Why can't the AI manage fitness levels?

[Ant26587]

I'm constantly coming up against sides that are starting players with 60-70% fitness, this happes in big leagues against big teams with huge squads.

It's abit unrealisitic to be playing against knackered players evey game, it isn;t even a matter of playing big names every game - I do that and sub them at 70 mins when we have multiple games.

[TCSawyer]

I've noticed this a few times playing against teams with 60-70% fitness, it happens everytime I play in the club world championship (when I make it). I don't see this as a bug I see it as an advantage!

[Saint_Lane]

Yes I agree with this.

Seen it across Europe's big leagues. Particularly teams with a heavy European schedule. Fatigue is to be expected for these teams, but the AI does not appear too competent in managing squad rotation and will drive players into the ground.

I'm my Las Palmas save when I've played Madrid or Barca after they have a mid week Europe tie they will routinely name the same XI despite heavy fatigue. I'm talking 60/65% Match Fitness. 

[Guest]

Seems like an expected consequence of the new dynamics feature & sounds worthy of a much deeper analysis. Are the affects of limited recovery time & fatigue due to selecting the same players reflected in individual performance drops & conversely is rotation to minimise this affecting team cohesion to an equally appropriate level?

[BigRoboCrouch]

Yeah i've noticed this aswell, and they also tend to start a lot of orange injured players.

AI seems to be much worse this year for some reason.

[Xterno50]

Yes, i've noticed this too playing in the MLS, sometimes 2-3 injured players will start

[cromscott]

Agreed. This needs to be looked at.

[phnompenhandy]

Well hang on - you're ASSUMING low condition affects the players. But does it? I have a very small squad of academy kids and am sometimes forced to play low condition players myself. If a cup match goes into extra time some can be sub-50%, yet I hardly notice any affect on their performance levels. Same goes for the AI teams.

[Flohrinho]

I think it actually mainly affects the risk to pick up an injury. At least I have seen totally exhausted teams having a go at mine for 90 minutes, no signs of slowing down so I'm not really sure it's that relevant apart from risking players getting injured.

[Guest El Payaso]

One decent way to get round this is to edit the AI managers to have 20 for squad rotation. Seems to help out a bit. 

[rdbayly]

I am increasingly of the opinion that physical condition and performance aren't related at all. I have experienced countless matches in which the AI switches to a high tempo aggressive strategy late in the game when most of their team are around 60% fit. They suddenly muster the energy to bomb up and down the pitch as if they were on steroids, showing zero signs of fatigue.

The game is telling us via the medical centre and other indicators to carefully rotate and manage injury risk. Yet when I see this time and time again I wonder why I bother.

[Bry]

5 hours ago, phnompenhandy said:

Well hang on - you're ASSUMING low condition affects the players. But does it? I have a very small squad of academy kids and am sometimes forced to play low condition players myself. If a cup match goes into extra time some can be sub-50%, yet I hardly notice any affect on their performance levels. Same goes for the AI teams.

Oh come on. If condition doesn't affect performance in FM18 then we may as well all pack up and go home.

"A player's condition represents how much energy and freshness he has. The higher this is the better he will perform and the less likely he will be to get injured, either during matches or in training." (guidetofootballmanager website)

[phnompenhandy]

8 minutes ago, Bry said:

Oh come on. If condition doesn't affect performance in FM18 then we may as well all pack up and go home.

"A player's condition represents how much energy and freshness he has. The higher this is the better he will perform and the less likely he will be to get injured, either during matches or in training." (guidetofootballmanager website)

Yes, I know what we've always been told and I accepted it for years; all I'm saying is that my experience doesn't correlate with such claims. Have YOU actually noticed those low-conditioned AI controlled players being noticeably slower?

What I DID see a couple of days ago was pretty cool: my assman informed me that my star striker had tweaked a groin. He was clearly struggling to move on the pitch. As I prepared to replace him he ran for a through ball and suddenly pulled up, hopped a couple of times and fell over, crocked. But that's an in-match injury; what I'm saying is that reduced condition due to fatigue doesn't appear to have the effect we all think it has.

[Bry]

3 minutes ago, phnompenhandy said:

Yes, I know what we've always been told and I accepted it for years; all I'm saying is that my experience doesn't correlate with such claims. Have YOU actually noticed those low-conditioned AI controlled players being noticeably slower?

What I DID see a couple of days ago was pretty cool: my assman informed me that my star striker had tweaked a groin. He was clearly struggling to move on the pitch. As I prepared to replace him he ran for a through ball and suddenly pulled up, hopped a couple of times and fell over, crocked. But that's an in-match injury; what I'm saying is that reduced condition due to fatigue doesn't appear to have the effect we all think it has.

If that's the case mate, it would be the most depressing thing since seeing The Last Jedi.

But yeah, it's good to challenge these things and you're right that players never look tired in the ME. (However even in real life, player tiredness is almost imperceptible in high level football, such are the incredible modern fitness levels.)

It would be an interesting experiment to run the same game twice with contrasting levels of fitness, or something akin to that. I don't have the editor so don't know if that's possible.

[Guest El Payaso]

36 minutes ago, phnompenhandy said:

Yes, I know what we've always been told and I accepted it for years; all I'm saying is that my experience doesn't correlate with such claims. Have YOU actually noticed those low-conditioned AI controlled players being noticeably slower?

I remember back in FM 2015 and a Copa del Rey final against Atlético Madrid. They had a mid-week tough UCL game (I think the semi-final) and the whole first team came into the final with 70-80% condition. Never won them so far in my save game but in the final I thumped them 7-0 without Atlético getting a single shot on goal. So in my experience it most certainly has a huge impact on performance and is also one of the reasons why we human users have a great advantage in a long league run while in single games the AI can actually perform really well.

