Subject: oSokker

Sorry, probably answered a million times, but how do I find what the calculations are for the skills sum at each of the positions.

Thanks.

Where i can find blue skin option when i activated osokker?

On hungary National Team main page:)

Sum coeficients:
i hope its clear, because i am to lazy to rewrite it :-))

aSumSkills = {
'all' : {
'tactical_discipline' : 0,
'form' : 0,
'stamina' : 1,
'pace' : 1,
'technique' : 1,
'passing' : 1,
'keeper' : 1,
'defender' : 1,
'playmaker' : 1,
'striker' : 1,
},
'allminstam' : {
'tactical_discipline' : 0,
'form' : 0,
'stamina' : 0,
'pace' : 1,
'technique' : 1,
'passing' : 1,
'keeper' : 1,
'defender' : 1,
'playmaker' : 1,
'striker' : 1,
},
'kee' : {
'tactical_discipline' : 0,
'form' : 0,
'stamina' : 0,
'pace' : 1.5,
'technique' : 0,
'passing' : 1,
'keeper' : 2,
'defender' : 0,
'playmaker' : 0,
'striker' : 0,
},
'def' : {
'tactical_discipline' : 0,
'form' : 0,
'stamina' : 0.5,
'pace' : 1.5,
'technique' : 1,
'passing' : 1,
'keeper' : 0,
'defender' : 1.5,
'playmaker' : 1,
'striker' : 0,
},
'mid' : {
'tactical_discipline' : 0,
'form' : 0,
'stamina' : 0.25,
'pace' : 1,
'technique' : 1,
'passing' : 1.5,
'keeper' : 0,
'defender' : 1,
'playmaker' : 1.5,
'striker' : 0.5,
},
'wing' : {
'tactical_discipline' : 0,
'form' : 0,
'stamina' : 0,
'pace' : 1,
'technique' : 1,
'passing' : 1,
'keeper' : 0,
'defender' : 1,
'playmaker' : 1,
'striker' : 0,
},
'att' : {
'tactical_discipline' : 0,
'form' : 0,
'stamina' : 0,
'pace' : 1,
'technique' : 1,
'passing' : 0,
'keeper' : 0,
'defender' : 0,
'playmaker' : 0,
'striker' : 1,
},

Is there a way for us to change those numbers somewhere?
I would chose some others when i see this. :)

if you use Opera then yes :-))))

:P

too much work to change :P (tried it once but go back to FF soon. On my tablet i use opera, but it has no add on functions there)

You can change it in Opera in file: oSokker_cis.js

can i use oSokker on Opera Mobile ?

I didnt try that. But if you have Opera standard, then it should be possible. You just have to find the User JS setting.

As far as i know, no it is not possible.

At the juniors page, the "est. level" is how percent correct?

Its linear regression = more weeks, more correct
after 10 weeks should be pretty correct

"after 10 weeks should be pretty correct "

I am sorry to disagree. Lineair regression needs a lot more weeks to be correct.

a difference of 1 between coach level and real level is a normal difference that can happen a lot.
5 5 5 5 5 5 5 5 5 6 and 4 4 4 5 5 5 6 6 6 6 can have perfectly the same real numbers.
Another problem with lineair regression is a high or low startingnumbers will ruin the estimation.

from around 16-20 weeks lineair regression works ok. But there are better ways to find out talent like :
JR Plus

or the excelsheet i made: talent prediction sheet (but for my sheet you have to put in all data manualy every time) One of the pro's of my sheet is that it works with a range. With guessed numbers like in sokker you can not point out easily one number and say that is his talent. There is a range where the talent propably lies.

example:
I have right now this player:
7 7 8 7 8 8 7 8 9 8 8 8

Lineair regression is 9,86

JR Plus makes it 8.17

my sheet makes it:
Gill Landers 7 7 8 7 8 8 7 8 9 8 8 8
talent - (3,75) - 7 7 7 7 8 8 8 8 9 9 9 9
talent - (8) - 7 7 7 7 7 7 7 7 8 8 8 8

the problem of lineair regression in it self is that it does not work with sokkerfacts, like there is no start at 7.5 so the first 8 of the coach can not be correct, that has to be a 7.
Both sheet estimations have the same fault distance between coachvalue and guessed value, both distance is 5

There is a smaller distance possible:
talent - (6,5) - 7 7 7 7 7 7 7 8 8 8 8 8
talent - (7) - 7 7 7 7 7 7 7 8 8 8 8 8

distance is 4. But it is risky to use this because of the small amount of numbers. there is no good reason to make the range as small as possible, because it will have a higher mistake range too :)

----
I do not say LR is not good to use, i only wanted to explain a little bit that people should look through the number they see in oSokker :)
(edited)
(edited)

I have another question.

When sorting in sokker it works from low to high. Is it possible to reverse that automaticly?

Well. Imho linear regression deals pretty well with coaches imprecision. Some kind of exponencial or better logarithmic regression would be better.

example:
I have right now this player:
7 7 8 7 8 8 7 8 9 8 8 8

Lineair regression is 9,86

JR Plus makes it 6.78

How do you know which estimation is better? that 9 could be statistical discrepancy.

my sheet makes it:
Gill Landers 7 7 8 7 8 8 7 8 9 8 8 8
talent - (3,75) - 7 7 7 7 8 8 8 8 9 9 9 9
talent - (8) - 7 7 7 7 7 7 7 7 8 8 8 8

well you have a range between 3.75 (actually should be 4) and 8, but you have no idea what the talent is. And it still could be more than 8, so you still do not have the correct MAX

Linear regression calculates what could be the next actuall skill next week. And your excell sheet does not take it in the equation. This is why your max is 8.