Guide: How to make a budget on the cost to develop a player?

[Al Svanberg]

New numbers after Al's game:


The Conditional Loss from a game -- indeed -- seem very random in relation to the individual fitness ability (and the rating too, Niko maybe this is something they have updated?).

Its a small sample size, but last game the avg was 18.75 now its 20. Will be interesting to see where we end up as this even outs with a bigger sample size. Does the fitness ability, or ability of the team, has any input on this number? Will a 8 star player loose as much conditioning as a 1 star player? The level of opponents matter? Position? How much effect does that 5% booster have?

The ability gain from the game seem to be on par with lvl 4 drills and Niko's numbers still.

[Al Svanberg]

First numbers from this season, will start doing some more tests this year to get a bigger sample size. Just need to remember to dump the condition status before games...



Seemingly zero correlation between Fitness Ability and Condition Loss in the game. Also, I had a lower avg. Con.Loss this games than the two other games I tracked, despite a 20% lower fitness ability avg.

The TEF in games keeps being closer to lvl 4 drills.

This is OT for this thread, will probably start a new thread on this topic when I get a bit more time.

[Al Svanberg]

I think I more or less can confirm that the gain/TEF in games is exactly the same as from lvl 4 drills.

Based on tracking individual players' gains in 34 games, the players gained 108 ability%. Had the same amount of Condition% as was lost in the games been used in practices, the gain would have been 108.1 ability%.

So on to the topic of how much condition% players looses in games, after tracking each player in about 5 games (some players have been rested a few games), I still have no clue what factors impact how much condition a player lose in games. Basically two explanations remain, (1) the condition loss is completely random in relation to the individual player (ie not in relation to bonuses or orders), (2) the condition loss is impacted by factors I have yet to understand.

The randomness of the numbers suggest alt. (1). But I am starting to see more and more trends that suggest something else. Lets look at this table:



The avg. Con.Loss of the players north of the red line -- ie 6 star players -- is 20.53%. The avg. Con.Loss of the players south of the red line -- ie 7 star players -- is 17,6%. This difference is quite significant. Like of the 13 games played by 6 star players, in 9 of them players lost more than 20% in condition. Of the games played by 7 star players, in only 2 of them did a player lose more than 20%. That diff to too big to be a coincidence, I think... Given that the sample size is not that insignificant anymore, this could explain a few things. Could the condition loss in a game be related to the ability gained? This is not the case on a game by game basis.Like a player can lose 22% con. in one game and gain 1 ability%, and another can lose 15% and gain 4 ability%. But this can be attributed to the "container function". We only see "gain" when it results in a container being filled up so that a full ability% is gained.

This also possess a few other questions. If you run condition bonuses that lowers the condition loss -- is the gain also lowered? The same with hard pressure, do you gain more if you play hard pressure and loose more conditioning? Some opts to always use the con bonus booster to save conditioning for development -- but I am not so sure that this helps. If the gain is lowered, nothing is gained by doing so.

What do you guys think?

[LeManiaque]

I really wonder the same about condition loss in match, especially if condition loss ties to skill gain. I don't think the losses are random, as other seemingly random aspects of the game have been stated not to be (like scout recommendations most recently). Pressing and arrows seem like logical, simple contributors.

We know something is calculating the field position or possession during each match (referring to the bar below the field that sways back and forth). So there could be a calculation representing km coverage during a match.

The game could be played "step wise" where each possession has an action tree depending on aspects of the overall team (for example 1amc2st could present different branches or attempt-able plays than aml+amr+st; maybe even allowing certain plays only if the quality of that area of the field is greater than X [explaining why some teams seem to attack in different combinations, especially with a Q difference]). In this way, players condition could drop only if they are involved in one of these branches plays.

[Al Svanberg]

Yeah, I will keep building up data.

There are a few numbers that stick out, another one is the diff between games. Some games, like G1, obviously seem to have had a lesser impact on the condition loss than like G5.

If anyone is starting a new account or has some old forgotten account that hasn't been used in a while, it would be interesting to see info on how the condition loss looks in those games.

[LeManiaque]

Somehow I missed post #10, great stuff

[ibangali]

Attachment 73974

I try to bring your updated info in here, can you please do it

mine showing as attachment not the real data, thanks