Drop rate data collection & anomoly tracking

#1

Lately I’ve been keeping a log of drops so I can get a rough idea of how likely items actually are, and to hopefully use this information to deconstruct the loot dropping process. Over the past couple weeks I’ve run over 700 dungeons keeping track of Drops of Interest, barely scratching the surface.

There have been moments where I’ve made some weird observations, which I’m currently hoping to back up with numbers:
•Loot Tier Potions increasing UT drops.
•Certain items dropping more often when the player count is within certain ranges.
•Dungeon Potion drop rates spiking at certain player counts.
•Multiple people getting UTs and STs or no one getting them.

I personally believe there are multiple steps to the loot dropping process, including one that looks at player count, a step that decides a range of how many items to give out on average (Loot Drop Boost influences this), and another step to upgrade the quality of items dropped (Loot Tier Boost influences this step).

I created this thread to

  1. Seek any insight people here might have about drop rates, including research they’ve done themselves.
  2. Hear about drop rate quirks you all might have observed.
  3. Ask for community collaboration in continuing to collect data and interpret the information.

I’ve read a few threads from years back, and some that even attempted colab data sheets like the one I’m working on myself. That kind of stuff is really inspiring and I’m here to do my part in demystifying drop rates.

#2

It’s always cool to see someone trying to dive into the stats for drop rates! A couple things-

700 dungeons is very, very little from a statistical significance standpoint. Especially when you are analyzing sooo many variables, you simply cannot come to so many conclusions from such a small sample size.

Loot tier potions have been confirmed multiple times by Deca to not impact UT/ST drop rates.

There is already a robust discord server dedicated to determining drop rates. If you would like to join them and their enormous and ever-growing data set you can do so here: https://discord.gg/zx24QJP. Do note that the bot handing the data has been down for a couple weeks, unfortunately, but I think it is better to wait for it to be fixed then to start over and lose 76,000 dungeon runs worth of data.

#3

I haven’t really observed any, or at least nothing that could not be explained by randomness.

E.g. “Multiple people getting UTs and STs or no one getting them.”. You might think a more uniform distribution would be reasonable, so e.g. if the drop rate is 1/100 and 50 people do the dungeon then one person gets a drop every other time. But RNG doesn’t work like that. It often results in lumpy distributions, odd spikes.

There are a couple of psychological reasons too why RNG seems odd. You get hundreds, if not thousands, of UTs, STs, while playing. But the ones you remember are the special ones, which can be especially rare drops, or multiple drops, or rare coincidences such as getting the same drop multiple times in adjacent runs.

And when one person says what they got other people are more likely to do so. So it may seem like you see more cases of multiple people getting drops just as at other times when just one person does they don’t announce it.

#4

Thanks for the reply. I’m well aware 700 is a drop in the bucket, especially since it’s across numerous dungeons. But multiple people working together can produce huge results :slight_smile: Thank you for the discord link, I’ll join that instead of starting my own most likely.

#5

Yeah lol that’s true. I have taken these things into consideration, but when it happens I still scratch my head and wonder what’s the deal. I still feel like there is something to learn about how bosses give items even if it’s just capricious rng when it comes to the actual rare stuff.Even if it’s just a curiosity, and not something that can be harnessed.

#6

I actually just looked into the server for the first time in a while and it seems that the project has stopped for an indefinite amount of time. The data is still available to look through by searching the bot commands channel.

#7

alright no problem. I’ll put my discord link back up

#8

Another anomaly I forgot to mention in OP, so far it’s been twice as likely for Limon to drop Sprite Wand when there’s two players (~20%) instead of solo (~10%).

#9

This topic was automatically closed 60 days after the last reply. New replies are no longer allowed.