Guide: Condition loss and Ability gain in games

[Al Svanberg]

BETA

This is very much on beta level so far. But I've done extensive testing for a while and wanted to put the results online in case anyone else can figure out what I am missing.

Here is the spreadsheet I am running so far:


Explaining it, a row is tied to a player. From the left: Cristelotti, my GK, has 6 stars, no arrows, 141% in Fitness Ability, in the first game tracked ("G1") he lost 22% in Conditioning, his avg, Conditioning loss is 20.5, low at the table the number 18.6 is circled, this is a significant number since its the avg conditioning loss of my players during 59 games tracked (its volume weighted so its a good number), at the top again, 1 is circled which means that my GK in G1 gained 1 ability% (like "1%+ Reflexes"), the number 0.15 is my GKs avg. TEF/Game (in other words, gain in ability per conditioning loss in %, he has a 0.15 TEF, if he looses 20% in conditioning in a game he will on avg gain 0.15x20=3 ability% (like +1 Reflexes, +1 Kicking, +1 Speed)), then TEF L1 and 4 is the TEF he has in practice and the Diff between games and practices.

Cristellotti is a GK, Cesar DL, Mante and Dailly DMC, Cafinha DMR, Zinidizu is a MC, Zap is a striker, Bozaka is a AML.

My conclusions so far:

-The gain in games is very close to what the gain would have been if the player had been trained with Lvl 4 Drills. Ie, about 30% higher than what the gain would have been with lvl 1 drills.

-I see zero correlation between -- individual -- fitness ability and conditioning loss in a game.

-The avg. con.loss for 6 star players is 19.7 and for 7 star players its 18.0. Ie, the less stars a player has, the more conditioning he loses. Does this have to do with the fact that he also gains a lot more ability%? I would believe so. (the long redline splits 6 and 7 star players)

-It still does not seem to be completely random. Like my MC Zinidizu loses a lot less Conditioning per game than my DMC Dailly. The diff is so big that it can't be a coincidence. Like Zinidizu has 4 of 7 games below 16% Con.Loss, my Striker Zapegol has never been below 16% in 8 games. I will keep looking at this.

Any input, ideas or thoughts to take this further would be greatly appreciated!

As always, if anyone want a copy of the excel document, drop me a note!

[LeManiaque]

Just quick thoughts:
I imagine a hard to clarify factor representing opponent tactics and formation impacting condition loss (and form rating). It seems to me that both injuries and cards are closely related to the two opposing formation and tactics combined, which makes easy to believe opponent also matters for condition loss.

An easy answer is condition loss is direct connection to position, easily tested by using your multi position players in 2 different spots over some matches. I doubt it.

Next would be considering simple factors like spaces between player and nearest opposing player. Your MC is probably often opposed to another MC in same position, while DMC may have opposing MC 1 space away. Could be different for DMC vs MC and DMC vs a "stack" of MC+AMC. Arrows might play into that if there are "additive effects", which would represent pressing or attacking teams.

[nikolgiorgos]

as I told you in another thread, i think you 'll find it useful if you also add in your excel/ studies, the rates of the players

[Al Svanberg]

Just quick thoughts:
I imagine a hard to clarify factor representing opponent tactics and formation impacting condition loss (and form rating). It seems to me that both injuries and cards are closely related to the two opposing formation and tactics combined, which makes easy to believe opponent also matters for condition loss.

An easy answer is condition loss is direct connection to position, easily tested by using your multi position players in 2 different spots over some matches. I doubt it.

Next would be considering simple factors like spaces between player and nearest opposing player. Your MC is probably often opposed to another MC in same position, while DMC may have opposing MC 1 space away. Could be different for DMC vs MC and DMC vs a "stack" of MC+AMC. Arrows might play into that if there are "additive effects", which would represent pressing or attacking teams.

Yeah, my thought exactly. I will look more closely at this. If the opponent has an AMC, does my DMC lose less conditioning?

I am not very familiar with the old version, but I was thinking to at least try to look at the play-by-plays made in games. It wouldn't be very precise, but is it possible to see any correlation between like how often a name is mentioned and conditioning loss? I plan to keep this spreadsheet going all season. I miss quite a few games, since I have to dump conditioning status before the game and capture the ability gain after it (the game crashes at times there), but it will give me a bigger sample size. Next step is to look at Al FC. See if there are any trends. Also, it would be very interesting to see numbers from like 2-4 star teams. If noting else, I could start an account just to get a grasp how how the numbers differs. Would also be interesting to see numbers from a team with zero condition bonus.

Is the Fitness Ability just cosmetic? I actually don't think any ability is -- just -- cosmetic. Albeit the impact of an ability quite possibly could be very marginal. Its also perfectly possible that individual abilities are lumped together with abilities of all other players in the same area of the field. Like, if you have 3 MCs and two are 8 stars and one is 1 star, it sometimes don't seem like there is any real difference in how they perform. OTOH, like on my teams if I have a weak DL and ML, it certainly seem like that side performs worse than if I have a strong DL and ML. So I don't think its possible on the basis of this to say that ability don't matter, but it might not matter on an individual basis all that much.

