18 outfield + 2 GK, and added some more pre-season friendlies.
Results:
Before final match - condition 95-100%, match sharp 62-100%
After final match - condition 69-100%, match sharp 78-100%
My observation here is that there were a significantly greater number of injuries, which led to low match fitness in a number of players throughout. Resting from training was also more frequent. You probably want just enough match sharpness so that low condition doesn't result in missed training days, which means either using a lighter training regime, or not scheduling too many friendlies.
I decided to have another go, cancelling the default friendlies and not adding too many. Because why not try and see if one can hold onto the optimal training regime.
Results:
Before final match - condition 96-100%, match sharp 89-100%
After final match - condition 67-100%, match sharp 92-100%
So if you manage things right, it does work pretty well. I used just 4 pre-season friendlies here. I'm sure with some minor adjustments, you could get maintain match fitness of all players at 95%+ without having to field any low match fit players in competitive games (a handful of the backup/sub players could be played in u23 matches for the first month after pre-season to get them to 95%+).
tl;dr Use Quick + Match Practice + Attackx2 + Double Intensity + Quick focus (Agility for GK) training regime every week, even on congested schedules. 'No pitch or gym work' for first 3 condition icons. Don't do too many pre-season friendlies, it's counter-productive, 4 or 5 might be best. Full rest training lowers match sharpness far too much, but you don't need to change rest to recovery sessions or increase intensity at lower condition.
It says which training regime as caption when you expand the image. That one is 'Walters_mixed = My dense but optimized regime based on theories above. 80 injuries.'
I can't remember what training schedule I used, but it doesn't matter because it was inferior to the HarvestGreen22 training regardless of whatever that image may indicate to you.
Now, the implications for squad composition. If you think about it, this is really something that hasn't been examined at all, in spite of how consequential we all know it to be - that ill-fated tendency of us to bloat our squads; what even is the actual optimal number of players for a first team squad?
In my view, the number is 20 outfield + 2 goalkeepers + the occasional promotion from reserve/youth squad or loan to cover an injury.
9 matches need sub replacements (10 x 9)
17/53 matches permit sub replacements (6 x 17)
26/53 matches permit sub replacements for high Nat players (4 x 26)
4 developing players (AML/AMR/DL/DR) get 35 matches
6 players (2 ST, 2 DM, 2 DC) get 26 matches
Remember that 25-30 matches/season is the ideal for player development. The reason why the wonderkids should typically be wingers not centre backs or strikers I would reason is that you don't want to interrupt your ST's bucketload of goals and young players aren't usually developed enough to fill the shoes of a proper DC, but also that wingers seem to tire fast and only have 1 of each position instead of 2 on the pitch.
It is definitely adjustable to some extent according to personal preference. Basically, for each wonderkid you add, make one of your starters a high Nat player so they can handle the reduced playing time.
Some extra info:
My figures are based on using the optimal HarvestGreen22 training regime, using rest not recovery. So it is safe to combine this info with those training regimes.
Stamina does NOT effect condition recovery post-match. It only affects condition usage during the match itself.
Having a physio reduces match sharpness decay significantly, but it doesn't matter how good the physio is or how many extra you have - you just need one for the benefit.
50% match sharpness or 1 natural fitness recovers condition to 66% vs 68% for 100% sharpness. 20 natural fitness recovers condition to 74%. I mention these select figures to illustrate that natural fitness is kind of important to condition recovery, and match sharpness less so, but overall condition recovery is pretty stable and predictable regardless of the player.
There is a way to kind of exploit match sharpness gain:
friendlies:
~10 minutes for 100 (1%) match sharpness gain
~25 minutes for 200 (2%) match sharpness gain
~43 minutes for 300 (3%) match sharpness gain
~60 minutes for 400 (4%) match sharpness gain
~78 minutes for 500 (5% - max) match sharpness gain
Source: EBFM
So if you want to play absolutely optimally, you could sub a player at 78 minutes in friendlies to maximize match fitness. And throwing a player on for 10 minutes gives the most efficient match sharpness boost, while you'd know not to do 7 minutes as it gives them nothing.
For competitive as I mentioned before, it's 9% max instead of 5% max. However, EBFM did find that sometimes players would inexplicably get 10-11% max for a reason he could not identify, but I would just treat this as a kind of bonus.
9% match sharpness gain per 90min competitive match @ 50-65% sharpness
8% match sharpness gain per 90min competitive match @ 70% sharpness
6% match sharpness gain per 90min competitive match @ 80% sharpness
2% match sharpness gain per 90min competitive match @ 90% sharpness
I feel like this topic deserves it's own thread and I've done some original research for it. And I think that in a way it's one of the last remaining actual challenges of the game to make it fun. If you don't use gegenpress tactic or pick/train high pace/acc then you lose a lot, simple as that. But even if you work out the mechanics of fitness management as I am trying to do, it remains a challenging and rewarding aspect of the game. In fact, just knowing how much it effects win rate to me brings back some significant enjoyment to the game.
