Hello again from droprates server! This time I’m posting keyper droprates because event started recently, when the event ends I’ll try to update this topic with newest droprates becaue we have people consistently logging them. If you got any questions please ask, because this new way of showing droprates might be more confusing for some people.
(Server invite link: https://discord.gg/ZgwDsM8)
It is from previous keyper event too, but from what Toastrz told me droprates haven’t changed.
Looks like wide confidence intervals. Redundant IMHO, I think we can get all the information we need from the (xx drops, yy runs) figures.
Edited it real quick, Red are lowest possible %, Orange are middle and Green are highest possible %, at least from what I understood myself, because I’m not one who programmed the bot.
If you understood you right, I’ve been doing that before I started working with Melo (Bot creator), but confidence intervals make it somewhat more accurate, right now we use 99% because when we get to higher amount of drops it gets a lot more accurate than >95%.
Yes, thought they looked wide. 95% is the normal default for stats, 99% is excessive without there being a very good reason for it.
But really we just need the numbers, so how many drops of each. That’s what people normally post. Confidence intervals in this case are meaningless as they don’t reflect any additional data so don’t tell you anything extra, they just make it harder to read the actual data.
wouldn’t this be affected by number of people in runs, whether you hit threshold, number and type of guaranteed drop, and like millions of other variables I have no idea about.
Do you really think Keyper has guaranteed event whites? If you hit SB you get a shard so there’s that to check.
The white bag drops from Keyper, like all white bag drops, have fixed % chances that are not affected by other people, or by guaranteed drops. Clearly you need to hit the SB threshold, but with Keyper that is straightforward (get some damage in before his final phase) and obvious when you do.
Can anything affect the rate? Loot drop potions/Lucky Clover are the only thing. Also the rate can vary over time, as can what drops. The big differences between some of the items could e.g. be due to some not dropping all the time.
All i know is from all this information gathering across multiple events, by multiple individuals, pogmur is one of the most common (aside from ray and kage) which tbh surprises me.
Basically my main got banned for “TP Cland”, I won’t got into details, but rn I have second acc Vaieforthe. I use this one as my forum acc because I see no reason not to. c:
And well, I quit but I might get back on motmg.
Vale here decided to play ROTMGTroller style one day with a cshield 
and got perma’d, his new acc is vaieforthe but I assume he uses this one on forums for the white star.
edit: ninja’d
I’m not sure this is particularly accurate. I understand the use of confidence intervals to give a range, but im not sure if its been calculated correctly, since I would expect the middle values to be calculated directly from the drops/runs (that is, the last four i would expect to have middle value 1 in 4969/2, being approximately 1 in 2500)
Cool stuff. This is a great idea if it can be executed properly.
I am not sure about using a 99% confidence interval for this kind of research, however. Without very specific method setups, that is not generally used. The range produced by that is very likely to be misleading with the quantity (and arguably, quality) of data available. I am also not sure if a confidence interval is warranted in this setting.
This is one of the problems I have with this, confidence intervals make stuff really really confusing. I already told bot creator about this and I’ll see what he does with that.
Hey, @Gidey and @Skandling, would you be able to join server? Melo would love to talk with you a little! 
Post anything you want to write here. Much better to keep the discussion in one place, and here is the best place for in-depth discussions on the game.