Subject: »[info]Dev Diary #82: Automatic release of NT-players from

I think low form is a way to force a club selling a player.

I'll explain it better in more detail:
My first 11 is all i am talking about! and not the keeper i bought in, my 1st choice keeper, and this has been going on weeks since start of season, my keeper was injured on NT duty the day before this season started , wonderful! this was the season i paid untold for to win my first premiership.

My main CM form dropped so much to very good so i had to sell, now my other 2 same problem one dropped to adqeuete selling that one, ok some players don't wanna play for me anymore.

*I went 1-0 up v Quarter Final champions leage opponent 1-1 till 81st min, beat Thamesmead 2-0 drew Bedzin 1-1
then lost to those same 2 sides Bedzin 5-0 and couldn't even beat bramford nothing changed but random form and my keeper

I got a divine keeper in but they stand in corner until tactical discipline better, season was over before it started i can handle that, but it shouldn't be , asif a newly bought keeper stands in the corner and forgets his whole career before him , chuckles , it's a quirky part of the game i can suffer, it's rare to get a keeper injury

You can't leave injuries same length when your playing double amount of games in a week , it's nonsensical

so injuries have to be 50% less as we are playing 50% more games , is that not 100% logical
(edited)

As a NT manager what i saw in a lot of years that the best trigger is when a player changes a team.
It looks like that some players just dont want to play for some teams.
It would be possible to check this by analysing the low form players for at least 2 seasons and how they change in one season after they are sold.
It is the same backwards, not one example, when i bought ~divine form player and his form went down very soon and stayed here.
(edited)

Tomori went from magical to very good in my team since 18yo now 27yo
my keeper has been with me over a season
all but 2 of my defenders were and ones that are not are in divine form
all my other players have been with my 2 seasons or more

The form is at least a change of +3/-3 every times. With a small set of data history it should easy to show (using simple math probability) that its purely a random roll of the dice. 3 chances for 4 to loose form, 1 chance for 7 to not change (until the player form is not 3 or 16 :-), and then and then...

i've had many -4 this season no +4 mostly just minus bar one player Hecquett, this has never happened before

yep, my bad. I wish it was +3 -3 but i miss something (or it change this season ?) I just check and i see a +5 form :-) ... well, so i dont know which is the range +6-6 ? no limit ? +18 -18 ? :-)

I think it's because form changes twice a week, while graph is updated only once. It may show even -6 decrease, which is misleading. In fact it was two consecutive -3 drops.

oh right its the explanation !

This new range of +6/-6 a week is not bad news. +3/-3 was ok for those players already high, but it wasn't for those who was low.

Again, im a not skilled in probability and statistics, but it seem not too hard to confirm that's form is juts a random dice of +3-3 each times.

I'll check form changes for all sokker players on Thursday, compared to today :) I'll divide it into two groups: those who played on Wednesday, and those who didn't, also I will make some comparison based on current form (to see if players with lower form are more likely to lose even more form and if players with high form are more likely to gain even more form).

For now i can say it's not just "a random dice roll", for example players who don't play matches have higher chance of losing form over longer time:

Form of players who played at least one match in the last ~month:


Form of players who played no matches in the last ~month:


(Of course there might be some correlation "player has low form so I don't play him", but I think most of these players who don't play at all come from bot teams, so they are pretty good "random population")

Data from my Players database

If it is a purely random draw we should see an equal distribution according to form number : and we observe this equal repartition (table 1. except a strange and slight inclination on 0).

Maybe there s a extra factor for those who play or not (table 2). But the population on table 2 is much lower than table 1, we cannot rule out the idea that those who are low in form do not play because of this. (bot teams players play so...).

Other tables, like the population repartition of +6/-6 form should confirm an equal repartition (except on poles, if a player has 0 he couldn't loose -3... ) and an median change near 0.
(edited)

Graphs are much easier to interpret than numbers :)

This is the form distribution of the players from all the teams in Spain a few seasons ago. It should be the same for the rest of countries:


And this is the form distribution of all the active teams in SA last week:

i clearly didn't understand your second graph (?) the distribution increase when form is higher ? It is quite the contrary to the overall data picture of Mikoos ^^

What's SA ? It look like a form distribution not only for 'active' teams, but competitive' ones, since those teams tends to keep higher player form and sell lower form its easy to change the random factor in that case :-)
(edited)

I have just checked 5 BOTs from my league and i found all of them have a majority of players with a low form, regardless of whether they played or not.

5 teams, 110 players:
- 24 players with form 0-1
- 2 players with form 16-17
- player with highest form [17] (within all 110 players) has not played even a minute.
- 11 out of 24 players with form 0-1, are playing regularly, so it is almost 50/.50 in my league.

Of course this is a low probe, but always shows something.
(edited)

ok maybe the players who don’t play loose form, but globally for those who play or not it seems that s globally a random factor happen (aacording to the equal repartition table 1)

Maybe you could compare the population of players who win or loose a match…

Most likely the players in bot teams have overall just low form (maybe because they don't have a coach assigned?), even if they play matches? So it looks even in my data, but in bot teams it's much worse and in active teams it's much better. I don't have data information if a player is in a bot team now, but I'll check it later today. :)

For now I can check player's form in my transfer database (all players sold after December 27th, aged > 23 to not get 9 form fresh juniors spike), and indeed here it looks a bit more like in Terrion's graph: