Loading...
  OR  Zero-K Name:    Password:   

Post edit history

drop elo by wining game?

To display differences between versions, select one or more edits in the list using checkboxes and click "diff selected"
Post edit history
Date Editor Before After
7/7/2018 9:31:19 AMDErankBrackman before revert after revert
7/7/2018 9:25:12 AMDErankBrackman before revert after revert
Before After
1 Was it really because of you winning? Or because of other ppl loosing independently that you played against previously? The only mechanism I can think of atm to explain loosing whr specifically by winning a game is that you played against the loosing players of that game just recently and then the system thinks that they were also worse in those recent games so that those games are rated worse for you. But I would think that this effect would not be enough to cause a "negative" win in most cases. 1 Was it really because of you winning? Or because of other ppl loosing independently that you played against previously? The only mechanism I can think of atm to explain loosing whr specifically by winning a game is that you played against the loosing players of that game just recently and then the system thinks that they were also worse in those recent games so that those games are rated worse for you. But I would think that this effect would not be enough to cause a "negative" win in most cases.
2 \n 2 \n
3 Or was it maybe not your whr mean but your whr visible rating that changed? Maybe this is caused by a whr deviation increase because something unexpected happened? I don't even know if whr has such a mechanism like TrueSkill to increase deviation on unexpected game results. Even if there were seemingly negative wins in whr I would think there would be much more with TrueSkill (without beeing actually negative because they only refer to visible rating, not actual rating). 3 Or was it maybe not your whr mean but your whr visible rating that changed? Maybe this is caused by a whr deviation increase because something unexpected happened? I don't even know if whr has such a mechanism like TrueSkill to increase deviation on unexpected game results. Even if there were seemingly negative wins in whr I would think there would be much more with TrueSkill (without beeing actually negative because they only refer to visible rating, not actual rating).
4 \n
5 Furthermore, I think that the current whr standard deviation cutoff to translate between real whr and visible whr should be increased from 20 to 25.