| Pages: [1] 2 :: one page |
| Author |
Thread Statistics | Show CCP posts - 1 post(s) |
Alyssah
   |
Posted - 2010.04.10 15:26:00 -
[1]
I use 7 slots for invention mainly drones, ammo, missiles and shield items.
As of late My success rate has plummeted to the following:
0/7; 1/7; 2/7; 1/7; 1/7; 2/7; and 1/7.I usually get about 3/7.
Is it just poor random luck or has there been unreported changes?
|
Arous Drephius
|
Posted - 2010.04.10 15:29:00 -
[2]
Yes, CCP changed the invention mechanics so that you, and only you get poor results. Sorry, but you'll have to biomass your character to get good results again.
|
frsd
Caldari
Kleinrock Heavy Industries
The Kadeshi
|
Posted - 2010.04.10 15:31:00 -
[3]
Strange, and i thought Invention succes rates were only to be decreased when Tyrannis hits the shelves. 
Sharing is a nice gesture. Stupid but nice... |
Alyssah
|
Posted - 2010.04.10 15:31:00 -
[4]
I've only been playing the game for 4 years so the paranoia effect has not kicked in yet. Sorry about your infliction.
Now, seriously.........
|
Furb Killer
Gallente
|
Posted - 2010.04.10 15:38:00 -
[5]
Normally that only happens when patches/expansions are deployed. Combined with worse rats, less faction rat spawns, worse missions and worse loot tables if we can believe the forums.
|
Don Pellegrino
The Tuskers
|
Posted - 2010.04.10 19:03:00 -
[6]
Every new patch has an incredible effect on most people: They start noticing each "bad luck" moment more than before.  Then they go to the forums and start making threads about "Stealth invention nerf" and the like. Omg... even if they reduced the invention success rate, prices would go up, but margins remain the same, ffs.
You simply had a few bad days, I hope next week will be better for you.
Welcome to the magic of random numbers.
|
Capt Fossil
Gallente
|
Posted - 2010.04.10 19:47:00 -
[7]
Random...........
Why is it still THAT random when I have trained ALL the relevent skills to level 5??? Why?
Very little reward for all that training time, IMO.
|
FunzzeR
Death of Virtue
MeatSausage EXPRESS
|
Posted - 2010.04.10 20:03:00 -
[8]
 Originally by: Alyssah
Now, seriously.........
Seriously
My last few runs batches of nine(Modules) were:
9/9, 7/9, 8/9, 7/9, 5/9, and 8/9
I think invention was stealth boosted.
PRAISE THE SCOTTISH FOLD!!
THEIR WILL SHALL BE DONE!! |
|
CCP Shadow
C C P
C C P Alliance
|
Posted - 2010.04.10 20:31:00 -
[9]
Moved to Science and Industry.
|
|
Don Pellegrino
The Tuskers
|
Posted - 2010.04.10 20:55:00 -
[10]
Originally by: Capt Fossil
Random...........
Why is it still THAT random when I have trained ALL the relevent skills to level 5??? Why?
Very little reward for all that training time, IMO.
I know you are probably trolling, but for the people reading this thread wondering what difference level 5 skills make:
Linkage
|
Zartrader
|
Posted - 2010.04.10 21:36:00 -
[11]
Edited by: Zartrader on 10/04/2010 21:39:01
Average chance does not mean actual results unless you do an infinite number of inventions. You're just on the edge of the bell curve and you only notice things when they are against expectations. With the amount of people inventing the probability of getting bad results for a good number of players is very high. Also true for those with very good results. You happen to be one of them.
Anyway, all my Inventions recently have been just above the expected percentage. That was also pure chance although more likely than your result, far from certain.
I also assumed you used a chance calculator and comparing like with like, how do you know previous invention success was not high?
|
Brock Nelson
Caldari
Flux Technologies Inc
SRS.
|
Posted - 2010.04.10 22:07:00 -
[12]
Originally by: Alyssah
I've only been playing the game for 4 years so the paranoia effect has not kicked in yet. Sorry about your infliction.
Now, seriously.........
Seriously? If you've played for 4 years, you would've already know by now that invention is purely chance based.
Store | Get SRS |
Herr Wilkus
|
Posted - 2010.04.10 23:24:00 -
[13]
Actually, I had a 0/8 today. I think OP is right. Something is fishy.
|
Jill Xelitras
|
Posted - 2010.04.11 00:02:00 -
[14]
If you want to keep statistics of your invention success, keep all data and not just the last 5 batches of x jobs.
The bell curve has already been mentioned in a previous post.
