I dont know but on the reset day i got the 10k bar, 2x 1k bar and 3x 100 bar so…
Its been 3 weeks, 9 is the most data I could collect so far.
Machine A gives red marbles 80% of the time and 20% of the time blue marbles
Machine B gives blue marbles 80% of the time and red marbles the other 20% of the time
Both machines are unlabled
Would it take you longer than 9 marbles drawn to figure it out?
In this case you get to draw 100 marbles from one machine and 9 from the other. Im pretty sure you could figure it out.
Showing ONE result that’s far from standard deviation does not mean that the underlying probability changed. To actually prove that Una chests are “nerfed”, you’d need to prove that there’re more than 0.00038% of such events happened, which means about more than 38 occurrences in a pool of 100000.
If I show this, what would be your conclusion? An extreme outline event happened, OR stone cutting percentage nerfed omg pls stop shilling for AGS?
Which of the following is more likely:
A. You being a lucklet
B. The same thing happening to everybody but you are the only one complaining
Yeah that doesn’t work you can’t use a sample of 9… your point is irrelevant
First week, can`t remember 
Second week i got 10k bar, 1k bar and 4x 100 bar + usual gold, from opening 3 big boxes.
This week i got nothing, only the gold.
Okay this week i got like 2.5k when last week it was almost 20k 
i bought 3 chests and all 3 combined was 3,984 gold so from each big chest i got 1,328… i feel like it’s actually nerfed. what are the odds that each chest to give the exact amount? 1328 each x3 ? pretty sneaky nerf to una gold.
Thats about the correct number. 1250 is the expected gold amount (The chests avgs around 2.5 per token spent).
i could swear this is way lower than i used to get before.
Nope, its always been an expected value of about 2.5-3.0 gold per una token since launch.
80 Bag = ~200-240 Flat Gold with a small chance for 100g Bars
200 Bag = ~500-600 Flat Gold with a small chance for 100g Bars and a Tiny chance for 1000g bars.
500 Bag = ~1250-1500 Flat Gold with a small chance for 100g Bars, a Tiny chance for 1000g bars and a rare chance for the chunky 10k bars.
People have stated that with the smaller bags you have a very SLIGHTLY better chance of getting the smaller bars then the 500 chest but you remove all possibility of getting the 10k bar.
I have to see if I can hunt down an old chart people used to get a decent sample size but It was found that the more consistent best gold gains were from the 200 small chest (had the highest avg gold return) and the 500 large chest has the worse avg return rate unless you pop the big chungus with the 80 bag just being “middle of the road”
I got my first ever Legendary Gold Bar AFTER the change with Una’s Coins. So your whole thread is invalid.
1 Like
I’ll just add my bit to it:
I have never gotten a 10k gold bar
I have seen 1 1k gold bar
This has been since I started playing in April, there has been no change in my drops and I generally see just over 1.2k per box.
Those that were getting massive amounts of gold, you were getting lucky. What you are now experiencing is what most of us have been experiencing throughout the game with Una boxes. How about you be glad you had that run of luck instead of complaining because you had no idea what the norm is and crying nerf?
1 Like
9 chests is statistically irrelevant and could happen entirely by random chance you’ve proven nothing…not saying it did or did not change just that your info and waiting 2 more weeks proves nothing you clearly don’t know how stats work. The fact that you are calling people dumb makes me feel free to not be nice to you.
Some light reading for you: Statistical significance - Wikipedia
Your p value isn’t sufficient your results are anecdotal and therefore prove nothing. If you wanted this to be relevant you’d likely need to do this over more like 100 weeks to prove anything and have real data based on the previous data you’re referencing not such an insignificant 9 chest sample size. Your 75% is reading the data wrong your 5% isn’t correct and even if it was there would still be a 5% chance you still are just wrong.
TLDR: Stay in school kids!
This statement proves 100% you have never published before.
A real published researcher would never, not ever, say insane things like “the data is clear” or any strong declarative statement like that.
No credible academic ever says such definitive statements even when repeated tests from peer reviews provide additional support. Not ever.
At very best, one would say “the data suggests”, “thus far the predicted model appears true given X data”, etc.
Nice try but your own stubborn insistence of your proof being definitive fact is all the reason to conclude you are at best a hack, and more likely a fraud.
I got 10 large chests, I don’t remember the exact names but I got 6 small and 2 medium gold bars. I remember it was pretty much 3 gold per token.
I have always taken the large chest. I do remember at least once gotten the largest bar, maybe two times in total. It’s pretty rare.
Ouch! zilla versus killa
I am going to update this post for all the trolls who don’t understand statistics, probability or how science is done. I have opened 6 more large gold chests with no thin bars, thick bars or giant bars dropped. That is 15 total with no thin gold bars, in korea data posted a while ago 100 large chests gave 75 gold bars. I have done a new two tailed z score to show the probabilities of this occuring if drop rates are the same between these 2 populations. The p value is less than .00001. That is less than a one in a 100 thousand chance. The standard significance for a paper to be published is less than .05 which is less than a one in 20 chance. The evidence is failry overwhelming that the drop rate of large gold chests has been changed for thin gold bars from what the korea players data mined in the past.
There is good reason for AGS/Smilegate to do this to combat gold farming bots. Go ahead and troll me for saying i cannot draw results from a sample size of 15. You are wrong.
[/quote]
No, it’s because you are wrong.
Both sample sizes are too small when testing the randomness of a random system. That’s why if you use your calculator you would see that it would only take a sample size as low as 2 to compare it to the 100 samples to still received a “significant” result.
That is logically absurd. A sample size of 2 to test a random system is stupid and if you think it is a reasonable sample size, then you aren’t just a normal level of ignorant but you are aggressively so.
1 Like
Yo this guy legit thinks he took statics using low sample size
HAHAHAHHA