You obviously didn’t pay attention in class, then.
“You can’t take samples from a population to make inferences about its parameters” is what you’re claiming, when that’s the entire reason the field of inferential statistics exists.
Haha, sadly there isn’t. We’ll have to try and get this going on the next PTR 
Crafting is the materials sink against botted goods (and possibly duped goods as well).
The only winners are RMT.
RMT’s are coming out on top in a lot of ways right now… -sad.
-pantz
To the OP fine job, thank you for providing amazing analysis. The noted 1/6 chance is not what is happening in game and that’s just sad.
wait… let’s add crafting luck… that’ll help… oh wait…
You do realize peer reviewed surveys often only use 1000 people as a sample size right?
1 Like
As far as I’m aware, AGS doesn’t post odds anywhere.
1 Like
I was referring to what was posted in the thread,
edited for clarity so people don’t get hung up on it.
-pantz
1 Like
First off I really appreciate the work. I have always wondered about this. you expected 9.1% and got 8.1%. Pretty close in my book tbh. Not enough difference imo to really call it unfair. It’s actually better than what I thought 600 was.
1 Like
Makes sense that heavy has worse rolls than light, have to make it hard as possible for tanks to get a good set, asmodeum is used most is very recipe and so is the most expensive and hardest to come by gotta keep you grinding the tread wheel for that one in a thousand roll that is actually a good.
I never understood why we can’t choose the perks for bags and tools, letting only the GS be random. Those arent game changing for PvP, there is no point in make the so expensive to make.
How we are still fighting for BASIC THINGS in the PTR 18 months later is so mind-boggling.
1 Like
Right, but at what point do you say the difference is big enough to be significant? That’s why we have to apply the binomial distribution to see the probability of the results by random chance.
So it tells us that the odds of getting 349/4300 or fewer (382 expected, -33) at 1/11 has a 1.2% of occurring by pure random chance, which is below the 5% error threshhold.
The important thing here is the sample size, since that helps reduce random deviation. At 1/11 the odds of getting 4/100 are actually higher than getting 349/4300, even though the former has a larger % difference from the expected.
Sorry dude, but these tests aren’t representative of the crafting community.
No one crafts without selecting max azoth.
Failure to do this values all this data pretty useless.
Can guarantee that the results would have been better if you did this. Don’t know how much better. But they would have been better.
Azoth only affects perk slots, not GS. You can still get non-gold 600s.
1 Like
My bad did not know this.
My point still stands though.
Most people are crafting only 595-600 and chasing Legos.
Going 590-600 and counting 600s is a bit redundant - no one crafts like this.
Yes, the point was to test if 600 is equally weighted across the window i.e. 1/11 or if it has lower odds than <600.
595-600 can’t be tested until next PTR goes live.
1 Like
Bags require dramatically less azoth, and the 600 chances for those seem even more abysmal, I’ll be doing these tests on the PTR with those to see what the odds actually are.
What is the point inknowing the odds of 590-600 crafting is what I want to know.
It doesn’t have any application, because everyone does 595-600.
It’s like me rolling 575-585 and counting how many 585s I get, just irrelevant data.
While I do not always agree with everything you post, this is some EXCELLENLY SUPERB research done to clear up confusion about crafting odds. Thank you for your efforts, they are GREATLY appreciated! 
1 Like
Did you, perhaps, miss this part of his post?
Or maybe this part?