| Pages: 1 2 3 :: [one page] |
| Author |
Thread Statistics | Show CCP posts - 0 post(s) |
YariLei
   |
Posted - 2011.01.10 06:50:00 -
[1]
For some reason, after updating to Incursion, I have started failing every single invention job I make (trying to do a basic frigate T2 invention). Previously I could get a success rate of approximately 40-50% with invention skills at level 4 and no decryptors. Now I've raised the skills to level 5 and updated to incursion and so far every job comes out with nothing of value. Do I really have to invest to expensive decryptors which cost about as much as the end product or am I just extremely unlucky? Anyone experiencing the same?
It seems such a waste to see all my time collecting datacores just being wasted in seconds, especially if previously I had no problems with lower skill levels...
|
Steve Thomas
Minmatar
|
Posted - 2011.01.10 07:32:00 -
[2]
Edited by: Steve Thomas on 10/01/2011 07:31:57
Welcome to the new Improved T2 lotto. (bascialy its the same thing its always been, your just going to run into streaks like that thanks to the way the Random number generator works)
the only way to realy deal with the problem is to plan to throw large numbers of Invention runs at the slot.
have fun!
|
Nikolai Kondratiev
Sphere Design Inc.
|
Posted - 2011.01.10 08:51:00 -
[3]
There are always good and bad streaks in anything chance-related. But on the long term it always comes close to your real chance %.
When I check my corp's invention stats, the success rate over the last 7/30 days is always around 49-50% (for most invention without decryptors @ ~48% chance + a few items invented with decryptors that increase invention chance) while daily success rate can sometimes go as low as 30-40% on bad days or as high as 60-70% on good days
_
Ore Table | PI Profits |
YariLei
|
Posted - 2011.01.10 08:56:00 -
[4]
Great, thanks a lot! I was wondering if they just made it super-hard to invent, but I suppose I'll just have to keep on putting in maximum amount of jobs I can (currently 5) and try to tweak it to a success-streak with a coupel of decryptors then. Thanks! 
|
AstarothPrime
|
Posted - 2011.01.10 10:15:00 -
[5]
Doing batches of 5 inventions at the time I found out that its either almost all fail or almost all hit.
5:0 and 4:1 are regular outcome so I guess random generator functions in batches as well :D
I.
|
Greg Huff
|
Posted - 2011.01.10 11:15:00 -
[6]
Success or Failure is determined when you start the invention job. As you can only start them one at a time batch size is irrelevant. You're only having a run of bad luck, but before you start using decryptors sharpen your pencils and work some numbers to make sure it's worth the added cost.
|
Jovan Geldon
Gallente
Lead Farmers
Kill It With Fire
|
Posted - 2011.01.10 12:21:00 -
[7]
Posting in this week's "ZOMG STEALTH INVENTION NERF" thread.
|
Aphrodite Skripalle
Galactic Defence Consortium
|
Posted - 2011.01.10 12:24:00 -
[8]
Because of the frustration and because we all know that eve is not bugfree at all it would be very helpful if you will get a more detailed answer why the invention failed.
Why not giving the user an answer with some statistic numbers like a percentage or whatever failed or just giving him the calculation of why the job failed ? There is a calculation behind that and it would really make a difference if you know what happens.
Invention of expensive BPC¦s like a marauder can easily cost you hundred of million isk and getting an explanation would be fair.
|
Nikolai Kondratiev
Sphere Design Inc.
|
Posted - 2011.01.10 12:34:00 -
[9]
Edited by: Nikolai Kondratiev on 10/01/2011 12:36:21
Assuming 50% base invention chance (which isn't too far from the 48.xx% you get with skills at 4 for modules) to make math easier. For a batch of 5 jobs, your chances are
- 5 success : 3.125%
- 4 success : 15.625%
- 3 success : 31.25%
- 2 success : 31.25%
- 1 success : 15.625%
- total failure : 3.125%
so 4 success/fails should be fairly common and full success/fails shouldn't be rare.
Edit for above : I assume server just rolls a random number between 0 and 1 and returns a success if it's <= to your invention chance, so it wouldn't really help to get a "Your invention failed because 0.738293923232 is greater than your invention chance of 0.4826" message.
_
Ore Table | PI Profits |
Honey Lovetrap
|
Posted - 2011.01.13 18:58:00 -
[10]
Edited by: Honey Lovetrap on 13/01/2011 18:59:10
I believe that CCP programmers or at least person who did program EvE RNG should be banned to ever come near computer EVER again.
12 in a row total failure...
Anyway this topic is enough reason to ask for reimbursement*, and ask that CCP publish RNG code.
