The problem with your theory is that you need to calculate drop chance for every player in party and come up with an average since the same 40 people don’t raid together.
Someone would have figured out that the best chance to get an item you want is to raid with 39 people who killed the boss a few times without getting the item so that the algorithm favors you (or just wait until a boss is killed n number of times and kill him for almost guaranteed loot if personal player stats aren’t tracked)
It would be easily exploitable, like we exploited crit chance in warcraft 3 where the system tries to even out the number of procs by starting with a very low chance on the first hit and than exponentially increasing to prevent no crit streaks.
Your idea creates more problems than it solves, and it makes no sense to implement it.
This is quite interesting, but as some have pointed out the drops that we see (that seem non-random ie same shabby item over and over, or the good ones ie binding not dropping) are not impossible at all just assuming pure randomness, sometimes they are even fairly likely like the Garr example - 10% chanc eof not getting a Garr binding in 1 year.
Personally we’ve got 3 Garr 0 Geddon, also we had 1 OG in ~7 months and now 4-5 OG’s in 2 months. It doesn’t feel random, but that’s prob more to do with our flawed intuition of what random truly means.
Later versions of WoW have after all slowly introduced different versions of “Bad Luck Protection”, as the true randomness isn’t that great an experience
That was an interesting read, I wonder if there’s a collection about such announcements by Blizzard somewhere? Even videos or interviews could contain some info