open All Channels
seplocked Science and Industry
blankseplocked Invention: Formulae
 
This thread is older than 90 days and has been locked due to inactivity.


 
Pages: first : previous : 1 2 3 4 [5] 6 7 8 9 ... : last (14)

Author Topic

Tahreem
Minmatar
The Littlest Hobos
Ushra'Khan
Posted - 2007.04.01 22:27:00 - [121]
 

to reply on some peepz asking about faction mods to help increase the chance..

I tried to use dread guristas cloaking device to get my chances higher for good result..

You can choose the item but when you activate the job it moans something like:

This job is to sophisticated for our scientists..

All needed skills are at lvl 4

Imprezina
Posted - 2007.04.02 03:49:00 - [122]
 

What will i get if I use 1-run bpcs for modules? for ships?

Aykido
Gallente
Lobster of Babel
Posted - 2007.04.02 08:48:00 - [123]
 

Trinity with a 300 run BPC you would get 10 runs +encryptor run modifier.

Trinity Faetal
Gallente
Little Garden
Posted - 2007.04.02 14:25:00 - [124]
 

Originally by: Aykido
Trinity with a 300 run BPC you would get 10 runs +encryptor run modifier.


yeah just found, thx Smile

Vladimir Tinakin
Caldari
Wife Aggro Productions
Posted - 2007.04.02 16:10:00 - [125]
 

Well, thanks to the stealth nerf on invention, my previous success rate datasets are worth exactly two things: Jack and ****. And Jack got podded.

Since then on TQ I've attempted about 10 hulk inventions with no success with 4/4/4 skills and a mixture of Test Reports and Stolen Formulas, and with covetors. I'm a bit peeved.

However, I had a few jobs pending on Singularity. Delivered them (4/3/3 skills there, and the jobs were no decryptor but with a covetor) and got 75% success....so maybe the stealth nerf is rolled back some with the new build? Oveur accidentally misplace the decimal point with the current invention success rate perhaps?

Guess we'll see tomorrow.

CampyloBacter
Gallente
Federal Defence Union
Posted - 2007.04.02 16:58:00 - [126]
 

I've noticed a strange phenomenon occurring in my invention of Cap recharger 2s. All skills at 4, and using fixed parallel link-capacitors as the meta item (second best I think)and initially just 10 run BPCs and no decryptors, I had ten successes out of twelve attempts, each however was for a 1 run T2 BPC. Then, when my nice new 300-run BPCs came of the photocopier, I changed to using them. My first four attempts with the 300 run BPCs but everything else the same (and in two of those jobs using the best Meta item) have all failed.

Has anyone else noticed a decreased chance of success with increasing BPC run size? Or could this just be down to chance?

The cynic in me thinks that the Devs want small scale invention to be more successful just to allow players to manufacture items for their own use, but to prevent inventors threatening the T2 BPO owner's stranglehold on the market. We're already seeing nerfs to invention, doubtless following bleating from the monopoly holders who fear the genuine competition which invention has the potential to introduce.


Qual
Gallente
Cornexant Research
Posted - 2007.04.02 21:40:00 - [127]
 

Originally by: Vladimir Tinakin
Well, thanks to the stealth nerf on invention, my previous success rate datasets are worth exactly two things: Jack and ****. And Jack got podded.

Since then on TQ I've attempted about 10 hulk inventions with no success with 4/4/4 skills and a mixture of Test Reports and Stolen Formulas, and with covetors. I'm a bit peeved.

However, I had a few jobs pending on Singularity. Delivered them (4/3/3 skills there, and the jobs were no decryptor but with a covetor) and got 75% success....so maybe the stealth nerf is rolled back some with the new build? Oveur accidentally misplace the decimal point with the current invention success rate perhaps?

Guess we'll see tomorrow.


Nerf? There has been no nerf. You are experincing really bad luck. It happens.