*Something that we all should do, not because form of protest, but because things should be changed. And there were rumors that some players have superb above average luck doing anything in game, while others have bellow average "luck."
I also firmly believe that if program is changed by individual it wouldn't be detected for years.
So in theory its possible that friends of developers be more "lucky" and not to cause any kind of market devaluation of anything else, others regular players who dont have theirs feet wiggling from developer sitting parts will have lower chances to succeed, to get escalation, good loot...anything that is chance based in EvE.
Now because of that and only for peace of our minds CCP should recheck theirs RNG and release at mechanic of how it calculates chances.
|
Lutz Major
|
Posted - 2011.01.13 20:23:00 -
[11]
 Originally by: Honey Lovetrap
Edited by: Honey Lovetrap on 13/01/2011 18:59:10
I believe that CCP programmers or at least person who did program EvE RNG should be banned to ever come near computer EVER again.
12 in a row total failure...
Anyway this topic is enough reason to ask for reimbursement*, and ask that CCP publish RNG code.
*Something that we all should do, not because form of protest, but because things should be changed. And there were rumors that some players have superb above average luck doing anything in game, while others have bellow average "luck."
I also firmly believe that if program is changed by individual it wouldn't be detected for years.
So in theory its possible that friends of developers be more "lucky" and not to cause any kind of market devaluation of anything else, others regular players who dont have theirs feet wiggling from developer sitting parts will have lower chances to succeed, to get escalation, good loot...anything that is chance based in EvE.
Now because of that and only for peace of our minds CCP should recheck theirs RNG and release at mechanic of how it calculates chances.
lololololol
I too demand to be in that benefited group! And to make it easier for CCP, here is the code to add: If (player == Lutz Major) Then probability = 1 Else probability = crap End
|
Nikolai Kondratiev
Sphere Design Inc.
|
Posted - 2011.01.13 21:13:00 -
[12]
Originally by: Honey Lovetrap
Edited by: Honey Lovetrap on 13/01/2011 18:59:10
I believe that CCP programmers or at least person who did program EvE RNG should be banned to ever come near computer EVER again.
12 in a row total failure...
In fact you should be asking to ban that person from using computers if 12 failures in a row NEVER happened.
_
Ore Table | PI Profits |
Akita T
Caldari Navy Volunteer Task Force
|
Posted - 2011.01.13 21:47:00 -
[13]
Edited by: Akita T on 13/01/2011 21:48:22
Originally by: Honey Lovetrap
I believe that CCP programmers or at least person who did program EvE RNG should be banned to ever come near computer EVER again. 12 in a row total failure...
You Fail Statistics Forever...
Even with a (relatively high) 50% success chance, there is a 1 in 4,096 chance that a 12-failure streak could happen (precisely the same odds of a 12-success streak happening).
Considering the estimated number of invention jobs, that should probably happen to at least one person every few days, if not more often.
For lower success rates, it's even more likely.
_
Make ISK||Build||React||1k papercuts
_
|
Velicitia
Open Designs
|
Posted - 2011.01.13 22:27:00 -
[14]
Originally by: Akita T
<smart stuff>
and this is why I should stay very far away from whatever it is Akita does in a given day...
|
Steve Thomas
Minmatar
|
Posted - 2011.01.14 01:44:00 -
[15]
Originally by: Velicitia
Originally by: Akita T
<smart stuff>
and this is why I should stay very far away from whatever it is Akita does in a given day...
um
you just posted in a forum that Akita posted in.
Need some ducttape remover?
|
KaarBaak
Minmatar
Sebiestor tribe
|
Posted - 2011.01.14 04:38:00 -
[16]
Edited by: KaarBaak on 14/01/2011 04:39:41
Factors which may influence invention success:
- are you delivering jobs at the same time of day when you complete them?
- are you delivering them in batches (select all->deliver) or one at a time?
- are you starting the jobs all at the same time?
- are you getting your datacores from market or your own R&D agents?
- are you playing full-screen or windowed?
- are you left-handed or right-handed? Trackball or mouse?
- do you have AA turned on?
- did you install any/all of those mysterious "optional" patches last Autumn?
- was your char created on an even date or an odd date? *
Any of the above factors carry equal weight when determining success of invention jobs.
[/sarcasm]
KB
*There is a well-known MMO that actually used char creation date/time as a part of it's RNG seed, and after years of player complaints it was changed when it was determined that some avatars were literally "born lucky."
EDIT: added sarcasm tag, just in case.
Circumstances rule men; men do not rule circumstances. --Herodotus, Histories
|
Honey Lovetrap
|
Posted - 2011.01.14 14:00:00 -
[17]
Originally by: Akita T
Edited by: Akita T on 13/01/2011 21:48:22
Originally by: Honey Lovetrap
I believe that CCP programmers or at least person who did program EvE RNG should be banned to ever come near computer EVER again. 12 in a row total failure...
You Fail Statistics Forever...
Even with a (relatively high) 50% success chance, there is a 1 in 4,096 chance that a 12-failure streak could happen (precisely the same odds of a 12-success streak happening).
Considering the estimated number of invention jobs, that should probably happen to at least one person every few days, if not more often.
For lower success rates, it's even more likely.
Oh darling, I never said that there isnt a chance that it can happen.
Statistically its possible that some players will have 90% success even when eve RNG says 25%, and will be more "lucky" on short therm and where is a chance to be lucky in short therm there is also a chance to be more lucky on a long run...its theoretical probability.
No matter how tiny it is, probability exists.
And its possible that you have effective invention chance equal to 1% after completing 10000 jobs with 50% success rate.
I am saying that those numbers should be balanced.
In another words, EvE RNG should be updated that it have memory of success or failures in the past of each character individually and with those numbers on its mind balance chance of invention process.
Because in theory even with 80% chance, and after 1m jobs you can fail every one of them. So how it is fair?
It is statistical probability to have one group of players having much better effective chance then other group of players.
In another words, every failure is result of other player success, and every your success is result of other players failure.
Invention is one big ISK drain implemented for second class of players who joined later to EvE universe, and most of them are annoyance to CCP. If this statement is correct that EvE RNG is fine as it is now.
|
Velicitia
Open Designs
|
Posted - 2011.01.14 15:22:00 -
[18]
Originally by: Steve Thomas
Originally by: Velicitia
Originally by: Akita T
<smart stuff>
and this is why I should stay very far away from whatever it is Akita does in a given day...
um
you just posted in a forum that Akita posted in.
I said *should*... never said that I'm smart enough to do so .
|
Berikath
|
Posted - 2011.01.14 16:24:00 -
[19]
Edited by: Berikath on 14/01/2011 16:24:38
Originally by: Honey Lovetrap
Oh darling, I never said that there isnt a chance that it can happen.
Statistically its possible that some players will have 90% success even when eve RNG says 25%, and will be more "lucky" on short therm and where is a chance to be lucky in short therm there is also a chance to be more lucky on a long run...its theoretical probability.
No matter how tiny it is, probability exists.
And its possible that you have effective invention chance equal to 1% after completing 10000 jobs with 50% success rate.
I am saying that those numbers should be balanced.
In another words, EvE RNG should be updated that it have memory of success or failures in the past of each character individually and with those numbers on its mind balance chance of invention process.
Because in theory even with 80% chance, and after 1m jobs you can fail every one of them. So how it is fair?
It is statistical probability to have one group of players having much better effective chance then other group of players.
In another words, every failure is result of other player success, and every your success is result of other players failure.
Invention is one big ISK drain implemented for second class of players who joined later to EvE universe, and most of them are annoyance to CCP. If this statement is correct that EvE RNG is fine as it is now.
Here, let me shorten that for you.
Originally by: Honey Lovetrap
I don't like being unlucky, so they should change EVE's RNG so it isn't actually a RNG.
Also, "wahh".
*** [ SIG] ***
Wish list for PI:
*One-click input routing
*Copy product, inputs & outputs in factories
*Launchpad upgrades: twice the space, twice the cost, half the hassle!
[ /sig ] |
RaTTuS
BIG
Majesta Empire
|
Posted - 2011.01.14 16:39:00 -
[20]
Originally by: Honey Lovetrap
...Large W[h]ine ...
let me offer you some cheese ;-
last 300 invention Jobs I've done have resulted in
48% for my 5/5/5
52% for my 5/4/4
68% for my 4/4/4
over the last 3000 it's pretty much as it should be [though the 4/4/4 is slightly higher]
now what I thin this means is you should stop @ 4/4/4 and then just end yourself
--
Join BIG
|
Flios Bror
Amarr
Doom Guard
|
Posted - 2011.01.14 16:51:00 -
[21]
Originally by: Nikolai Kondratiev
Edited by: Nikolai Kondratiev on 10/01/2011 12:36:21
Assuming 50% base invention chance (which isn't too far from the 48.xx% you get with skills at 4 for modules) to make math easier. For a batch of 5 jobs, your chances are
- 5 success : 3.125%
- 4 success : 15.625%
- 3 success : 31.25%
- 2 success : 31.25%
- 1 success : 15.625%
- total failure : 3.125%
This is only the case for uncorrelated die-rolls. It's quite clear by now that the rnd() of Eve produces colored output, so your calculation is incorrect. Because we don't know the color of the noise, we cannot determine what the chances actually are.
