Ephemera drop rate?

[ShadowPlagueXx]

The wiki says its 20%, but thats clearly not accurate since I've done 11 kuva liches and after the recent update i've done 17 sisters without getting an ephemera

Can anyone share their total liches and total dropped ephemera so I can compare the actual drop chances?

https://imgur.com/a/UqKGgEt

[(PSN)biskuitbau]

I did 6 kuva liches and got two ephemeras. So I guess the 20% isn't so accurate

[Krankbert]

6 and 9 aren't big enough samples to draw any conclusion.

[ShadowPlagueXx]

Just now, Krankbert said:

6 and 9 aren't big enough samples to draw any conclusion.

yes but the overall ephemera drop rate is 20% so with 20 liches done I should expect around 4 right?

[_Euphoria_]

ı just want taht sweet Gloriana Ephemera and nothing else, but stupid RNG wont drop me one

[Monolake]

New ones are more like 10%  according to people who ran hundreds of sisters combined

[(XBOX)C11H22O11]

59 minutes ago, ShadowPlagueXx said:

yes but the overall ephemera drop rate is 20% so with 20 liches done I should expect around 4 right?

You can make 100 and still not get any ephemera, unlikely but possible, it's just random it took me 5 to get the Zarina while my friend got it on the first lich

[Jarriaga]

1 hour ago, ShadowPlagueXx said:

yes but the overall ephemera drop rate is 20% so with 20 liches done I should expect around 4 right?

No. It's RNG. You can potentially always see an Ephemera or never see them at all even at 20%. For what it's worth: I've generated a total of  8 Sisters and 2 had an Ephemera.

[Venus-Venera]

vor einer Stunde schrieb ShadowPlagueXx:

The wiki says its 20%, but thats clearly not accurate since I've done 11 kuva liches and after the recent update i've done 9 sisters without getting an ephemera

Can anyone share their total liches and total dropped ephemera so I can compare the actual drop chances?

https://imgur.com/a/UqKGgEt

those 20% mean you almost never get it. you can farm it for half a year and still not get it. or longer.

the chance does not increase with every kill and that is the problem. the thing is an absolute waste of time!

[Fabpsi]

Hi there @ShadowPlagueXx  ^^

as a few other people already said here: RNG! 
Yes...this is never the answer players are looking for, but it still is a very valid point to make.

e.g. I did so far 12 Sisters of Parvos, and got 4 Ephemera:

Spoiler

But I don't go around and say "look the ephemera-dropchance is even higher than I thought it would be/ than they said!" (because that's not how it works 😅)
I got lucky so far, and other people got unlucky. There really is nothing more to it.

TL&DR:
RNG, nothing unusual, working as intended.

Still I hope that you will get some of the new ephemera soon! 

Best of luck & happy hunting Tenno!
Cheers,
Fabpsi~

[vanaukas]

People laugh at me or say that "I don't understand how math works" in this forum for poiting out how the average user understands RNG and then we have this thread....

[DaMasque]

My highest record of not getting a ephemera is 12 lichs.... on my 3 sister and I already lost hope.

RNG~ Dance with me~~~ Oh RNG~~~~

[Surbusken]

I figure it will be the same as with kuva liches, only better.

Since people are faster to farm the items, there will be more of it. So you start out selling whatever you get, then once the prices drop you buy the specific things you wanted.

[Birdframe_Prime]

Yeah, the problem with RNG is that when it says 20%, it means 20% overall.

There is no 'if I take on 5 liches, one of them will have the drop'.

If you were to take a million liches, then yes, on analysing that drop rate it would work out at 20% of them having Ephemera.

However there is nothing to say that you can't go 60 in a row without an Ephemera and then get 10 in a row, miss the next 20, then get another 10 in a row. Meaning the average was 20%. That's how true random generation can work.

Not fun, but functional.

[(PSN)Lollybomb]

I got an ephemera on my first go.  Clearly the drop rate is 100%. /s

Quick look around about statistics, minimum sample size should be at least 100.
Unfortunately most of the information I'm finding is about surveying populations, not testing random events.  Of course given that a sample size of 20, one extra or one less will swing the result wildly, I'd say it's a bad sample size based on that alone.  I mean, +/-5% result from just one extra event?

 

[(PSN)Lollybomb]

Wheee, forum lag creating double posts...

 

[(XBOX)CI shadow2397]

2 hours ago, ShadowPlagueXx said:

The wiki says its 20%, but thats clearly not accurate since I've done 11 kuva liches and after the recent update i've done 9 sisters without getting an ephemera

Can anyone share their total liches and total dropped ephemera so I can compare the actual drop chances?

https://imgur.com/a/UqKGgEt

 

2 hours ago, (PSN)biskuitbau said:

I did 6 kuva liches and got two ephemeras. So I guess the 20% isn't so accurate

It’s called rng you aren’t garunteed anything

2 hours ago, ShadowPlagueXx said:

yes but the overall ephemera drop rate is 20% so with 20 liches done I should expect around 4 right?

The drop rate is 20% so for every 5 lichens you should expect at least one, however in case you forgot this is a system that runs on rng meaning, it’s a CHANCE not a garuntee 

[MqToasty]

2 hours ago, ShadowPlagueXx said:

yes but the overall ephemera drop rate is 20% so with 20 liches done I should expect around 4 right?