Qual
Gallente
Cornexant Research
Posted - 2007.04.02 21:44:00 - [128]
 

Originally by: CampyloBacter


Has anyone else noticed a decreased chance of success with increasing BPC run size? Or could this just be down to chance?



Chance. I do my module invention using 300 runs. ~70% succes over time. (No change before and after patch. Large sample base of 100+ jobs.)

Vladimir Tinakin
Caldari
Wife Aggro Productions
Posted - 2007.04.02 22:29:00 - [129]
 

Originally by: Qual

Nerf? There has been no nerf. You are experincing really bad luck. It happens.


Just seems to be a lot of anecdotal evidence pointing to a general upping of the difficulty to invention...it could be random chance, admittedly, but I'm left with a nagging feeling that the devs are doing some behind the scenes, ignore-the-man-behind-the-curtain tweaking of invention odds.

In any event, I'm going to launch a new round of testing after 1.4.1 to recheck the baseline.

Anfauglith
Retired Soldier Ltd.
Posted - 2007.04.03 01:12:00 - [130]
 

I do think some tweaking has been made, cause the sucess rate has dropped A LOT. Using decryptors I've even yet to get a single sucess. (The one adding more runs)

And without decryptors I've gotten 2x 10-runs out of 5 tries on warp disruptors II's and 1/4 on vagabonds.

Two step
Aperture Harmonics
K162
Posted - 2007.04.03 03:42:00 - [131]
 

I don't buy the nerf theory either. My success rate is still close to 75%, just where it was pre-patch. This is with about 75 runs before and another 75 after the patch.

One thing I have noticed is that the meta level of the item has a huge effect on your success rates. For runs where I have been poorer and bought a crappier named item, I am seeing success rates lower than 50%.

Plutoinum
Mercenaries of Andosia
Veritas Immortalis
Posted - 2007.04.03 03:47:00 - [132]
 

Edited by: Plutoinum on 03/04/2007 04:37:44
Originally by: Anfauglith
I do think some tweaking has been made, cause the sucess rate has dropped A LOT. Using decryptors I've even yet to get a single sucess. (The one adding more runs)

And without decryptors I've gotten 2x 10-runs out of 5 tries on warp disruptors II's and 1/4 on vagabonds.



Yes, I did 6 jobs yesterday, all failed.

( Of course, it can all be bad luck. But the chances to start a set of 6 jobs and see them all failing is low, if your success rate was like 2 out of 3 before. And if I count the two misses of the 4/2 result, it's already 8 misses in a row. )

Moonaru Izu
Caldari
Posted - 2007.04.03 07:47:00 - [133]
 

I've been following this thread for quit a while now. As I'm planning to get into invention soon (damn interfaces :) I thought I might as well prepare myself as good as I can and this thread delivers Cool

However, the recent "indications" that there have been some stealth nerfs got me thinking. Not hindered by any hands-on experience of inventions (and therefor surely no expert), I think that there were some nerfs, because it would make sense. Let me explain....

ATM, everbody, including the kitchen sink, is getting into invention it seems. A lot of them focussing on the modules and ships which are insanely overpriced, like hulks, cloaks and some others.

If all you inventors would suddenly be succesfull at inventing and building them and put them on the market, the prices would not only drop, they would probably crash. Invention is a very good initiative but it should not be that it makes the T2 stuff so common (and thus low-cost) that is will become mainstream, that was not the intention of CCP I recon.

So, I think that CCP has been monitoring the success rate of certain inventions and has start fiddling with the numbers of some of them to make sure that these items remain, kinda, exclusive and we do not end up with noobs flying T2 fitted ibis's.

This could explain why some of you get the feeling that the succesrate is not as high as it was and others do not experience this. My hunch is that the persons experiencing the stealth nerf are into inventing the more exclusive T2 stuff (like hulks).


Anfauglith
Retired Soldier Ltd.
Posted - 2007.04.03 09:17:00 - [134]
 

The idea was to get more producers getting into the T2 market Moonaru Izu. As it stands only the high-price mods/ships will give you any real profit.

With 1/4 sucess on a vagabond for instance you would actually EARN isk by buying a built a ship for 200m and selling your datacores.

With current invention and buildcost if you have a 100% sucess ratio the cost for a vaga is roughly between 80-90m.
(With Mech Eng cores @ 4m/each)

I'm not good at math, but 4 tries with one ship built came down to 230m. So the people with t2 bpo's have no need to worry since invention cost is always going to make the cost higher for inventors, not to mention the negative ME & PE.
Decryptors, from what I've seen, was not suppose to lower the % fo sucess and yet all tries failed with it, but managed 2/5 with warp disruptors without decryptors.
(The decryptor adding more runs that is, I'm aware of one decryptor having a x0.8 mod to sucess)

Khajit Smitty
Minmatar
MisFunk Inc.
Daisho Syndicate
Posted - 2007.04.03 10:02:00 - [135]
 

T2 BPO owners will always make profit, however it wont be the 300% markup sort of profit and more like a 25% to 50% markup.

anyways lets not loose sight of the objective here.

So far on sisi with all relvant invention skills at level 1 and trying to invent a module, i have the following stats to share :

Total jobs : 50
Total success's : 15
Total failure : 35

Job stats :
Item type : Module
No decrytor used at all.
Base T1 item used in all jobs.

Minmatar Encyption Methods : Lvl 1
Science skill 1 : lvl 1
Science skill 2 : lvl 1

More stats to come......




Khajit Smitty
Minmatar
MisFunk Inc.
Daisho Syndicate
Posted - 2007.04.03 10:02:00 - [136]
 

Edited by: Khajit Smitty on 03/04/2007 10:00:35
dble post, stupid forums

Ozzie Asrail
Caldari Provisions
Posted - 2007.04.03 12:24:00 - [137]
 

I havn't noticed any nerf at all, admittidly I've only been inventing on-off a week or 2 and only seroiusly this weekend. Don't really have massive amounts of runs done.

BCUII
10 Invents
Base Item - Standard BCU I
Decryptor: 1.3x success, +4 runs
8 Successes
1 Fail
1 Fail using the 0.4x, +9 runs.

Sensor Boosters II
2 Invents
Base Item - Standard SB I
Decryptor: 1.3x success, +4 runs
1 Success
1 Fail

Magstab II
2 Invents
Same as sensor booster

425mm Rail
5 Invents
425mm 'Scout' - 3rd best
1.3x, +4 decryptor
2 Sucesses
3 fails

I was a bit disapointed with the 425mm rail but the bcu was real successfull. Only got 3/3/3 for skills so possibly thats what hurt the 425's chance.

Overall - 19 invents - 12 Success - 7 Fail = 71.4% success rate and probably 250mil+ profit once everything sells. I'm happy enough tbh.

Plutoinum
Mercenaries of Andosia
Veritas Immortalis
Posted - 2007.04.03 15:36:00 - [138]
 

Well, just had a 5 out of 5 for overdrives without any decryptor. I have started another 5 invention runs. I'm really interested to see, how it goes on...

GayL0rd
Gallente
Gobshyte Inc.
Posted - 2007.04.04 10:30:00 - [139]
 

Is there any chance that the sucess rate is dynamic like the loyalty points and mission rewards that depend on average globals run time?

So, if there is alot of success on some mods for a week or so, the sucess rate for that particular mod drops. Would explaine why some people notice a dorp but others dont. And it insures that inventin doesnt cracs the market for some mods/ships. It would fit in with what CCP manage oter stuff like the missions.