[Guest]

5 hours ago, Bry said:

Oh come on. If condition doesn't affect performance in FM18 then we may as well all pack up and go home.

"A player's condition represents how much energy and freshness he has. The higher this is the better he will perform and the less likely he will be to get injured, either during matches or in training." (guidetofootballmanager website)

All the testing I did prior to FM18 indicated that low condition had very little affect on match performance, this may or may not apply to 18 but the main issue I saw was a lack of any drop in decision making & other mental attributes, those should see a greater impact when suffering from fatigue or low physical condition in matches.

[Bry]

16 hours ago, Barside said:

All the testing I did prior to FM18 indicated that low condition had very little affect on match performance, this may or may not apply to 18 but the main issue I saw was a lack of any drop in decision making & other mental attributes, those should see a greater impact when suffering from fatigue or low physical condition in matches.

If you did all that testing Barside, then fair enough. Did you provide SI with the results of your testing and raise it as a bug? Condition management is supposed to be an integral part of FM and if it has little/no effect that would be shocking.

Edit: Or if not, would you mind sharing how you conducted the test and found your results? I'm gonna try a few of my own.

[Svenc]

17 hours ago, Barside said:

All the testing I did prior to FM18 indicated that low condition had very little affect on match performance, this may or may not apply to 18 but the main issue I saw was a lack of any drop in decision making & other mental attributes, those should see a greater impact when suffering from fatigue or low physical condition in matches.

I don't find the post anymore, but I think it was argued that the match fitness levels at which "performance" are impacted are quite lowly. All I did find was a reply I got last year when being worried about AI.   In particular, as all the soak tests seem to be done with AI likewise, so it seems to made sense to me that I can't even remember an injury crysis for my teams likewise, as they are consistently at better fitness rates (requires no masterclass whatsoever), so were less likely to suffer the added injury caused by tiredness.

Purely anecdotical I had a play-off match in some recent iteration where the oppositition side was down to a curious sub 70% match fitness from kick-off across the squad (likely a bug), upon which the play-off was a walk in the park. I know that's not entirelly fair, but sometimes you wonder whether things are toned because AI would suffer first here... [This is still written under the shock impression how widespread nonsensical AI tactic decisions are, which had been out by FM 13 already.] It seems the more complex the factors to consider, the less robust AI is going to be, in a sense.

 

[phnompenhandy]

5 hours ago, Svenc said:

I don't find the post anymore, but I think it was argued that the match fitness levels at which "performance" are impacted are quite lowly. All I did find was a reply I got last year when being worried about AI.   In particular, as all the soak tests seem to be done with AI likewise, so it seems to made sense to me that I can't even remember an injury crysis for my teams likewise, as they are consistently at better fitness rates (requires no masterclass whatsoever), so were less likely to suffer the added injury caused by tiredness.

Purely anecdotical I had a play-off match in some recent iteration where the oppositition side was down to a curious sub 70% match fitness from kick-off across the squad (likely a bug), upon which the play-off was a walk in the park. I know that's not entirelly fair, but sometimes you wonder whether things are toned because AI would suffer first here... [This is still written under the shock impression how widespread nonsensical AI tactic decisions are, which had been out by FM 13 already.] It seems the more complex the factors to consider, the less robust AI is going to be, in a sense.

 

Yes, this is what I have been trying to point out. It's not necessarily a bug; rather we have the wrong expectations (admittedly due to being implicitly misled by SI).

The play-off match: in some leagues such as in Scotland and now the Vanarama play-offs in England, the lowest team that qualifies for the play-offs can find itself playing a ridiculous 4 matches in just over a week, so those low conditions are not unrealistic. As a human manager I've learnt to make this factor work in my favour. For example, last night I played the biggest match in my young club's history: in the cup our non-league club saw off Edinburgh City, Montrose and Clyde to draw the mighty Hearts. For game at home we drew, leading to a money-spinning away replay. Thanks to a Hearts' red card we survived into extra time. Many of my players were down to 60% condition, but playing well. With the knowledge I've expressed here, I was not tempted into substitutions too quickly and was able to wait until I'd taken the lead to make tactical substitutions and hold off for a historic victory.

[Miravlix]

On 12/19/2017 at 15:30, BigRoboCrouch said:

Yeah i've noticed this aswell, and they also tend to start a lot of orange injured players.

AI seems to be much worse this year for some reason.

They aren't injured.

 

There is 3 states.

Orange Knock

Red Injury

Orange Recovery.

 

I know it can be confusing that Orange means two things.

 

Time to go home I guess, because I really don't see how fatigue effect AI teams.

[Lem]

It doesn't need to manage it, players perform perfectly well at low condition or with knocks, even late into the game or in extra time.

[HUNT3R]

Performances are affected from either 60% or 65% (I can't remember which, but probably 60%) so there's no issue with a 70% Condition player on the pitch.

[DP]

That’s a bug then I’d suggest. 

Below 90% and the text says the player is ‘tired’. It’s at this point most people make the decision to rest the player. If it’s only 60% where there’s visible impact then the cue needs to change for the player as we’re being misled. 

[phnompenhandy]

1 hour ago, DP said:

That’s a bug then I’d suggest. 

Below 90% and the text says the player is ‘tired’. It’s at this point most people make the decision to rest the player. If it’s only 60% where there’s visible impact then the cue needs to change for the player as we’re being misled. 

As I've pointed out, it's not a bug, but we have been misled and the text ought to clarify the issue.