View Single Post
  #1  
Unread 05-14-2012, 08:36 PM
K1R4D3L K1R4D3L is offline
The Undying
Interface Author - Click to view interfaces
 
Join Date: Feb 2012
Posts: 34
thoughts on best way to count group kills

I feel like I'm talking to myself a lot on here, but I'll keep going anyways...

I've been giving a lot of thought to ways that I could count group kills in KillTracker, and I'm hoping for some opinions. I haven't done much grouping, so it isn't something I have a great sense for, and all of the options have their pros and cons.

First off, if I do this, I definitely want to count them as a separate metric so that you can choose whether or not to see them, and because the accuracy will be significantly lower than for your own kills. I would also, obviously, only count them if you are in fact in a group and in-combat during the kill. Easy enough...now for the hard part - I have 3 main approaches as I see it:
  1. only include kills by PartyMember names - this will miss any kills by their pets, soldiers and devices, but keep false positives fairly low
  2. include all nearby kills (still have to be in a group and in-combat at the time) - this will clearly eliminate misses altogether, but could include many false positives in busier areas
  3. provide a button to send any new kills while in the group (unsent) in some format to the fellowship/raid (should know which automatically) chat channel, which KT could pick up and count for any group member running the plugin - this would be the most accurate if everyone was using KT consistently, but would miss if anyone in the group making kills wasn't using the plugin or didn't, manually, click the button to send the data before they or anyone wanting their data left

Another possible tweak, especially for option 2, would be to only include kills if you gained XP at the same time - this would further reduce false positives, but would obviously miss any non-XP-supplying (trivial) kills.

I could even do a hybrid of #1 without the XP check and #2 with the XP check, which at this moment would be my choice. I just don't see anyone liking option #3 with the manual click required (but maybe I'm wrong), and this hybrid feels like the best balance between misses and false positives.

Anyone have any thoughts? Anyone?
Reply With Quote