Kill:Death Ratio

i disagree with that. numbers actually do give a lot of information about player's proficiency. take a look at that:
http://www.damnsmallserver.eu.pn/
(very early version and i suck at websites so it may not work properly on some browsers, if at all)
check the following players and try to guess their proficiency, playstyle and teamplay looking only on their stats:
moshermadnes
nirvanajm
krzywousty
julado123br
Simroth

and now their actual proficiency and playstyle

Spoiler: answers

moshermadnes - very skilled builder (hints: large hammer and spike kills percent, sucky KD)
nirvanajm - good knight (mostly sword and bomb kills, average KD)
krzywousty - great knight (very good KD, knows how to use bombs)
julado123br - terribad builder, burden to team (KD, almost no spike kills [and a lot spike teamkills, not listed]
Simroth - good builder (same as moshermadnes)

 

Have you independently verified these conclusions? Like, have you spent a lot of time observing these players in game?

Pardon me if I'm wrong but from you post it seems like you are giving these descriptions purely from the statistics which in no way means your descriptions are true.

For example moshermadnes could be a very poor builder who doesn't do important builder-y things like build shops or ladders but instead runs around just trying to hammer and spike everyone. I have some trouble reading the spreadsheet but I believe it shows that he has a negative score, suggesting a low tendency to build and possibly a high tendency to grief.

You could also look at someone with a good archer KD like 60-20 and think "wow there's a great archer" but it could actually be someone who randomly lobs arrows from the back of his base and disconnects from the serve when enemies get close, which is the kind of behaviour KD tracking tends to promote.

edit: I don't want to come off like I think what you're doing is bad, I think that this stats stuff you guys are coming out with is brilliant and that you are doing really important pioneering. However I'm speaking on a larger scale where I do believe that if improperly done statistics could have a really negative impact on the community.

 

i assure you that ive seen every player of that fifth in action for more than 5h. thats why i chose them.

 

You just should remove these stats. Its just bugged atm and there is a lot of camp-archers . They are just here as stat whores..

 

The way to prevent that is to track if they stay the entire game

 

keep in mind that this is KAG, not soldat. Averange KD on my server is 0.6 (partially due to map designs, most of them has bottomless pits). Also, there is a bug that if someone throws a bomb and dies the bomb becomes "neutral" - can kill both team-mates and enemies and is counts as selfkill.

because of the nature of KAG i was wondering if selfkills should be accounted when calculating KD and points.
more info here: https://forum.kag2d.com/threads/the-original-damn-small-server.7146/page-2#post-137706
i think that depends on if the server uses limited or unlimited units.

 

I do agree with this.. you should make a POL on this saying if you want K/DR or not..... I personaly would prefer to have this because its better for clans.