Quote:
Originally Posted by falkao
indeed, this would be another approach to deal with the data but is less informative. It would be ideal to test covariates.
|
I think that on the contrary, looking at how different variables such as avatar characteristics or kill time affect loot would be most interesting, if you are trying to understand the loot system.
Basically,
EU is a
zero-sum game. This means that your gain is someone else's loss, and vice versa. Since
MA "siphons away" resources through decay, average player will necessarily loose, on average. Your (very informative, by the way) findings show more specifically how much you are expected to loose by killing a given mob.
It doesn't mean that every player will loose, however, some will certainly win. Running probit/logit on a bunch of characteristics could then tell you what factors, if any, make you more or less likely to be a winner, rather than a looser. The main problem problem are data - it will be hard to collect large and comprehensive sample. Another problem is that any such findings will be useless as soon as they are made public, for obvious reasons.
Lastly, you've done a great job so far. Kudos for taking your time and doing first (to my knowledge) serious analysis of the loot system.