Have a look at this info about normal distribution. See the the picture in the right top corner, those are bell curves.
If you have calculated your average rate of success using one of the many helpful eve related sites then this average represents the peak of a bell curve. Your actual success on each job is either 0 or 1. On a batch of 8, you can see anything between 0/8 and 8/8. The more jobs you consider, the more unlikely it gets that you stray away from the expected average. Once you have 100 or even 1000 jobs in your statistic, you will be very close to the pek of the bell curve (the expected average). At that point you can have 20 failures in a row and it will only throw you 20/1000 = 2% away from your average.
I hope this helps a bit.
|
Tellenta
Gallente
Invicta.
Cry Havoc.
|
Posted - 2010.04.11 00:20:00 -
[15]
Originally by: Alyssah
I use 7 slots for invention mainly drones, ammo, missiles and shield items.
As of late My success rate has plummeted to the following:
0/7; 1/7; 2/7; 1/7; 1/7; 2/7; and 1/7.I usually get about 3/7.
Is it just poor random luck or has there been unreported changes?
It sounds like you need to learn how to practice proper eve voodoo.
My typical reaction to that scenario is to halt all invention produce all my blueprints wait 1 downtime of having no t2 bpc's in my hanger and starting again. No, really I do this it works. I know just about every inventor I have talked to has some sort of 'voodoo' they perform when things start going awry. You need to figure out what actions you need to take to appease the eve gods to get back the the roughly %50 you should be getting for module invention.
|
Misty Lilly
|
Posted - 2010.04.11 17:14:00 -
[16]
Modules, Drones, Ammo have been reported to have a base invention of 40%, modified up based on skills.
My overall history of these items - without decryptors - is 7,382 successes out of 14,909 attempts; or 49.51%
However, over the past 3 months only - I've had 2,556 successes out of 5,069 attempts; or 50.42%
I attribute the improvement to continuing skill training improvements and not to a change in the mechanics.
My suggestion is to keep a longer term history so when you have the few runs of 0/7 you don't get too disappointed. Similarly, don't get overly excited when you get the 7/7 successes. It all balances out in the end.
I must also admit, there are days when I'll do 200+ invention jobs, so I don't even notice the various ups-and-downs anymore....
|
Ikserak tai
|
Posted - 2010.04.12 14:55:00 -
[17]
Originally by: Alyssah
I've only been playing the game for 4 years so the paranoia effect has not kicked in yet. Sorry about your affliction.
Now, seriously.........
Fixed it for you.
It could be infliction, as in self-inflicted pain I suppose, but then by playing Eve we're all inflicting it on ourselves, no?
YOU'VE NEVER ROCKED 'TIL YOU'VE UNDOCKED. |
Rashmika Clavain
Gallente
|
Posted - 2010.04.12 15:19:00 -
[18]
Your sampling metrics are rather poor.
Please make a note of the next 1000 invention jobs and we can dicsuss it further.
|
Umega
Republic Military School
|
Posted - 2010.04.12 20:52:00 -
[19]
I was playing poker last night.. typical friendly sunday occurance.
And I got three straight full houses! Amazing.
Then the next 25 hands I pulled out only 5 low pair.
What are the odds..
----------------------------------------
-Treat the EVE Market like you're a pimp and it is your 'employee'.. freely fondle it as you wish and make it pay you for it- |
Taedrin
Gallente
Xovoni Directorate
|
Posted - 2010.04.12 21:45:00 -
[20]
Edited by: Taedrin on 12/04/2010 21:51:17
Originally by: Jill Xelitras
If you want to keep statistics of your invention success, keep all data and not just the last 5 batches of x jobs.
The bell curve has already been mentioned in a previous post.
Have a look at this info about normal distribution. See the the picture in the right top corner, those are bell curves.
If you have calculated your average rate of success using one of the many helpful eve related sites then this average represents the peak of a bell curve. Your actual success on each job is either 0 or 1. On a batch of 8, you can see anything between 0/8 and 8/8. The more jobs you consider, the more unlikely it gets that you stray away from the expected average. Once you have 100 or even 1000 jobs in your statistic, you will be very close to the pek of the bell curve (the expected average). At that point you can have 20 failures in a row and it will only throw you 20/1000 = 2% away from your average.
I hope this helps a bit.