That is the "expected value", or basically the most probable outcome from 0 to 20.  But the odds of getting exactly 4 is only:

0.8^16 * 0.2^4 * 20choose4 = 21.8%

The odds of someone being in your shoes (getting none) is:

0.8^20 = 1.15%

So basically, 1 out of every ~87 players will be in your shoes.

Personally, I got 3 ephemeras out of the 6 Corpus Liches I did so far (but only 1 out of 18 Grineer Liches I did in the past), so...  Just remember that RNGesus giveth, RNGesus taketh away...

[MqToasty]

21 minutes ago, (PSN)Lollybomb said:

I got an ephemera on my first go.  Clearly the drop rate is 100%. /s

Quick look around about statistics, minimum sample size should be at least 100.
Unfortunately most of the information I'm finding is about surveying populations, not testing random events.  Of course given that a sample size of 20, one extra or one less will swing the result wildly, I'd say it's a bad sample size based on that alone.  I mean, +/-5% result from just one extra event?

 

Actually, the margin of error calculation is pretty simple.  It's just 1/sqrt(sample size).  So when sampling populations (say, for a survey), the typical rule of thumb is 1000 samples for a +/- 3.2% margin of error.  Nothing intrinsically magical about this number, but just widely accepted because it "seems good enough".  At 100 samples the margin of error is +/- 10% and at 20 samples the margin of error is +/- 22.4%.

[(PSN)Lollybomb]

5 minutes ago, MqToasty said:

Actually, the margin of error calculation is pretty simple.  It's just 1/sqrt(sample size).  So when sampling populations (say, for a survey), the typical rule of thumb is 1000 samples for a +/- 3.2% margin of error.  Nothing intrinsically magical about this number, but just widely accepted because it "seems good enough".  At 100 samples the margin of error is +/- 10% and at 20 samples the margin of error is +/- 22.4%.

I figured someone would correct me on error and minimum sample size.  Which is fine.  I never took proper statistics classes, but I do know at least how to calculate fairly simple probabilities.  In my defense, this was just a quick search, and 100 came up as pretty much the "absolute minimum", and my +/-5% is just how much of a change a single event can make.  I'm not surprised the actual error rate is more than quadruple that.

Plus now I know how to calculate margin of error, so I can start throwing that at people who run around saying "I did 10 of this and didn't get a single one!  30% chance my butt!"

[MqToasty]

2 minutes ago, (PSN)Lollybomb said:

I figured someone would correct me on error and minimum sample size.  Which is fine.  I never took proper statistics classes, but I do know at least how to calculate fairly simple probabilities.  In my defense, this was just a quick search, and 100 came up as pretty much the "absolute minimum", and my +/-5% is just how much of a change a single event can make.  I'm not surprised the actual error rate is more than quadruple that.

Plus now I know how to calculate margin of error, so I can start throwing that at people who run around saying "I did 10 of this and didn't get a single one!  30% chance my butt!"

It's alright, statistics is a specialized field with a lot of math and assumptions built-in, so just knowing the basics is good enough.  Yep, most of us do not perform and track nearly enough runs of anything to get any results with high confidence, so just about every single one of these topics are just people venting.  Which I can understand, to an extent...

[(PSN)biskuitbau]

2 hours ago, (XBOX)CI shadow2397 said:

 

It’s called rng you aren’t garunteed anything

The drop rate is 20% so for every 5 lichens you should expect at least one, however in case you forgot this is a system that runs on rng meaning, it’s a CHANCE not a garuntee 

So it's kinda like gacha? It's rng but guaranteed every 5 liches?

[(XBOX)C11H22O11]

Just now, (PSN)biskuitbau said:

So it's kinda like gacha? It's rng but guaranteed every 5 liches?

No

[vanaukas]

6 minutes ago, (PSN)biskuitbau said:

So it's kinda like gacha? It's rng but guaranteed every 5 liches?

How many times do you need to read that this is pure RNG and no prize/ephemera/drop is guaranteed no matter how many runs/liches/sisters you do?

[Birdframe_Prime]

5 hours ago, (XBOX)CI shadow2397 said:

The drop rate is 20% so for every 5 lichens you should expect at least one, however in case you forgot this is a system that runs on rng meaning, it’s a CHANCE not a garuntee 

No, my friend, that would be a system of 'One ephemera guaranteed in 5 runs or less'. We know that system exists in other games, there's a faux RNG system that's used for some Loot Box systems, where if you go a certain amount of rolls without getting a high tier, you're guaranteed a high tier. But you still aren't guaranteed the one you want.

Warframe at least lets you pick the reward you're going for and all you're rolling for is whether you get it or not.

Warframe runs on true RNG.

If you have 5 runs with 20%, that's 5 separate attempts at a 20%, they aren't linked.

If you take what's called an aggregate view, where the more you do, the more chances you've had, meaning the more likely you are to get one over time? That's where the percentage starts to rise.

With an RNG of 20% the aggregate of 5 runs is barely over 50%.

After 10 runs you can be 90% sure of having got one, but there's still a chance that you won't. Your rate goes up to nearly, but not quite, guaranteed after 21 runs or higher.

However, as I said above, there's actually a chance that you could go 50-60 runs without getting one, only to get the next 5 or 10 runs with an ephemera every time. That's how actual RNG works.