Subject: Junior Talent Prediction Tool for non plus users (based on sokke

thx

updated to version 1.0.0.9
- screenshot can be taken


(edited)

Hello.

I have a little problem. I see, that under 3 weeks, the program doesn't show me the players... but I looked at my SV profile, and 1 player is still missing. Is there another criteria beside this "at least 3 weeks in school" thing?

This player is missing: His history in the school is: 4 5 3 3 3 3 4 4 4. As you can see, he is in my school more than 3 weeks, and the program doesn't show me his graph.

thank you.
(edited)

hello,
please take a look in your sv xml folder for the newest file of type juniors_82264_******.xml
open this file and look if you can find your lost player here. if he is not in this file he will not shown.
junior regression tool starts parsing the newest junior.xml file because there must list the actual player in school. then it parsing the second newest file and collecting the data and so on.
the sokkerviewer has its own database and the xml files are only used for downloading the data.
i see you have plus. in two or three weeks comes a new version of the tool were you can download you data direct from sokker.org and you have no missing data.
best regards

Hello.

I have checked my xml file, as you've said and I can see there my missing player. So I don't know what can be the problem. :) I have refreshed my SV also, so the xml file is really fresh.
No, I don't have Plus in Sokker.
The new version will be mentioned here, so we can dowload it from this topic?
I hope it will be good, and none of my "students" will be missing from the regression table. :)

Hzoli

Junior Tool dont work any more ?

nice job. I've also used regression to predict talent:)) But, Im not good at programming

Could you tell me how did you make it? what I must learn to write program like this?

(edited)

my method was ..


subplot(3,2,1);
x = [1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31];
y = [6 5 7 6 6 7 7 9 9 8 9 9 8 9 9 10 10 10 10 10 11 12 12 12 12 13 13 13 13 14 15];
x=x'; y=y';
A=[x ones(size(x))];
c=(A'*A)\(A'*y);
ya=c(1)*x+c(2);
plot(x,y,'o',x,ya,'-');
grid on;
xlabel('x');
ylabel('y=F(x)');
ylabel('level');
xlabel('week');
title('Dziedziel');

subplot(3,2,2);
x = [1 2 3 4 5 6 7 8 9 10 11 12 13 14];
y = [5 5 5 5 6 6 7 7 7 7 8 8 8 8];
x=x'; y=y';
A=[x ones(size(x))];
c=(A'*A)\(A'*y);
ya=c(1)*x+c(2);
plot(x,y,'o',x,ya,'-');
grid on;
xlabel('x');
ylabel('y=F(x)');
ylabel('level');
xlabel('week');
title('Marecki');



:D


[URL=http://imageshack.us/photo/my-images/545/dsdadsada.png/][/URL]



I had to copy y (levels) from skviewer every time

so that was time-consuming.:)


(edited)

http://i.imgur.com/FXbyT.png

Ryszard Jarzebski

7 weeks at school
average pop ~2.55

CorrC equals 0.89


How is it possible?

This program uses plusdata

It seems to me that this team has only 7 weeks plussubscription. That makes the program unreliable for youthllayers that are longer then 7 weeks in youth.

Besides that, this program is just a tool. The talent that is predicted isjust that, a prediction. With (lets say) less then 15 points of data it is often wrong.
(edited)

Plusdata?

:O
It's been a long time since I last used it (plus)
not less than 5 years



Besides that, this program is just a tool. The talent that is predicted isjust that, a prediction. With (lets say) less then 15 points of data it is often wrong.

If we have large amount of data its very very accurate:)

I've just checked it with my program (lol) and it seems that there're no mistakes with regression or sth

I have also 2.8 talent (jarzebski)


(edited)

Oops sorry i was mistaken, i tought of an other youth estimationtool.

This tool uses sokkerviewerdata. If you replace plus with sokkerviewer in my previous post it is correct ;) it seems you did use sokkerviewer for only 7 weeks or so :)

If we have large amount of data its very very accurate:)

yup :) but as long as you see talent below 3 it is not accurate :D
(edited)

:)

Thank you, you will help me a lot.

---

188cm, you are right he is big!
;)