Username: Passwort:
Neue User - Registrierung
Moderator: SueQ , coan.net 
 Backgammon

Backgammon and variants.

Backgammon Links


Liste der Diskussionsforen
Modus: Jeder kann schreiben
In Postings suchen:  

27. Dezember 2008, 02:01:44
AlliumCepa 
Thema: Re:
grenv: Quote: "TC: It's the same thing using different words."
No offence, man, but it makes a huge difference by adding the word "average". The way you stated it the first time ("# of moves per day"), it sounds like player's moves are calculated daily and reset at server's midnight. Maybe your idea indeed was to propose the "average daily moves during the last 30 consecutive days", but you didn't make it clear. All I did was to bringten up your idea, because the "average" thing wasn't too ovbius.

Furthermore, to make the ratio more precise, there is one thing (that I can think of) to avoid. Here is an example: We have a player who had 100 running games during the last 30 days and he makes 100 moves per day. So as average, the guy has ratio=1. Now imagine on day 31st a huge tournament kicks in and suddanly there are 100 more games to play, totaling to 200. During our guy sleeps (and hasn't made any moves) his ratio will drop to 0.5, becouse "now" he has twice as much games as before. So my suggestion is: we don't make the formula "in real time", but rather a "moving average" that deals with 30 complete days in order to smoothen such peaks.

In other words, here is a sample "Activity Ratio" formula: AR=[Σi=1..30(Mi/Gi)]/30,
where:
i - day number, runs from 1 to 30;
Mi - moves made during day i. Calculated at server's midnight;
Gi - number of running games at the end of day i.

Of course, I don't pretend to be thorough.

Datum und Zeit
Freunde Online
Abonnierte Foren
Vereine
Tip des Tages
Copyright © 2002 - 2024 Filip Rachunek, all rights reserved.
Zurück nach oben