[None] |
Captain Pompous
Is Right Even When He's Wrong So Deal With It
|
Posted - 2011.01.15 10:32:00 -
[22]
Edited by: Captain Pompous on 15/01/2011 10:33:02
Originally by: RaTTuS
now what I thin this means is you should stop @ 4/4/4 and then just end yourself
Real nice 
---
☻♥ Problem? Therapy sessions ♥☻ |
Nikolai Kondratiev
Sphere Design Inc.
|
Posted - 2011.01.15 10:44:00 -
[23]
Originally by: Flios Bror
Originally by: Nikolai Kondratiev
Edited by: Nikolai Kondratiev on 10/01/2011 12:36:21
Assuming 50% base invention chance (which isn't too far from the 48.xx% you get with skills at 4 for modules) to make math easier. For a batch of 5 jobs, your chances are
- 5 success : 3.125%
- 4 success : 15.625%
- 3 success : 31.25%
- 2 success : 31.25%
- 1 success : 15.625%
- total failure : 3.125%
This is only the case for uncorrelated die-rolls. It's quite clear by now that the rnd() of Eve produces colored output, so your calculation is incorrect. Because we don't know the color of the noise, we cannot determine what the chances actually are.
Not sure about others (and I'm certainly not going to make that kind of stats on job history to confirm) but the distribution feels totally normal here, lots of 4-6 success out of 10 jobs, 2-3 or 7-8 success are "uncommon" and rarely getting 0-1 or 9-10 success. And the feeling was the same when I was only using 5 slots for invention.
_
Ore Table | PI Profits |
Tres Farmer
Gallente Federation Intelligence Service
|
Posted - 2011.01.15 17:24:00 -
[24]
To add insult to the whole injury thing.. just recently I did put 10 freighter inventions into the cooker.. 7 JFs came out.. all jobs where started to yield 1run, ML-1, PL-1 prints.
Usually the average for this is 3.5 out of 10 to be successful.
support
Public Idea Tracker | 24hr PLEX |
Lutz Major
|
Posted - 2011.01.15 20:43:00 -
[25]
Originally by: Honey Lovetrap
Statistically its possible that some players will have 90% success even when eve RNG says 25%, and will be more "lucky" on short therm and where is a chance to be lucky in short therm there is also a chance to be more lucky on a long run...its theoretical probability.
No, it's not. It is correct, that on short term (ie 10 even maybe 20) someone can have a 90% chance (0.977% for 9 out of 10 and 0.018 for 18/20), but on long term (90 out of 100 = 0.00000000000000136554263874631% = IMPOSSIBLE) it is quasi zero.
Originally by: Honey Lovetrap
And its possible that you have effective invention chance equal to 1% after completing 10000 jobs with 50% success rate.
Can you please solve the equation (100 over 10000) / 2 ^ 10000 ... that would be the probability. My computer breaks when I try to solve it. Originally by: Honey Lovetrap
I am saying that those numbers should be balanced.
They are ... just make 1000+ invention runs and you will have (nearly) balanced results
Originally by: Honey Lovetrap
In another words, EvE RNG should be updated that it have memory of success or failures in the past of each character individually and with those numbers on its mind balance chance of invention process.
I'm speachless
Originally by: Honey Lovetrap
Because in theory even with 80% chance, and after 1m jobs you can fail every one of them. So how it is fair?
Not in a million years !
Originally by: Honey Lovetrap
It is statistical probability to have one group of players having much better effective chance then other group of players.
No it's not. It is your disability to understand probability!
RNG
|
Isonik Umarrio
|
Posted - 2011.01.15 22:23:00 -
[26]
I document my success rates, and for ammo, before incursion, my success on a 100 bpc invention run would be from 35-52%, my science skills are all at 4 (all 20 physics/engineering yada yada or so), i am using a hyasyoda lab, no special stuff like "test reports."
since incursion, i have done 2 100 bpc runs on ammo and my success is less than 30% on both.
i will post an update in a week once i have done my third 100 run and another personal run of 55 bpc's.
|
Kharylien
Gallente
Masked Rider Project
|
Posted - 2011.01.15 23:59:00 -
[27]
Edited by: Kharylien on 15/01/2011 23:59:43
Originally by: Isonik Umarrio
I document my success rates, and for ammo, before incursion, my success on a 100 bpc invention run would be from 35-52%, my science skills are all at 4 (all 20 physics/engineering yada yada or so), i am using a hyasyoda lab, no special stuff like "test reports."
since incursion, i have done 2 100 bpc runs on ammo and my success is less than 30% on both.
i will post an update in a week once i have done my third 100 run and another personal run of 55 bpc's.
