These forums have been archived and are now read-only.

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

Science & Industry

 
  • Topic is locked indefinitely.
123Next page
 

Invention success rates nerfed??
Author
Previous Topic Next Topic
Dread Nanana
Doomheim
#1 - 2012-07-23 16:14:45 UTC
Has CCP nerfed invention success rates in last few weeks??

My invention rates are down from 52% down to 46% since last month. Did anyone notice anything in this time period?
Ginger Barbarella
#2 - 2012-07-23 16:18:34 UTC
I've noticed a SLIGHT downward trend, but nothing more than anecdotal, really. Given the other non-peeveepee changes in last few months I wouldn't be surprised if the overall success rates were nerfed.

"Blow it all on Quafe and strippers." --- Sorlac
Steijn
Quay Industries
#3 - 2012-07-23 16:47:51 UTC
mine are down significantly over the last 2 weeks, just puitting it down to the RNG atm.
Tsadkiel
Aliastra
Gallente Federation
#4 - 2012-07-23 17:09:15 UTC
at what statistical confidence level do you make those claims? > . >

because these posts pop up all the time
Volar Kang
Kang Industrial
#5 - 2012-07-23 17:47:44 UTC
I just finished 15 invention jobs and had 7 successes. I'm not feeling any nerf.

@Tsadkiel, I almost spit up when I saw your post, nice one mate.
shar'ra matcevsovski
Doomheim
#6 - 2012-07-23 18:06:28 UTC
Dread Nanana wrote:

My invention rates are down from 52% down to 46% since last month.


mind telling us what you are inventing with wich decryptors? 46% does not sound too bad tbh.

shar'ra phone home
Smohq Anmirorz
State War Academy
Caldari State
#7 - 2012-07-23 19:33:27 UTC
Unless you're plugging numbers into the binomial probability formula, your post reads like this:

Last month I flipped a coin 10 times and got 5 heads. This month I only got 4. Did someone change coins?



Start here:

http://www.mathwords.com/b/binomial_probability_formula.htm


After that you can quantify the probability that you get that many successes in that many tries.
Dread Nanana
Doomheim
#8 - 2012-07-23 19:47:25 UTC
Just to add,

Long term average, over about 80,000 invention attempts is 49.4%. My average is just over 50%, mostly due to perfect skillz Shocked

http://games.chruker.dk/eve_online/invention_chance.php?base_item=none&decryptor=none&base_chance=40&skill_encryption=5&skill_datacore_1=5&skill_datacore_2=5&precalculated_chance=hidden

But over the last thousand invention attempts, success rate is down to 46%. It seems about 4% down since last month and the 25+ month long average (average monthly success rate is going up as other people in corp get better skills)


So no, this is not about a 10 invention run getting 0 What?. Could be just chance, but it is getting a little on the outside of standard deviation from the long term mean.
Pipa Porto
#9 - 2012-07-23 22:01:43 UTC
According to this: http://stattrek.com/online-calculator/binomial.aspx

460 (or fewer) successes out of 1000 with a single trial probability of success of .494 has a probability of .017 (or a 1.7% chance for each 1000 attempt group), so it's not particularly unlikely to happen by random chance.

The problem is that, most likely, you picked your data looking back. If you want statistical rigor, you need to say: "The Next Thousand Runs are going to be my test."

EvE: Everyone vs Everyone

-RubyPorto
Derath Ellecon
University of Caille
Gallente Federation
#10 - 2012-07-24 01:26:05 UTC
Yes they got nerfed along with deadspace module drops, melted Nanoribbon drops and officer spawns.

Glad to see this pops up in more places than just missions and wormhole forums.
Casiella Truza
Ecliptic Rift
#11 - 2012-07-24 02:15:20 UTC
This is a troll, right?

I never can tell the difference anymore between a troll and extreme incompetence.
Kusum Fawn
Perkone
Caldari State
#12 - 2012-07-24 02:44:37 UTC
posting just to post in an "invention nerfed" thread.

this is not among the areas of my expertise

Its not possible to please all the people all the time, but it sure as hell is possible to Displease all the people, most of the time.
Smohq Anmirorz
State War Academy
Caldari State
#13 - 2012-07-24 03:55:52 UTC
Dread Nanana wrote:
Just to add,

Long term average, over about 80,000 invention attempts is 49.4%. My average is just over 50%, mostly due to perfect skillz Shocked

http://games.chruker.dk/eve_online/invention_chance.php?base_item=none&decryptor=none&base_chance=40&skill_encryption=5&skill_datacore_1=5&skill_datacore_2=5&precalculated_chance=hidden

But over the last thousand invention attempts, success rate is down to 46%. It seems about 4% down since last month and the 25+ month long average (average monthly success rate is going up as other people in corp get better skills)


So no, this is not about a 10 invention run getting 0 What?. Could be just chance, but it is getting a little on the outside of standard deviation from the long term mean.


