These forums have been archived and are now read-only.

The new forums are live and can be found at https://forums.eveonline.com/

EVE General Discussion

 
  • Topic is locked indefinitely.
 

Invention stealth nerf (or bug) ?
Author
Previous Topic Next Topic
Taipion
Adeptus Petrous
#1 - 2012-12-11 12:33:12 UTC
Hello everone,


I know, as always, when playing with RNG, this might be just coincidence, but chances are at about 1000:1


Since the big "Retribution" patch, my invention-success dropped so low, one could call it "highly unlikely".

I tend to run a lot of inventions, and I happen to know binomial distribution, and, while calculating the probability of the results, it looks that the actual invention chance is at 0.4 now, where it should be 0.504 as all related skills are at V.
This is a big difference!

Usually, invention results would even out, if taking a big enough sample, but since retribution, they wont do it anymore!


Is that a bug, a stealth nerf, or just plain bad luck in high amounts?!
Beckie DeLey
Garoun Investment Bank
Gallente Federation
#2 - 2012-12-11 12:43:03 UTC
So, by "a lot" you mean like 10000 invention jobs, right?

My siren's name is Brick and she is the prettiest.
Chribba
Otherworld Enterprises
Otherworld Empire
#3 - 2012-12-11 12:43:33 UTC  |  Edited by: Chribba
I think we need more data to know for sure. I know I could invent for a week and have really bad results, and then another week with good results. So just one week after Retribution I think is too soon to know exactly. I will be checking my own data a bit further ahead anyway bc for me at least a week is too soon.

Also, if I am to share my success rate, I have done close to 35,000 invention jobs. And I have a success rate of 47,58%

And since Retribution I actually have a 50% rate (altho haven't done many this week so that's not very accurate).

/c

★★★ Secure 3rd party service ★★★

Visit my in-game channel 'Holy Veldspar'

Twitter @ChribbaVeldspar
Lexmana
#4 - 2012-12-11 12:46:20 UTC
How much data do you have? Also, in a group of 300k EVE players it is not that strange if one observes something one day with 1000:1 odds.
raskonalkov
Tie Fighters Inc
#5 - 2012-12-11 12:50:29 UTC
Yes, it was.
Lysa Riay
Pator Tech School
Minmatar Republic
#6 - 2012-12-11 12:51:17 UTC
I'd say the run of bad luck started before the patch hit!

But I'd also add that CCPs random number generator has been known to produce blocks of good or bad results for at least 7 years.
So 10 to 100 bad results doesnt a stealth nerf make!


30ish years ago I ran a test on some "random number generators" & most required millions (and some 10s or 100s of millions) of results to show that they were random, not a handfull!

Could Klang take the (now rotting) pizzas & the launcher out as they leave the building plz.
raskonalkov
Tie Fighters Inc
#7 - 2012-12-11 12:57:36 UTC
^ I like that, because its realistic.
Taipion
Adeptus Petrous
#8 - 2012-12-11 13:25:12 UTC
Chribba wrote:
I think we need more data to know for sure. I know I could invent for a week and have really bad results, and then another week with good results. So just one week after Retribution I think is too soon to know exactly. I will be checking my own data a bit further ahead anyway bc for me at least a week is too soon.

Also, if I am to share my success rate, I have done close to 35,000 invention jobs. And I have a success rate of 47,58%

And since Retribution I actually have a 50% rate (altho haven't done many this week so that's not very accurate).

/c


Well, then, it might be just plain bad luck.

Maybe I am just too used to predictable results, before Retribution, you could basically take any streak of ~200 inventions, and it would already even out and come pretty close to what it should be.
I have been inventing for 3 or 4 years now, and I never encountered anything like that.
Darwins Hammer
Doomheim
#9 - 2012-12-11 13:31:46 UTC
Quote:
binomial distribution



Isn't binomial distribution a double peak?


Also what are you inventing. Im sure that success rates over time are related to itemss


You can have a long bad patch inventing warrior IIs compared to inventing Golums. As if you did not you would quickly go broke on the later.
Taipion
Adeptus Petrous
#10 - 2012-12-11 13:57:01 UTC
Darwins Hammer wrote:
Quote:
binomial distribution



Isn't binomial distribution a double peak?


Also what are you inventing. Im sure that success rates over time are related to itemss


You can have a long bad patch inventing warrior IIs compared to inventing Golums. As if you did not you would quickly go broke on the later.


Uhhm, you don't sound like you know how invention works, for all things that are not ships, you can have exactly 50,4% chance with perfect skills and not decryptors.

And with "binomial distribution", I mean that I know how to use formualas like that to calculate probabilities.