Sorry, I'm still not following this.

FrostBurn said:

One player = 1 : 2406568 (One player being /p1 and /p2)

Four players = 1 : 1052873 (Four players being /p7 and /p8)

ATMA's probability listed above is the probability of a Razorswitch dropping when you kill a Hell Bovine at /p1 and /p8. From the above, the probability of a Razorswitch dropping at /p8 is much higher than at /p1.

Delreich said:

Those are probability per kill, so while increasing the chance of getting a drop will increase the overall chance of getting a certain item, the chance per drop is still the same.

But the probability per kill, as listed by ATMA, is higher at /p8 compared to /p1. On the other hand, you say that the chance per drop is the same. So probability per kill is not the same as chance per drop? If not, then what is chance per drop?

Delreich said:

If you run cows on /p1 until 100 items have dropped, you will have the same probability of getting a Razorswitch as if you run on /p7 til you get 100 items. You will have to kill more cows on /p1 to get those 100 itmes though.

But the probability on /p8 is higher than on /p1, as given by ATMA. Which is why I don't get it.

Skinhead On The MBTA said:

Yeah, but if there's a higher chance of any item dropping (as there is if there are more players in the game), then surely there is a higher chance of the unique version dropping, is there not? That'll be why ATMA gives better odds of Razorswitch dropping for more players in the game.

That is how I read it from looking at ATMA's results. Firstly, more items will drop (i.e. the 'no drop' chance is lowered when more players are in the game). Secondly, based on ATMA's listed probabilities, there is a higher chance of a Razorswitch dropping on /p8 compared to /p1.

This is all getting rather confusing. I understand the core principle of 'more players in the game does not increase the chance of a roll being of higher quality', but ATMA appears to contradict that by giving better probabilities at higher player settings.