Quote:
Originally Posted by mrproudbeard
[You must be logged in to view images. Log in or Register.]
Heh, I only had math reasons why that chart was bogus. Rad to see non-math reasons why it's also garbage!
dasgoot... here's some tips to make your data better. This is not a comprehensive list.
Use a time range instead of a hard "300 kills" because that's way more apples-to-apples than what you did.
Graph all guilds in a single graph and use a grouped bar chart. You can set some lower bounds to make this not super crazy, like only guilds that have > 20 kills and/or deaths in the time range, or whatever lower bound makes sense to be able to actually compare the data without the chart getting too dense. This would allow real data comparison to happen.
Since there are allegations of dropping tag to avoid leaderboard hits or something... you probably need a bit of data scrubbing to make sure people who do that get grouped into the correct guild for every kill/death instance. A fun side statistic would be a count of occurrences where someone dropped tag for a death and then was back in the guild later. See which guilds are most guilty of this because that seems like good FQ material to me!
That all sounds like a lot of work though, so sticking with this totally worthless narrative is probably more time effective for you. Math abuse makes me sad.
|
lol
why would a time range be more "apples to apples"
it was a direct comparison between holocaust's and azrael's last 300 kills allocated by guild.
think u might be a retard.