Just a college student here, but from what I understand, Invention follows binomial distribution (since it is essentially a true/false event that happens with a given probability). So let's say you have 10 runs, with each run having a 50% chance of success. According to my formula sheet that I was given at the beginning of the semester, we use the formula:
P(x)= n!/(x!*(n-x)!) * p^x*((1-p)^(n-x))
where x = number of successes
where n = number of trials
where p = probability of success for each trial
So for x = 1 success, n = 10 trials and p = .5 chance of success per trial, we have .00976 chance of one (and only one) trials succeeding. That's a 1% probability - very low to be certain but not unheard of.
The wikipedia entries state that binomial distribution will only approximate normal distribution under certain circumstances. The wiki entry gives a rule of thumb n*p > 5 and n*(1-p) > 5. Since we used p = .5 in our example above, n*p = 5 and n*(1-p) = 5, so the example above is the most ideal approximation we can get to normal distribution without further increasing the number of trials. Let's say you get REALLY greedy, and decide to run an invention job on a battleship with a max run decryptor. This gives you p = .15 (according to this). Our formula gives us P(1) =.347 ~ 35% chance of happening. Now, this is the chance of one, and ONLY ONE success. If we also include the possibility that we have zero successes, this increases to 54%.
Just glancing at a table of cumulative binomial probabilities, I see that for p = .3 and n = 10 we have P(x<=2) = 38%.
Now, let's take the OPs example: "0/7; 1/7; 2/7; 1/7; 1/7; 2/7; and 1/7." This is 7 batches of 7, so presuming that these are consecutive batches, we have 49 trials, with 8 successes. The OP said that 3/7 is normal for the OP, so that's p = .42 (Which seems to be pretty close to the chance of inventing a module with the worst possible skills). Using Excel's BINOMDIST() function, I find for p = .42, n = 49, x = 8 you have P(x <= 8) = .0001 probability (.01%). So, if the OP is being truthful here, either the invention chances have changed or the OP has experienced a very, very rare event! This is of course presuming that the OP is being truthful here, and isn't handpicking results. If the OP is handpicking events AT ALL, then we can only look at each batch individually, and not together as a whole.
If we take only the 0/7 batch, we have P(x<=0) = .02 (2%), which is rare, but within the realm of imagination. Furthermore, P(x<=1) = .134 (13%) and P(x<=2) = .377 (38%). We see that the P(x<=2) events are actually not that rare.
In conclusion, if the OP hasn't been manipulating the selection of data at all, then the OP is justified in fearing some sort of invention change. However, if we consider that only the people who suffer these rare events post on the forums....
----------
Originally by: Dr Fighter
"how do you know when youve had a repro accident"
Theres modules missing and morphite in your mineral pile.
|
Misanthra
|
Posted - 2010.04.13 01:34:00 -
[21]
Don't want to think about the math...all I know is recent inventions been good and bad to me. was like 4/10 for scourge invention and like 8/10 for juggernaut.
I call invention in game gambling personally...sometimes I win big and some days I just burn up alot of cores and bpc's for nothing lol
|
Backfyre
|
Posted - 2010.04.13 18:54:00 -
[22]
Edited by: Backfyre on 13/04/2010 18:54:55
I recently tracked almost 100 drone inventions and about bullseyed the expected success rate of 48% for all L4 skills. Early on, I had a stretch where I got 2 in 12. Today, I was 3 for 5 on marauders. Some days you win, others you lose.
|
SurrenderMonkey
|
Posted - 2010.04.13 21:08:00 -
[23]
Breaking news:
Pseudo-randomly generated numbers are pseudo-random.
---------------
Faction-Militia:Player-Alliance::Newbie-corp:Player-corp |
Taedrin
Gallente
Xovoni Directorate
|
Posted - 2010.04.14 01:47:00 -
[24]
Originally by: SurrenderMonkey
Breaking news:
Pseudo-randomly generated numbers are pseudo-random.
Still, the chances of getting 8 successes or less out of 49 consecutive trials is very, very low. The OP has either witnessed a very rare event (my numbers came out to a .01% chance), or invention algorithms have been changed. Or most likely the OP was handpicking results which makes statistical analysis impossible.
----------
Originally by: Dr Fighter
"how do you know when youve had a repro accident"
Theres modules missing and morphite in your mineral pile.
|
Lord Fitz
Project Amargosa
|
Posted - 2010.04.14 03:48:00 -
[25]
Originally by: Furb Killer
Normally that only happens when patches/expansions are deployed. Combined with worse rats, less faction rat spawns, worse missions and worse loot tables if we can believe the forums.
Strangely enough the next expansion could very well actually have worse loot tables, depending on your POV.
|
Enthral
|
Posted - 2010.04.14 04:14:00 -
[26]
Originally by: Taedrin
Originally by: SurrenderMonkey
Breaking news:
Pseudo-randomly generated numbers are pseudo-random.