OK, fair enough. Just pointing out how it can and will be read by a lot of people without the hard data.

Pipa Porto wrote:

According to this: http://stattrek.com/online-calculator/binomial.aspx



lol...cheater.
Dread Nanana
Doomheim
#14 - 2012-07-24 05:53:30 UTC
Pipa Porto wrote:
According to this: http://stattrek.com/online-calculator/binomial.aspx

460 (or fewer) successes out of 1000 with a single trial probability of success of .494 has a probability of .017 (or a 1.7% chance for each 1000 attempt group), so it's not particularly unlikely to happen by random chance.

The problem is that, most likely, you picked your data looking back. If you want statistical rigor, you need to say: "The Next Thousand Runs are going to be my test."


OK. Thanks.

I guess I'll just have to wait for more data! Lol
Pipa Porto
#15 - 2012-07-24 06:01:36 UTC
Smohq Anmirorz wrote:
Pipa Porto wrote:
According to this: http://stattrek.com/online-calculator/binomial.aspx



lol...cheater.


Cheaters Prosper.

EvE: Everyone vs Everyone

-RubyPorto
Ginger Barbarella
#16 - 2012-07-25 18:00:26 UTC
Tsadkiel wrote:
at what statistical confidence level do you make those claims? > . >

because these posts pop up all the time


That's part of the problem with making claims/prognostications on intarwebs posting boards: One man's "significant statistical level" is another man's "you're not doing it enough."

"Blow it all on Quafe and strippers." --- Sorlac
Pipa Porto
#17 - 2012-07-25 21:41:35 UTC
Ginger Barbarella wrote:
Tsadkiel wrote:
at what statistical confidence level do you make those claims? > . >

because these posts pop up all the time


That's part of the problem with making claims/prognostications on intarwebs posting boards: One man's "significant statistical level" is another man's "you're not doing it enough."


Post your P-values then.
http://en.wikipedia.org/wiki/P-value

Also, post before you run your test, so it's clear that you're not picking some anomalous streak looking backwards.

EvE: Everyone vs Everyone

-RubyPorto
Ginger Barbarella
#18 - 2012-07-26 00:57:50 UTC
Pipa Porto wrote:
Ginger Barbarella wrote:
Tsadkiel wrote:
at what statistical confidence level do you make those claims? > . >

because these posts pop up all the time


That's part of the problem with making claims/prognostications on intarwebs posting boards: One man's "significant statistical level" is another man's "you're not doing it enough."


Post your P-values then.
http://en.wikipedia.org/wiki/P-value

Also, post before you run your test, so it's clear that you're not picking some anomalous streak looking backwards.


First, I don't play forum epeen games.

Second, I doubt very seriously ANYONE would run any single run, with a single toon with fixed skill levels, for a sufficient level to keep the statistics wonks happy. My success rate I look at overall; some things I have great success rate overall on (Adaptive Nano, 66.67% for 396 runs since accumulating data), some numbers can't be trusted (Heavy Assault launchers, 65.22% for 298 runs, some using MetaX launchers to help the success rate, some not), and others are just a "statistical" nightmare: Mjolnir Fury Lights have a rate of 1 in 25, for 73 runs). But those numbers aren't anywhere NEAR enough for a real statistical analysis; I know this, but I'm not losing any sleep over it.

In a nutshell, waving the epeen over statistical non-sense in a game is just foolish, given the probability that CCP will tweak (or mess up code) that changes the odds without telling the community at large (see recent "discovery"). Be happy with what you have, and if you're not, examine what you're doing to see what you can do better (or stop doing).

As I said already, my OBSERVATION is purely anecdotal; I've seen a slight decrease, but it very well can fall into the grey area.

"Blow it all on Quafe and strippers." --- Sorlac
Emperor Salazar
Remote Soviet Industries
Insidious Empire
#19 - 2012-07-26 11:51:30 UTC
I ran 10 invention jobs the other night.

Not a single success.

Did CCP nerf invention to 0%???????????????????????????????????
Emperor Salazar
Remote Soviet Industries
Insidious Empire
#20 - 2012-07-26 11:52:49 UTC
Ginger Barbarella wrote:


As I said already, my OBSERVATION is purely anecdotal.


Anecdotal is just a fancy word for useless, when it comes to observations.
123Next page