Subpage under development, new version coming soon!
Asunto: oSokker
Somehow i like Venezuela, so i wont react on your question. The others already did.
There is a small mistake in this option:
When I use it I get the first 2 numbers of weight where height should be.
When I use it I get the first 2 numbers of weight where height should be.
Something is wrong with the talent calculation. +3 in 13 weeks, oSokker gives talent 11.10 :/
Talent prediction is based on linear regression. You can check the talent yourself. What were the skills of the player in those 13 weeks?
Seems that i am more busy when i am unemployed as when i worked in a company ;-).
Maybe i had free time in work then :D
Maybe i had free time in work then :D
You can check the talent yourself.
I did check it myself and it's +3 in 13 weeks. And after 13 weeks it should start to give a somehow steady prediction, but 11.10 is way of 4.33 ;)
I did check it myself and it's +3 in 13 weeks. And after 13 weeks it should start to give a somehow steady prediction, but 11.10 is way of 4.33 ;)
"And after 13 weeks it should start to give a somehow steady prediction"
No, that depends on the numbers. And below 15 weeks all data is (in most cases) not very reliable. And i saw cases that needs 22 weeks to get to a good indication (not only with LR but also with other methods)
if you say it is 4.33 based on 3 up in 13 weeks, that is almost as not correct as 11.10. You can not simply divide 13 by 3. You do not even know how good your numbers are.
No, that depends on the numbers. And below 15 weeks all data is (in most cases) not very reliable. And i saw cases that needs 22 weeks to get to a good indication (not only with LR but also with other methods)
if you say it is 4.33 based on 3 up in 13 weeks, that is almost as not correct as 11.10. You can not simply divide 13 by 3. You do not even know how good your numbers are.
I looked at it again and oSokker could be right if week 0 is low prediction and week 1 high prediction.
With 'you can check yourself' I meant you can perform the linear regression yourself. It depends on the actual data not just the first and last value. Your method is an easier and therefore rougher estimate.