PROJECT --- Elemental chests loot analysis

I’m sure it has been discussed either in this thread or others, but can’t seem to locate it…I’m interested in what the algorithm is for determining the loot outcome for elemental chests. For example, how many possible rolls does each category get and what are the categories?

Algorithm wasn’t determined…
But drop rate % per roll were updated in OP.

Hi @Rowan1 and @Rfm , from data this is how it works:
-First one roll for crafting item: nearly always 4* and tiny chance of 3*
-Two rolls for ascension mats, 3* or 4* are possible
-One roll for an attack item, supposedly evenly distributed between 3 and 4*
-One roll for gems, 10, 20, 30, 40 or 50 gems can be obtained
-Three rolls for silver, troop or heroes tokens
-One roll for loot tickets, you can get 0,3 or 6
-One roll for emblems, it is possible to get 10, 20, 30 or 50 (forty is not possible)
-One roll for a flask, either world energy, raid energy or titan energy (supposedly 33.33% each)
-One roll for a trainer, either 1,2 or 3*
-One roll for Valhalla coins (2, 3, 5, 10, 20 or 30 coins)
-One roll for costume keys, supposedly 1,2 or 3 keys.

Now, with respect to “the algorithm” we can only have educated guesses. I will explain what I mean: We can only suppose the rolls are independent, so each is obtained by a separate random number and probabilities are independent from roll to roll, however we cannot claim that is true.

All the % rolls I have computed are considering that assumption however it could be the case that results are correlated (so as example it is way more possible to get 3 token silver or 3 heroes token together rather than what should be allowed in case of independence). Another dumb example would say that the algorithm could say: If the chest gives at least 1 four star AM, then impossible to get 50 emblems or more than one hero token, in which case it is obvious future rolls depend on the previous ones so they are correlated.

I actually plan on testing that assumption, however the amount of data is still too low to test on those assumptions and get clear cut results. Once I am done with this round of Finals in my Ph.D. classes I will test that properly and see what can be said about it.
Hope everything is clearer and do not doubt on asking more questions in case it is not!.

3 Likes

Well done. Very thorough explanation. One follow-up question though-regarding the two rolls for ascension mats…it looks like my last chest resulted in one compass. is there three possible outcomes? nothing, a 3* or a 4*?

@Rowan1 The last one you posted in the other thread you got a compass and a chainmail shirt. The shirt counts as an AM roll, because it is a 3* AM, even though it is a farmable one (and thus rather disappointing to pull).

1 Like

I do not think it is possible, I would check your screenshot or maybe you just forgot you got a scabbard or something like that?
It is always a 3* (unfarmeable or farmeable) or a 4* (from the data over 400 elementals)

@NotEye reminded me that there was a chainmail shirt. Thanks for your input guys!

@ierazo sorry I was not counting how many chests were opened before my last elemental chests.

So I have posted these numbers as 0. (I have posted 3 chests in a row, as I also forget to do it earlier)

That is fine, thanks @PlayForFun !

1 Like

You are welcome. Thanks for doing these statistics for us :slight_smile:
(BTW my first titan chest is also posted in the other form :slight_smile: )

1 Like

Excellent!! Thank you so much !
I am planning to post a mini summary with the low volume of data I have from war chests and in a few weeks for titan chests so stay tuned !!

2 Likes

My last 17 elementary chests got just one epic rise item (apricot blade), what happened to those chests? nerfed too? these chests are so rare that it should be mandatory to have at least one item of epic rise according to their element.
But unbelievably, absolutely nothing comes, I already gave up on anticipating chests, because they are giving items as common as normal chests.

My best loot from elemental chest => 150 holy monsters

image

Terrible

@ierazo are you still updating this and your other tracking projects (war and titan chests). I have been plugging in data, but wasn’t sure if you have sufficient data or are done with the projects.

Hi @Hurriphoonado , I am still updating the data but haven’t processed results. I have been a bit busy because I am studying for my Comprehensive exams of my Ph.D. program, but I want to get results soon, post them and their analysis, and then I will decide if I really want to continue running the projects or not (because it takes quite a bit of time). So in summary I plan on posting results hopefully within 2 weeks.
Thanks for continuing putting data !

2 Likes

@ierazo, question for you on this project.

Will you be updating the OP to contain the new Emblem Quantities?

1 Like

That is the plan @Guvnor .the same link works to submit the data, and I will parse it in a separate datafile to do the analysis!

2 Likes

maybe this needs a new breakpoint since emblem revamp?

@ierazo. Thanks for your wok and good luck with PhD thingy. :slightly_smiling_face: