Forums

[Updated on: Mon, 16 April 2012 19:37]

Good graph Francois, thx for the efforts you put in it !

Beginners after 100 games should also be super low-ranked, but remain around 1350 or 1400 points.....

This could be a starting score then without a provisional ranking.

A special Hug and kiss to An-Team, who should have stayed retired from TTR like he told the community in a very loooooooooooong pm some years ago !

[Updated on: Wed, 18 April 2012 22:32]

On a un deuxième ..o...c qui chasse les débutant d'ipad. Et certains qui sont au dessus de 1770 chassent des brêles

Question :

Who between jimmy and anteam is the best hunter of ipad ?

[Updated on: Sun, 22 April 2012 08:35]

Sysyphus = hardliner ^ 2

I don't respect Jimmijazedzed very much, I never did, I simple don't know him and his ranking is much, much too high.

I do respect An-Team. He is not a top player but a good one. His ranking was over 1800 and there he was clearly overrated. But I browsed his history an I think he was a true 1700+ easily. Less than every 5th of his games were against beginners.

As for myself, I don't open games and I step into every game 1300+. That's my way for years.
I don't step into games called beginners only, and since I'm aware of the situation I don't step into games of players with exactly 1500. But there is no chance to recognize a new player. When a 1480 opens a game, how can I know if this is a (almost-)beginner?
That I'm getting the start is not of much help to judge the situation. If a player is giving start all the time, it is correct to play against him, because if he manages 1400 without start than he is a real 1470, so the points I get from winning against him are correct.

The only problem is the starting ELO of 1500 and not dropping much in the first games...

[Updated on: Thu, 26 April 2012 11:08]

[Updated on: Wed, 16 May 2012 07:28]

[Updated on: Wed, 16 May 2012 07:29]

Thx for the hint, that !1 means, this player is a newbie . Didn't know that, and reading this, I really get fear, that I played one of them. There really were some games, that I won easily, perhaps she/he was a newbie too. Her/his name was dea1...



Carsten

[Updated on: Tue, 01 May 2012 20:54]

DoW could re-design and recalculate the whole rankings easily!
It's a simple database with 30.000.000 or so entries.

Just delete the games of the known cheaters beforehand

And a new #1 history would be nice
e.g.
20041201-20050131 Angel (max. 1850)
20050201-20050831 kid (max. 1965)
20050901-20051031 pilke (max. 2010)
...

Ideas for the re-design:
-) 1500 has to be the exact overall average (forever).
-) Players that don't play should continually drop in rankings
-) Players that play should equally rise
===> those two things can happen once a day - just drop non-players by 0.1 and raise every player with played games the same amount (e.g. 3000 players didn't play, so they drop by 0.1 each; 10.000 games where played, that means ~25.000 players*games, so a player that played 8 games get 300/25000*8=0.096 pts.
-) Maybe the provisional rankings formula should be "normal" for the beginner, but much lower for the experienced player. The new delta can be balanced by raising the gain of the playing players as described above.

[Updated on: Thu, 03 May 2012 11:14]

MaximoR - Ginglet schrieb am Thu, 17 May 2012 23:41

I take manualy the 600 pages of scores

wow
There should be a possibility to write a script that reads the scores automatically!

[Updated on: Fri, 18 May 2012 13:39]