For the love of God scroll up and read what Akita T and Lutz Major posted and then learn some statistics or never post again.
Originally by: Flios Bror
It's quite clear by now that the rnd() of Eve produces colored output
Proof or STFU.
|
Invention Nerf
|
Posted - 2011.01.16 07:03:00 -
[28]
Originally by: Isonik Umarrio
I document my success rates, and for ammo, before incursion, my success on a 100 bpc invention run would be from 35-52%, my science skills are all at 4 (all 20 physics/engineering yada yada or so), i am using a hyasyoda lab, no special stuff like "test reports."
since incursion, i have done 2 100 bpc runs on ammo and my success is less than 30% on both.
i will post an update in a week once i have done my third 100 run and another personal run of 55 bpc's.
I swear to god, I didn't do anything!
|
Greg Huff
|
Posted - 2011.01.16 15:06:00 -
[29]
Originally by: Invention Nerf
Originally by: Isonik Umarrio
I document my success rates, and for ammo, before incursion, my success on a 100 bpc invention run would be from 35-52%, my science skills are all at 4 (all 20 physics/engineering yada yada or so), i am using a hyasyoda lab, no special stuff like "test reports."
since incursion, i have done 2 100 bpc runs on ammo and my success is less than 30% on both.
i will post an update in a week once i have done my third 100 run and another personal run of 55 bpc's.
I swear to god, I didn't do anything!
Seriously??? You created and alt just so you could post with that name?
|
Invention Nerf
|
Posted - 2011.01.16 20:01:00 -
[30]
Originally by: Greg Huff
Originally by: Invention Nerf
Originally by: Isonik Umarrio
...
I swear to god, I didn't do anything!
Seriously??? You created and alt just so you could post with that name?
Yes. Funny, isn't it? And my second alt is called Sarcasm!
|
Ticker Troon
|
Posted - 2011.01.16 21:01:00 -
[31]
I just had a 10/10 run of Incursus -> Ishkur invention.
This is clearly a reverse nerf.
I want this fixed.
Now.
|
Wyke Mossari
Gallente
Staner Industries
|
Posted - 2011.01.16 22:58:00 -
[32]
Originally by: AstarothPrime
Doing batches of 5 inventions at the time I found out that its either almost all fail or almost all hit.
5:0 and 4:1 are regular outcome so I guess random generator functions in batches as well :D
I.
I agree, there seems to be definite clumping in the results, and there are two other things that I think affect the results.
There also seems to be a beginners luck fudge that significantly improves the chances for success the first couple of dozen or so jobs, this also appears to kick in after most patches, like something is being reset.
Characters that also do significant amounts of other research jobs, ME/PE/Coping also appear to earn a success bonus.
|
Linda Flamewalker
|
Posted - 2011.01.17 01:58:00 -
[33]
Originally by: Wyke Mossari
I agree, there seems to be definite clumping in the results, and there are two other things that I think affect the results.
There also seems to be a beginners luck fudge that significantly improves the chances for success the first couple of dozen or so jobs, this also appears to kick in after most patches, like something is being reset.
Characters that also do significant amounts of other research jobs, ME/PE/Coping also appear to earn a success bonus.
ANd where is your data backing this up? Or is this just your gut instinct?
|
RaTTuS
BIG
Majesta Empire
|
Posted - 2011.01.17 17:13:00 -
[34]
Invention rates have changed again ..
up to this morning
4/4/4 52%
5/4/4 52%
5/5/5 58%
and I bet they'll change again after the patch..
so you see the RNG hates you ... [or something]
--
Join BIG
|
RaTTuS
BIG
Majesta Empire
|
Posted - 2011.01.21 17:19:00 -
[35]
as a potential 49.56
I got
4/4/4 52.17%
5/5/4 48.75%
5/5/5 50.0%
up to incursion 1.1.1
--
Join BIG
|
Talon Dreyua
|
Posted - 2011.01.21 23:15:00 -
[36]
Ok, I am never one to complain, but I finally got frustrated enough to even look this subject up. Before incursion I was running a life time tally of near exact 45% success, I have been reviewing my success over the last 200 jobs, and it sits at 38.5% (77/200). I agree that previously I would have a few bad batches, but this is getting really frustrating.
|
Initori Kawashime
|
Posted - 2011.01.22 09:26:00 -
[37]
Edited by: Initori Kawashime on 22/01/2011 09:31:34
umm...
|
RaTTuS
BIG
Majesta Empire
|
Posted - 2011.01.25 17:24:00 -
[38]
Edited by: RaTTuS on 25/01/2011 17:26:40
Total
4/4/4 48.26%
5/5/4 64.86%
5/5/5 50.40%
Since Incursion 1.1.1
4/4/4 57.50%
5/5/4 64.85%
5/5/5 36.36% 
edit was stupid me not posting anythign
--
Join BIG
|
wr3cks
Reliables Inc
Majesta Empire
|
Posted - 2011.01.26 00:14:00 -
[39]
Originally by: Ticker Troon
I just had a 10/10 run of Incursus -> Ishkur invention.
