Re: [pbmserv-dev] Solitaire game scoring

[Ragnar Wikman IB <rwikman@ra.abo.fi>]

On Thu, 15 Apr 2004, Cameron Browne wrote:

> Date: Thu, 15 Apr 2004 15:41:15 +1000
> From: Cameron Browne <cameronb@optusnet.com.au>
> Reply-To: pbmserv-dev@gamerz.net
> To: pbmserv-dev@gamerz.net
> Subject: [pbmserv-dev] Solitaire game scoring
>
> I'm wondering how to best handle scoring in solitaire games.
>
> Robots implements a high score table in which only the highest score for
> each player is shown. This is a bit dull since a player's standing will not
> change very often, and almost never if they achieve a freak high score.
>
> Would it be better to use the player's average score over all games
> instead? The total number of games played is given by Wins+Losses+Ties, so
> after a game the player's new standing could be calculted as follows:
>
>                     (old_standing * total_games) + new_score
>    new_standing  =  ----------------------------------------
>                                 total_games + 1
>
> This is not a true average, just an approximation using the available
> information. The more games a player finishes, the less impact each new
> score would have.

Hmm, people can have freak low scores as well - would it be possible to
use, say, 'average of top ten scores' or would that require too much new
programming/data storing?

> Cameron

Ragnar

* Ragnar Wikman, Abo Akademi University  * "Every soul is  *
* Lecturer in Computer Science           *  worth a song." *
* mailto:rwikman@abo.fi                  *  -- Veikko Lavi *