Still, the chances of getting 8 successes or less out of 49 consecutive trials is very, very low. The OP has either witnessed a very rare event (my numbers came out to a .01% chance), or invention algorithms have been changed. Or most likely the OP was handpicking results which makes statistical analysis impossible.
Posting a thread on bad success rates is, by definition, handpicking results. I don't see evidence the OP ever posted when he was having good results. I've never seen anyone scream "Wow, I just got 8 out of 10 successes three times today!" But this happens to people all the time.
I have thousands of invention jobs under my belt, and I typically do five or six runs of ten jobs each, which I've done since the beginning of invention. With maxed skills, my success rate is very consistently over 50%, when measured over the course of several days, several weeks, or even several months. I've noticed no changes, no increased incidence of "bad luck", and no increased incidence of "good luck".
Contrary to popular belief, invention is not a gamble, but it is a commitment. You can't just toss up a few jobs of whatever item strikes your fancy and expect results...
|
Kun'ii Zenya
|
Posted - 2010.04.14 06:49:00 -
[27]
Edited by: Kun''ii Zenya on 14/04/2010 06:51:14
FGS...hasn't this topic been covered enough?
Please, get a probability/statistics book....
That goes for you too Taedrin.
|
ingenting
20th Legion
Sodalitas XX
|
Posted - 2010.04.14 15:30:00 -
[28]
L2Statistics, you're only having a bad streak, and someone else is having a good streak.
|
Taedrin
Gallente
Xovoni Directorate
|
Posted - 2010.04.14 19:27:00 -
[29]
Originally by: Enthral
Originally by: Taedrin
Originally by: SurrenderMonkey
Breaking news:
Pseudo-randomly generated numbers are pseudo-random.
Still, the chances of getting 8 successes or less out of 49 consecutive trials is very, very low. The OP has either witnessed a very rare event (my numbers came out to a .01% chance), or invention algorithms have been changed. Or most likely the OP was handpicking results which makes statistical analysis impossible.
Posting a thread on bad success rates is, by definition, handpicking results. I don't see evidence the OP ever posted when he was having good results. I've never seen anyone scream "Wow, I just got 8 out of 10 successes three times today!" But this happens to people all the time.
I have thousands of invention jobs under my belt, and I typically do five or six runs of ten jobs each, which I've done since the beginning of invention. With maxed skills, my success rate is very consistently over 50%, when measured over the course of several days, several weeks, or even several months. I've noticed no changes, no increased incidence of "bad luck", and no increased incidence of "good luck".
Contrary to popular belief, invention is not a gamble, but it is a commitment. You can't just toss up a few jobs of whatever item strikes your fancy and expect results...
What I mean by handpicking results is that he is conveniently ignoring certain results. Going back to the OPs statement:
"0/7; 1/7; 2/7; 1/7; 1/7; 2/7; and 1/7."
You will notice that the OP has actually told us remarkably little here. It is up to the reader to make certain assumptions to make any sort of sense from this data. Are these contiguous results? Did the OP look at a greater collection of samples and pick out the worst ones to present us?
When I ran my numbers, I presumed the following scenario:
"What is the probability that there will be 8 or fewer successes out of 49 trials when the probability of success for each trial is 3/7?" I determined that there is a .01% chance that this would happen. One out of ten thousand.
Consider you own *module* invention history. How many 49 *consecutive* jobs yielded 8 or fewer successes? Chances are, you don't have any.
But anyways, the OP hasn't been here in awhile, so I presume that means that invention chances - in fact - have NOT been changed. He simply observed a very rare event, or was lying. This is further supported by the fact that nobody else is on the forums complaining about such flukes. I suppose it is also possible that the OP accidentally tried to invent something with a much lower chance of success, like hulks or battleships instead of modules. That would also explain his bad streak of luck.
The argument that I am trying to make here is that what the OP has described is NOT a common event, and is in fact significant. Given what the OP knew, he was justified in coming to the forums and ask if anyone else is having similarly bad luck.
----------
Originally by: Dr Fighter
"how do you know when youve had a repro accident"
Theres modules missing and morphite in your mineral pile.
|
Borun Tal
Minmatar
|
Posted - 2010.04.14 21:30:00 -
[30]
Nope, my industrialist's success rate overall has been increasing as I go from L4 skills to L5... My VERY rough average right now is about 35%; two days ago a small run of drone inventions were 13 successes for 14 attempts.
|
| |
|
| Pages: [1] 2 :: one page |
| First page | Previous page | Next page | Last page |