One impact of the Fitness Ability can be that a player plays better with a high fitness ability if he has low conditioning, IE that a player for example with 200% in fitness ability will perform better with 50% conditioning than a player with 50% fitness ability and 50% conditioning. Or, the fitness ability does impact the conditioning loss, but that the impact is so small that it doesn't stick out. Like if the impact of a player having 200% in Fitness compared to a player with 100% in Fitness is that the former will lose like 10% less conditioning -- that deviance could drown among all other factors. Another is of course that a team with 200% avg fitness ability will loose less than a team with 100% avg. fitness ability, but that the ability only matters as an avg of the team, not on an individual basis.

If anyone have some an account out there with 2-4 stars, I would be very grateful if you could just dump a screenshot from the fitness center before and after a game!

[Al Svanberg]

as I told you in another thread, i think you 'll find it useful if you also add in your excel/ studies, the rates of the players

I have!

But after 5 games, I saw no impact whatsoever! I wonder if this could have been changed with some update, since I have seen a lot of talk from a few years back that it has an impact. But from the data I saw it was completely random. I will post some charts later when I have the other laptop. One guy with 10 rating could loose 15% and another with 10 rating could loose 23%.

[LeManiaque]

I'm with you on the play by play, plan to pay more attention to individual plays myself. (There may be interesting conclusions about stories like team with many failed corner attempts with no receiver).

I mentioned on another thread I am wondering about a semi-complex "decision tree" as a main gameplay mechanism. If that were the case, each decision or action could have a specific condition loss associated with the players involved.

For a loose example, play starts from goalie with long kick to ML, ML receives and dribble fast, makes a cross while st tries to sprint to break his mark, st fails to receive and opposing DC wins skill check to clear the ball. In this case, goalie might lose .5 "units", ML .7 and st 1; opposing DC might lose 1, and a base of .2 for players away from the play; add modifier for pressing, mentality. Players who succeeded in their piece of decisions would improve rating, others fall.

[Jeeves]

A complete study on condition loss would be too complicated. The factors that can influence it are many, some of which probably unknown, and you can't really make controlled experiments, because each match is different.

Have you considered the age of your players? Older guys seem to tire out faster, I've noticed. Do you play with the same tactics and formation each time? When using attacking mentality for example, your strikers will tire out faster than your defenders. Marking and pressing style also affect the condition loss.

Like LeManiaque suggested, I've also noticed that the opponent's players are an important factor in this. If for example you use an AMC, it will lose more condition if your opponent uses a DMC, compared to when he is left alone in that spot. Player ratings are also affected by this, so you may be able to find a correlation there, but nothing conclusive, I suppose.

I really like the idea, however, and would like to see this study pass the beta stage! I'll try running some friendlies and isolate tactics and factors that influence the condition, when I find the time.

Good Work!

[Petrov2g]

Nice one mate !

[Al Svanberg]

Thanks guys, I very much appreciates your input!! Fantastic comments!

Niko-
Here are some old data I have with ratings. I tracked rating for about 5 games, but it just seemed all over the place. What do you think? Maybe I should watch it some more.


LeMa- Yeah, it could definitely be like that.

Jeeves- Yeah, my guess is that our conclusion will be something like the avg. is 18.5% loss per game if you have a x% team and that the individual condition loss can variate about +/- 4-5% based on how involved the player is etc.But its still interesting! Every bits and pieces tells you something about the game. Who knows what conclusions can come out of a study like this, all of a sudden you stumble across something you never considered.

Like I for example also bought into the idea that using the conditioning bonus bumper and not press the play was smart because then that conditioning saved could be used in practice, but now I wonder if that is the case. Since there seem to be a direct correlation between conditioning lost and ability gained.

Great idea about age being a factor. Since I also has looked at TEF, I have only included the players I have age 18-21, since I only have comparable training data from them. But I will star to look at vets too. I have some Free Agents I picked up this season to round out the roster that will be interesting to study.

It would be very interesting to see what numbers you get!

[Al Svanberg]

This is how the numbers of gain from the games breaks down in relation to lvl 4 drills:

6 stars players:
Condition lost in games: 354
Ability gained in games: 81
TEF in practice lvl 4 drills (see below): 0.21
Ability gained if same condition loss has been used to practice (354x0.21): 83

7 stars players:
Condition lost in games: 703
Ability gained in games: 100
TEF in practice lvl 4 drills (see below): 0.14
Ability gained if same condition loss has been used to practice (703x0.14): 98

So the gain over a bit of a sample size is basically identical.



So if you have lvl 1 drills, play your players as much as possible. If you have lvl 4 drills, from a gain point of view it does not matter if you play or train them.