Einzelnen Beitrag anzeigen
  #61  
Alt 14.04.2020, 19:07
Tibono Tibono ist offline
Resurrection
 
Registriert seit: 22.05.2018
Ort: Frankreich
Alter: 63
Land:
Beiträge: 599
Abgegebene Danke: 3.635
Erhielt 1.406 Danke für 477 Beiträge
Aktivitäten Langlebigkeit
6/20 7/20
Heute Beiträge
0/3 ssssss599
AW: King Besitzer als freiwillige Tester gesucht

 Zitat von germangonzo Beitrag anzeigen
Hi Eric,
where did this formula (171,4) comes from?
When I use the formula I calculated I get e.g.


npm 125 250 500 1.000 2.000 4.000 8.000 16.000 1.000.000
ELO 1.113 1.215 1.318 1.420 1.522 1.625 1.727 1.829 2.440

Finetuning with new results and better fitting in the low-ELO-area are possible,

Best regards,
Frank
Hi Franck,

I agree the table you show here is the correct output of the formula you calculated.
You can compare your table values with the orange spots (& orange line drawn) in my latest graph; this is it, we are on the same page.

But my concern is, this orange line (your table, so), does not hit any close enough the blue squares, apart from the 125 nodes spot (7 on the x-axis) and 500 nodes spot (9 on the x-axis). The orange line diverges a lot from the others and does not fit the trend.

I trust you calculated it correctly, but most probably you used other points as input for interpolation. The input sample I used is page 5 (results so far from games shared here):
Code:
Nodes     	125	250	500	1000	2000	4000	8000	16000
Elo     	956		1339	1618		1951	1932	2203
Excel provided automatically the blue line (feature: add a trend line, choice linear) - It obviously fits better than the orange one.
So, to answer your question, the so-called "171,4" formula is the one Excel displays for the blue line.
Kind regards,
Eric
Mit Zitat antworten