This is clearly a reverse nerf.
I want this fixed.
Now.
No, the game just feels sorry for you because you're inventing things that are unprofitable.
Anyone who thinks that the RNG is well-designed needs to play the game for a few days. This entire game (client, server) is obviously one giant heap of spaghetti code. Since a good rand() function is relatively computationally expensive, and since the game uses so very many of them, it would not surprise me in a little if it was optimized in a way that made it streakier but faster to calculate.
|
dungabee
Minmatar
Molden Heath Software Company
|
Posted - 2011.01.28 22:39:00 -
[40]
As someone who does a lot of invention:
4/4/4 character.
Module invention:
pre-3rd part of Incursion: ~50%
since 3rd part of Incursion: ~30%
|
Eraseri
|
Posted - 2011.01.29 13:15:00 -
[41]
Confirming I have too noticed that the chance of invention seems to be dropped somehow.
|
Invention Nerf
|
Posted - 2011.01.29 14:54:00 -
[42]
Originally by: Eraseri
Confirming I have too noticed that the chance of invention seems to be dropped somehow.
Confirming I have noticed that the IQ rating seems to be dropped somehow.
|
DejaPeaux
|
Posted - 2011.01.29 16:04:00 -
[43]
Well for me, the pass rate has been very good, 4/5,5/5,3/5,5/5,4/5 i'm doing ammo..
Long may it continue.
|
dungabee
Minmatar
Molden Heath Software Company
|
Posted - 2011.01.30 11:56:00 -
[44]
All the highsec Incursions are in Amarr space. Incursions must be using the same Random Number Generator as inventions
|
V'Thalla
|
Posted - 2011.02.01 17:24:00 -
[45]
Originally by: Honey Lovetrap
In another words, every failure is result of other player success, and every your success is result of other players failure.
This statement is wrong. Each and every attempt is independent of any attempts prior or following.
|
RaTTuS
BIG
Majesta Empire
|
Posted - 2011.02.01 17:33:00 -
[46]
Total for base rate 40%
4/4/4 53.17%
5/5/4 51.926%
5/5/5 42.86%
Since Incursion 1.1.1
4/4/4 53.75%
5/5/4 55.26%
5/5/5 38.10%
5/5/5 is being stupid a lot recently
--
Join BIG
|
Malik Barrelhouse
|
Posted - 2011.02.04 04:44:00 -
[47]
Ok before Hulkaggedon was anounced I had a very good success rate at inventing Hulk BPC's at least 35%. When hulkaggedon was announced I went and invest in extra datacores to build up my BPC's for building after hulkaggedon to capitalize on the destruction. I have since had a 5% invention ratio. I have even used dycyptors. Anyone else having a less than average success rate?
|
Turat
|
Posted - 2011.02.04 04:59:00 -
[48]
i have had a pbad string last 3 days 4out of 11 have been bpcs oh well :(
|
Malik Barrelhouse
|
Posted - 2011.02.04 05:29:00 -
[49]
Originally by: Turat
i have had a pbad string last 3 days 4out of 11 have been bpcs oh well :(
I got 1 out of 20 in the last three days. I ussualy get close to 50%
|
Kerr avonn
Minmatar
Standard By 10
|
Posted - 2011.02.04 16:07:00 -
[50]
Just suck it up and deal with it i guess.
Does **** me off though when iv just wasted 50mill in decrypters 
|
Novee InFeldspar
|
Posted - 2011.02.05 01:26:00 -
[51]
Heehee, some of these posts made me chuckle.
320 inventions over the past four days (same item); my success rate (both engineering and the encryption skills at 4, stands at 48.75%.
156 successes vs 164 failures.
Not bad, pretty close to my all time (8171 jobs logged for that one item) success rate of 48.72%). 3981 successes vs 4190 failures.
Doing 10 at a time, I generally see the common 5/5, 4/6 or 6/4. Sure I see higher success rates than that, but very rarely do I get 8 or more successes from 10 jobs. Never seen a 10/10, either way.
Seems to be working fine for me, no difference noted.
I agree with the above poster who stated the random number generator may be simplified to speed calculation.