My overall impression is that counter-intuitively to my mind, it's better to sacrifice condition for match sharpness. Maybe this is why the AI often/always plays players with 87% condition without any rotation. It's because low match sharp players suck in terms of performance, low condition not so much (injury risk is bigger downside here). The other thing about it is that you can deal pretty easily with the problems of low condition - rotate them, pick low injury proneness players, rest them. Match fitness is much trickier, there is only one modest offset for it (natural fitness), it also has a high injury downside, you have to pay the price of lower performance to get their match fitness up if you neglect it, and as I'll show below even friendlies have a limit to recovering match fitness so there's no get out of jail free card really.
I've taken this graph from EBFM's video on match sharpness. What it's telling you is that the difference between 90% and 100% match sharpness is 33% difference in win rate.
For comparison, the difference of pace/acc 10 > 18 is 23.5% win rate according to HarvestGreen22.
To boot, match sharpness significantly impacts injury rate, even at 90%:
So how do we optimize for 100% match fitness of all players every match? And also, how to handle condition in concert with match fitness? There's a lot of variables to consider, it's fascinating to contemplate, but I will try and lay things out in the most straightforward way:
Starting at 100% condition, 90 minutes will reduce to 75% condition (typically).
Condition & match fitness update at midnight (00:00), starting the midnight that may be just a few hours after the match.
It takes 10 days for full recovery to 100% condition.
75% (day 1 - 16:00) > 83% (day 2 - 00:00) > 90% (day 3 - 00:00) > 93% (day 4) > 95% (day 5) > 96% (day 6) > 97% (day 7) > 98% (day 8) > 99% (day 9) > 100% (day 10)
In FM24, Man City (English Premier League) has the following schedule in a save I looked at:
67 games (excluding national games):
38 league
13 cup
16 continental
Total recovery days between matches:
53 x 7 days
4 x 3 days
5 x 2 days
1 x 16 days
2 x 13 days
1 x 11 days
1 x 9 days
So there are 9 matches where condition recovery will be insufficient, and another 53 which would be sub-optimal. With a weaker team with a less filled up schedule, such as Burnley, I found it could be as low as 6 + 35. In both cases, we would conclude a player should be ideally be rested every 2nd match.
But before we continue, let's look at the other half of the picture, match fitness:
after 3 weeks, match fitness dropped to 98% (recoverable in 1 game) (20 nat)
after 4 weeks, match fitness dropped to 94% (20 nat)
after 6 days, match fitness dropped by 1% (10-12 nat)
after 8 days, match fitness dropped by 2% (10 nat only)
after 9 days, match fitness dropped by 3% (10 nat only)
after 11 days, match fitness dropped by 4% (10 nat only)
after 12 days, match fitness dropped by 5% (10 nat only)
after 13 days, match fitness dropped by 6% (10 nat only)
after 14 days, match fitness dropped by 7% (10 nat only)
after 15 days, match fitness dropped by 8% (10 nat only) (99% > 91%)
after 18 days, match fitness dropped by 11% (10 nat only)
after 21 days, match fitness dropped by 15% (10 nat only)
after 4 weeks, match fitness dropped by 22% (10 nat only) (99% > 77%)
The gain rates are:
76% > 88% (1st friendly)
88% > 90% (2nd subsequent friendly)
90% > 91% (3rd subsequent friendly)
91% > 93% (subsequent competitive match)
Friendlies gain match fitness at 55% rate of competitive. As you can see, this is adequate for when the player has low match fitness, but once it hits around ~90% you run into the following complication: You can't have a friendly the day straight after the last one, so you are likely giving at least 2 days rest, yet only gaining 1% match fitness with the friendly.. so potentially you are losing 1% match fitness to later gain 1% match fitness. Hence, at 90%+ match fitness, competitive matches for gaining are recommended, and that means you have to consistently give players playing time - you'll note that the decay rate increases as a player becomes less match fit. Edit: On reflection and doing some testing, this part about friendlies hitting diminishing returns is in error. Unfortunately, I may have to re-evaluate certain conclusions I've drawn based on this presumption, which I'll revisit tomorrow.
If you do the math with the league schedule previously mentioned:
10 nat player needs to play every 8 days
12 nat player needs to play every 11 days (2 in 3 matches)
20 nat player needs to play every 21 days (1 in 3 matches)
So while for condition, a player should ideally play no more frequently than 1 in 2 matches, for match sharpness, a player typically needs to play either every match or 2 in 3 matches in most cases.
I accidentally pressed submit on this thread, so I will add more detail and my theory on squad composition in a followup post soon.
8 vs 20 drib, with 20 pace/acc/jump + 8 other attributes
8 drib:
+52, 83
+77, 90
+67, 92 (came 2nd)
+83, 92
20 drib:
+107, 97
+99, 95
+84, 97
+104, 100
Overall result: +30% goals, +10% points
-------------
140 PA 'ideal' templates with 8 drib and freed up CA reassigned randomly to other attributes vs 10-18 drib in my templates. I consider this test a more realistic indication of what you will get if you favor or disfavor dribbling.
8 drib:
+127, 104
+110, 106
+119, 102
10-18 drib:
+121, 107
+146, 107
Overall result: +13% goals, +3% points
--------------
So dribbling is quite significant, even with high physicals.
Note that my tests for this were poorly controlled due to laziness, but I feel it's adequate as a general indication. So it might be +20% or +40% rather than +30% say, but it's clearly a sizable difference. Maybe in the future I will do some more proper testing, so I can present it alongside results of other attribute combos or whatnot.