forums.marpirc.net

Ataxx scoring has been discussed formerly at
viewtopic.php?t=4444

But we've planned by mail with Chad to change the scoring for ataxx & clones, and to harmonize it with gfire2's scoring (which is the same game).

Several things are not good in actual ataxx scoring and should be improved.
1- sometimes you can win taking all opponent's blobs ;
for example 17-0 should be counted (49-nb blocks)-0
[note that gfire2 is counting like that : look my own recent input, I was given a 49-0]
2- scoring with nb blobs at the end is bad, blobs difference is better ;
example=24-17 (8 blocks) has to be rewarded better than 25-24 (no blocks)
3-with actual scoring, a defeat could be more rewarded than a win !!!
example=24-25 better than 22-21 (6 blocks) : nonsense

So we should just count :
1 win = 100 + (blobs difference)
1 tie = 0 (it's possible, Timmykins has a tie in one of his inputs)
1 lost = blobs difference (negative)
[still keeping 1 credit per game]

But before making all edits and changes, I post there to know if someone is against these changes

on a historical note, i made the initial ataxx scoring because i thought the difficulty #of the opponoent 1700 2200 etc reflected the difficulty of the game. it does get harder but that's no reason to add a bunch of math to the equation, the othello/gfire scoring systems work well so they should work here to.

There is an argument to make the 17-0 game worth less than a 49-0 game, since they are definitley different games; if you win with 49 pieces on the board instead of 17 or a number less than 49, you've allowed a lot more risk. but i won't make that argument now :)

Ok : 1 week and nobody has something against new scoring ;
I do the changes