Aristedes
Posted - 2007.04.04 11:36:00 - [140]
 

Edited by: Aristedes on 04/04/2007 11:36:24
some data

invention job no.1 (before revelations 1.4)
Skills: 4/3/3
BPC 300run ME0 PE0
item meta-level: 4
decryptor: none
Total success's : 13
Total failure : 7

invention job no.2 (before revelations 1.4)
Skills: 4/3/3
BPC 300run ME0 PE0
item meta-level: 4
decryptor: assembly instructions (used in 1 job only - finished with success)
Total success's : 14
Total failure : 6

invention job no.3 (after revelations 1.4)
Skills: 4/3/3
BPC 300run ME0 PE0
item meta-level: 4
decryptor: none
Total success's : 15
Total failure : 5

invention job no.4 (after revelations 1.4)
Skills: 4/4/3
BPC 300run ME0 PE0
item meta-level: 4
decryptor: none
Total success's : 16
Total failure : 4

invention job no.5 (after revelations 1.4)
Skills: 4/4/3
BPC 300run ME0 PE0
item meta-level: 4
decryptor: none
Total success's : 15
Total failure : 5

invention job no.6 (before revelations 1.4)
Skills: 4/4/4
BPC 15run stabber bpc ME0 PE0
item meta-level: 1? (stabber)
decryptor: assembly instructions (+4 run +30%success rate)
Total success's : 0
Total failure : 3

Pale NPastey
Posted - 2007.04.04 15:18:00 - [141]
 

The formulas you number crunchers out there come up with just make my head hurt. How can you guys go back and forth and assume you're correct with such limited information. Forgive me for trolling but the hardcore formulas presented here are a joke! Can't we agree that the majority of what we're all workng under (for the foreseeable future) is educated guesses and assumptions? I do (assume) think it's safe to assume that CCP will be making stealth adjustments to the sucess percentages of individual items for a while. 1) theres no way to predict the outcome until you cut it loose and see what happens, it can be so easy that they're as available and cheap as T1. 2)It seems that it'd be asking for headaches to announce every tweak to every item (cloak sucess percentages being dropped 2% today at downtime, something else increased 5%, etc. and see that daily) People would blow a gasket over every item and adjustment.
The successes/failures, attempts and general information is the only helpful stuff here. We can't come up with hard fomulas without a lot more time and information. Course....it would be nice if ccp would just release the percentages..
FYI-5 Hulk atempts, all using 10 run (max) bpc's with ME/PE 0, coveter as base item and 1.1 sucess decrypter
first 3 attempts, skill were 3/4/4 - All Failed
4th and 5th attempt had skills at 4/4/4-1 success. 2-run -4ME/PE

Daerkannon Shimmerscale
Paxton Industries
Paxton Federation
Posted - 2007.04.04 15:35:00 - [142]
 

Edited by: Daerkannon Shimmerscale on 04/04/2007 15:37:32
Originally by: Pale NPastey
The formulas you number crunchers out there come up with just make my head hurt. How can you guys go back and forth and assume you're correct with such limited information. Forgive me for trolling but the hardcore formulas presented here are a joke! Can't we agree that the majority of what we're all workng under (for the foreseeable future) is educated guesses and assumptions? I do (assume) think it's safe to assume that CCP will be making stealth adjustments to the sucess percentages of individual items for a while. 1) theres no way to predict the outcome until you cut it loose and see what happens, it can be so easy that they're as available and cheap as T1. 2)It seems that it'd be asking for headaches to announce every tweak to every item (cloak sucess percentages being dropped 2% today at downtime, something else increased 5%, etc. and see that daily) People would blow a gasket over every item and adjustment.
The successes/failures, attempts and general information is the only helpful stuff here. We can't come up with hard fomulas without a lot more time and information. Course....it would be nice if ccp would just release the percentages..
FYI-5 Hulk atempts, all using 10 run (max) bpc's with ME/PE 0, coveter as base item and 1.1 sucess decrypter
first 3 attempts, skill were 3/4/4 - All Failed
4th and 5th attempt had skills at 4/4/4-1 success. 2-run -4ME/PE


To answer your question, we make assumptions because that is part of the scientific process. Gather some data, make a theory and establish a possibility. Gather more data and see if that guess is right (it usually isn't). Refine your equation until you get something that accounts for 99% of all cases. Then you're probably very close to the actual result. Sometimes your assumptions prove to be wrong and you have to change those.