To the poster who thought toons who engaged in other research etc got better results... in my experience no. No change.
|
Untraceable Alt
|
Posted - 2011.02.11 08:04:00 -
[52]
Originally by: RaTTuS
Since Incursion 1.1.1
4/4/4 53.75%
5/5/4 55.26%
5/5/5 38.10%
5/5/5 is being stupid a lot recently
Makes sense... CCP wanted to nerf invention but they couldn't do it for newer players, or they would leave the game, so they decided to nerf only the professional inventors 
|
Bobhead
Caldari
Imperial Motive Power Corporation
|
Posted - 2011.02.11 20:11:00 -
[53]
Anyone ever consider that invention success might be tied to the number of said items on the market at certain given times?
|
Invention Nerf
|
Posted - 2011.02.11 21:43:00 -
[54]
Originally by: Bobhead
Anyone ever consider that invention success might be tied to the number of said items on the market at certain given times?
Yes, but it is a bullsh!t scenario, instead I just randomize all invention successes again.
|
Hakkar'al Gallente
|
Posted - 2011.02.12 11:11:00 -
[55]
Why is invention luck based, not that I can complain - have 4/4 so far (so of course the 5th will fail =) )? I would rather see some kind of player influenced method of improving the outcome of invention. Which brings me to my suggestion of researching skill books and/or blueprints.
Researching skill books and/or blueprints of the related T1 object for which you wish to have a T2 blueprint increases the probability of a successful T2 invention job.
How I'd see it if I want to improve invention outcome for a HAC blueprint.
1.Get one T1 cruiser bpc,
1a. Optionally get 1 cruiser skill book,
1b. Optionally get 1 HAC skill book
2. Add these to a pre-Invention "Study" job
3. Outcome of this job raises probability of success depending on object you put in (possibly of a new skill related to invention)
4. If new skill is implemented repeat until desired level of skill.
5. Proceed as now but with better chances of positive invention outcome
Pretty sketchy I know, but I think it would be neat. I don't like complete randomness in stuff I invest time in.
|
Invention Nerf
|
Posted - 2011.02.12 13:19:00 -
[56]
Originally by: Hakkar'al Gallente
Why is invention luck based, not that I can complain - have 4/4 so far (so of course the 5th will fail =) )? I would rather see some kind of player influenced method of improving the outcome of invention. Which brings me to my suggestion of researching skill books and/or blueprints.
Researching skill books and/or blueprints of the related T1 object for which you wish to have a T2 blueprint increases the probability of a successful T2 invention job.
How I'd see it if I want to improve invention outcome for a HAC blueprint.
1.Get one T1 cruiser bpc,
1a. Optionally get 1 cruiser skill book,
1b. Optionally get 1 HAC skill book
2. Add these to a pre-Invention "Study" job
3. Outcome of this job raises probability of success depending on object you put in (possibly of a new skill related to invention)
4. If new skill is implemented repeat until desired level of skill.
5. Proceed as now but with better chances of positive invention outcome
Pretty sketchy I know, but I think it would be neat. I don't like complete randomness in stuff I invest time in.
*cough, cough*
Raising the probability level doesn't make something any less random!
Also how is your method any different than to learn the corresponding invention skills on a higher level?
On a side note, I hope you trained for missiles, because the hit success of your drones and Gallente pew-pew are based on the same RNG as invention is.
|
Adrian Idaho
|
Posted - 2011.02.12 22:31:00 -
[57]
Edited by: Adrian Idaho on 12/02/2011 22:31:14
Originally by: wr3cks
Since a good rand() function is relatively computationally expensive, and since the game uses so very many of them, it would not surprise me in a little if it was optimized in a way that made it streakier but faster to calculate.
Complete and utter bull****, you have no idea what you're talking about. For counterexamples to your statement, look at xorshift RNGS (theoretical background) or the Mersenne Twister û both are very efficient, high-quality RNGs. |
RaTTuS
BIG
Majesta Empire
|
Posted - 2011.02.15 10:30:00 -
[58]
Total for base rate 40%
4/4/4 51.32% [above]
5/5/4 44.68% [above]
5/5/5 52.66% [above]
Since last patch
4/4/4 46.88% [below]
5/5/4 46.84% [below]
5/5/5 45.45% [below]
though 5 has been a lot batter, [and has been inventing other things not on this list above average]
--
Join BIG
|
dungabee
Minmatar
Molden Heath Software Company
|
Posted - 2011.02.15 18:13:00 -
[59]
Every time one of these threads appear (and there have been a LOT since Incursion 1.1) you get the same RNG-trolls coming out to insist it's ok because it's all random.
Funny that the same working-as-expected RNG keeps giving me the same mission 5x in a row from the same agent.
Funny that the same working-as-expected RNG keeps generating Incursions in Amarr space.
Funny that the same working-as-expected RNG has given an Invention success rate down by 10-20%.