There is no evidence to suggest that CCP is tweaking invention secretly without telling the player base. Most of the samples presented here are, individually, far too statistically small to make any good observations from. Even with an 80% success rate failing 10 jobs in a row isn't that improbable.

Incidentally Newton's equations for physics are still being used today even though Einstein's equations superseded his because Newton's equations are good enough for 99% of what we need them for even if they have since been proven to be incorrect (and even Einstein's equations have faults).

Meau
Fatal Furballs
Posted - 2007.04.04 18:11:00 - [143]
 

Edited by: Meau on 04/04/2007 18:29:20

@Pale NPastey

Firstly i agree with Daerkannon.

Furthermore one formula seems to be fully(100%) understood, output run number.
Might not be the most valuable one(that one being the success chance) but its a success and can be helpful.

About formulae for success chance. You will probably never be able to calculate success changes for all modules, cause its sure there is a base chance in there and that one is hard to pin down and could chance over time. Still having an approximate table would be extremely helpful.

BUT what imho is most important at the current stage of this "scientific" endeavor is discovering the dependance of success on skills and meta-level.
As this should be the same for all items, there is hope, with enough data, and it would be very nice to have that knowledge.

DeODokktor
Caldari
Dark Templars
The Fonz Presidium
Posted - 2007.04.04 21:24:00 - [144]
 

Originally by: Ricdic
Yeh cloak has always seemed to be an exception to the T1 rule. Hell, we never even got the T1 bpo's seeded.


The tech1 proto cloak bpo's were seeded in like November.

Kaaii
Caldari
Kaaii-Net Research Labs
KAAII-NET
Posted - 2007.04.04 22:04:00 - [145]
 



Something ive been curious about...

In corp jobs, one player "starts" an invention process, his stats are checked against the job reqs. The job finishes sometime later.

Another player "finishes" (delivers) the job. This has no effect on the outcome right?? If player 1 has all skills, player 2 has none but delivers? Im thinking "probably" not, but I had an instance where my skills were lower than the starters(job) skills and when i delivered the job, it failed, bad luck i suppose but it got me thinking...

oh and for some numbers folks
Ive been playing with heavy launchers:
10 jobs-failed 4
skills rocket/mech/ency 4/4/4
caldari tunning instructions
84 runs total

Kaaii

Daerkannon Shimmerscale
Paxton Industries
Paxton Federation
Posted - 2007.04.04 22:11:00 - [146]
 

Originally by: Kaaii
Another player "finishes" (delivers) the job. This has no effect on the outcome right?? If player 1 has all skills, player 2 has none but delivers? Im thinking "probably" not, but I had an instance where my skills were lower than the starters(job) skills and when i delivered the job, it failed, bad luck i suppose but it got me thinking...


I would think that is safe to assume that whomever delivers the job has no effect on the outcome of the job since that's consistent with how the rest of the manufacturing discipline works. If I pop in a job at PE5 and later a corp mate with PE3 delivers it, it doesn't suddenly cost more minerals.

Meau
Fatal Furballs
Posted - 2007.04.05 11:06:00 - [147]
 

Originally by: Daerkannon Shimmerscale
Originally by: Kaaii
Another player "finishes" (delivers) the job. This has no effect on the outcome right?? If player 1 has all skills, player 2 has none but delivers? Im thinking "probably" not, but I had an instance where my skills were lower than the starters(job) skills and when i delivered the job, it failed, bad luck i suppose but it got me thinking...


I would think that is safe to assume that whomever delivers the job has no effect on the outcome of the job since that's consistent with how the rest of the manufacturing discipline works. If I pop in a job at PE5 and later a corp mate with PE3 delivers it, it doesn't suddenly cost more minerals.


/agree

The outcome of the job is probably determined in the moment you start it, it just doesnt tell you for obvious reasons.
Why do i think that? Cause which type of output BPC you get(when there are several options, i.e. with cloaks or rifters) is also determined in the moment of starting the job(and this, it tells you).