I wonder what else these trolls think is random?
Global warming, fruit machines, online poker?
|
RaTTuS
BIG
Gentlemen's Agreement
|
Posted - 2011.03.28 10:37:00 -
[60]
Sine there have been more invention whines
Total for base rate 40%
4/4/4 51.20% [above]
5/5/4 52.88% [above]
5/5/5 42.58% [below]
Since last patch
4/4/4 51.11% [above]
5/5/4 53.33% [above]
5/5/5 38.24% [below]
5/5/5 is just not having the good luck at all - though the number of jobs he's done is way down [like 10% of the other 2]
--
Join BIG
|
RaTTuS
BIG
Gentlemen's Agreement
|
Posted - 2011.04.07 09:21:00 -
[61]
Since we have a new patch
Total for base rate 40%
4/4/4 51.00% [above]
5/5/4 52.99% [above]
5/5/5 45.45% [below] [but improving]
Since last patch
4/4/4 0%
5/5/4 0%
5/5/5 0%
5/5/5 has been increasing
--
Join BIG
|
Yith Za'bolazhi
|
Posted - 2011.04.07 11:20:00 -
[62]
Why or why do poeple have to necro these disfunctional forum threads.
|
RaTTuS
BIG
Gentlemen's Agreement
|
Posted - 2011.04.07 12:10:00 -
[63]
because we have a new patch and we'll get a spate of invention has been nurfed threads
--
Join BIG
|
Doctor Badger
|
Posted - 2011.04.07 14:58:00 -
[64]
Originally by: dungabee
Funny that the same working-as-expected RNG keeps generating Incursions in Amarr space.
Amarr holds sov in more empire systems than any other faction by a large margin. Of course there's going to be a disproportionate number of incursions into Amarr space.
Amarr: 702
Caldari: 326
Gallente: 388
Minmatar: 280
|
RaTTuS
BIG
Gentlemen's Agreement
|
Posted - 2011.04.15 16:32:00 -
[65]
Total for base rate 40%
4/4/4 50.65% [above]
5/5/4 53.17% [above]
5/5/5 47.23% [below] [but improving]
Since last patch
4/4/4 48.28% [slightly below]
5/5/4 56.56% [Above]
5/5/5 58.14% [Above]
|
Doddy
Excidium.
|
Posted - 2011.04.15 17:27:00 -
[66]
Originally by: Honey Lovetrap
In another words, EvE RNG should be updated that it have memory of success or failures in the past of each character individually and with those numbers on its mind balance chance of invention process.
So the RNG should not be a RNG at all.
|
Florence Valentine
|
Posted - 2011.04.15 17:55:00 -
[67]
I would like to point you to this excellent read, Alex's Adventures in Numberland.
Very well written, and gives a nice explanation of the "law of large numbers", which should help.
I.e. Your invention success rate over many, many, many attempts will average out at exactly the calculated rate found with this equation.
The "law of large numbers" also explains why gambling is a mug's game, but you knew that already.. right?
|
RaTTuS
BIG
Gentlemen's Agreement
|
Posted - 2011.05.16 10:29:00 -
[68]
Total for base rate 40%
4/4/4 51.84% [above]
5/5/4 50.88% [above]
5/5/5 48.69% [below] [but improving]
Since last patch
4/4/4 53.85% [Above]
5/5/4 43.81% [Below] [had a really bad run for a few days]
5/5/5 54.19% [Above]
|
Tehg Rhind
|
Posted - 2011.05.17 15:09:00 -
[69]
It is possible to calculate expected ranges of # of successes (X) for a total # of inventions (n) by using a normal approximation of a binomial distribution. It's pretty simple really. But all this data everyon keeps giving out is totally worthless if they don't include the # of trials (n).
Tempted to calculate the %iles (5/95) for 10/100/1k runs. Hmmmm. Yeah I'm going to do that.....
When CCP decides to do some overhaul of the market/industry stuff they should look at the way that past invention jobs are reported, showing the input characterisitics and including some aggregates of the data (n runs under these conditions, x successes)
|
Phizban
|
Posted - 2011.05.17 19:32:00 -
[70]
This one time, I invented something and it didn't work.
Sunny D and Rum, yum yum. |
RaTTuS
BIG
Gentlemen's Agreement
|
Posted - 2011.05.20 14:17:00 -
[71]
Total for base rate 40%
4/4/4 51.65% [above]
5/5/4 51.25% [above]
5/5/5 48.48% [below] [but improving]
Since last patch
4/4/4 70.00% [Above]
5/5/4 77.78% [Above]
5/5/5 54.55% [Above]
|
| |
|
| Pages: 1 2 3 :: [one page] |