Mazhius
New Horizon Technologies
Posted - 2007.04.05 11:44:00 - [148]
 

Edited by: Mazhius on 05/04/2007 11:52:14
Quote:
succes_rate= (1+encryption.lev/10)*(1+related_science1.lev/10)*(1+related_science2.lev/10)*(1+(meta_item.lev*5+5)/10)*modyfier_decryptor


I suspect this is latest formula that we assumed is close to reality?

with skills at lvl 4, 1.3 decryptor and meta_item.lev4 it gives result of 12.4852(1.4*1.4*1.4*3.5*1.3) and this number just confuses me (I'm not the maths specialist though.. Smile )

shouldn't it be devided by 100, not 10 in each case? it would give result of 1.8279 (1.04*1.04*1.04*1.25*1.3), what cout be assumed a 82.79% succes chance?

Ofcourse it should also be multiplied by a 'base chance' - which I suspect is not a costant but variable figure. As someone previuosly spotted, this number could vary based on global attemts or succes figures. I can assume that it changes on each dt, summarising all activities of previuost DT to DT period (EDIT: or rather week - as per dt is too short to make reasonable influece)per item - in this way it would not only add randomnes to invention, but also would influece prices of t2 items - making them more stable due to higher cost per item for inventor (what is also should be quite important issue for dev and players).

Thus, if the above is close to truth, invention of less profitable items has a bigger chance of succes as less people are attempting it.

But I wont be surprissed if I'm totally wrong... Smile

Anfauglith
Retired Soldier Ltd.
Posted - 2007.04.05 12:27:00 - [149]
 

Vagabond invention:
Before "pressumed stealth-nerf"
4/6 sucesses, no decryptors.

All related skills at lvl 4.

After:
0/5 sucesses, no decryptors.
1/4 sucesses with decryptors.

All related skills at lvl 4.

Salvis Tallan
Gallente
Sparse Enterprises
Posted - 2007.04.05 13:30:00 - [150]
 

Originally by: Mazhius
Edited by: Mazhius on 05/04/2007 11:52:14
Quote:
succes_rate= (1+encryption.lev/10)*(1+related_science1.lev/10)*(1+related_science2.lev/10)*(1+(meta_item.lev*5+5)/10)*modyfier_decryptor


I suspect this is latest formula that we assumed is close to reality?

with skills at lvl 4, 1.3 decryptor and meta_item.lev4 it gives result of 12.4852(1.4*1.4*1.4*3.5*1.3) and this number just confuses me (I'm not the maths specialist though.. Smile )

shouldn't it be devided by 100, not 10 in each case? it would give result of 1.8279 (1.04*1.04*1.04*1.25*1.3), what cout be assumed a 82.79% succes chance?

Ofcourse it should also be multiplied by a 'base chance' - which I suspect is not a costant but variable figure. As someone previuosly spotted, this number could vary based on global attemts or succes figures. I can assume that it changes on each dt, summarising all activities of previuost DT to DT period (EDIT: or rather week - as per dt is too short to make reasonable influece)per item - in this way it would not only add randomnes to invention, but also would influece prices of t2 items - making them more stable due to higher cost per item for inventor (what is also should be quite important issue for dev and players).

Thus, if the above is close to truth, invention of less profitable items has a bigger chance of succes as less people are attempting it.

But I wont be surprissed if I'm totally wrong... Smile

I still believe this base success chance does not change each dt or anything like that(unless its a patch and done on purpose). The data files give something like 0.43 for a base chance, and then if we multiply it by 1.04*1.04*1.04*1.25*1.3 we get 0.84 which is far more likely and looks right.


Pages: first : previous : 1 2 3 4 [5] 6 7 8 9 ... : last (14)

This thread is older than 90 days and has been locked due to inactivity.


 


The new forums are live

Please adjust your bookmarks to https://forums.eveonline.com

These forums are archived and read-only