open All Channels
seplocked Market Discussions
blankseplocked Invention cost analysis, and the decryptor market (+ inv. XLS sheets).
 
This thread is older than 90 days and has been locked due to inactivity.


 
Pages: [1] 2 3 4

Author Topic

Akita T
Caldari Navy Volunteer Task Force
Posted - 2007.09.11 16:00:00 - [1]
 

Edited by: Akita T on 07/12/2007 05:32:11

By now, probably most of the faithful readers of this forum section will have at least tried their hands at invention, so I'll assume most of you have much more practical experience as I have.
If you haven't, in the next couple of posts you'll find a comprehensive breakdown of what you need, what you want, and what you get for it.

Being a relative newbie in invention, so take the "ASSUMED VALUES" with a grain of salt... or two, or three...
The only thing that doesn't change is the math attached to these posts, you only need to replace it with your own self-experienced values.
I am however experimenting right now massively on the test server, and so far even if I wasn't able to pinpoint exact values for invention chances (need a lot more "samples" at different skill levels and on various types of items), I was able to confirm the validity of "everything else".

One thing to note though, if your "sample size" is quite small (say, 5 tries at something with 10% chance), your results might not be of statistical significance.
Of course, if you are willing to contribute with your experienced values, please, go ahead, do so... no matter how much actual experience you had (just don't forget to mention that, though).

___

We're already at XLS v.10 (K.I.) / v.11 (A.T.) which take all issues talked about down below into account Wink
KI10 version is simplified scientific notation and saves display space.
AT11 version assumes you have a huge display resulution so you don't care if values are coma-separated, has purty kulorz and more comments !

Of course, it's still a work in progress Wink
___

Trinity (v.15) invention spreadsheet.

The correct me/pe/runs bonus values from now, also included freighters to inventables.
The "guesswork" chance to invent (several ships, especially the smaller ones, have had their invention chance boosted) is auto-linked to actual base chance to invent by default, change if you think you know better.

Default decryptor prices are intentionally left blank, as the different races' decryptor prices are also fluctuating wildly.
Fill those values in yourself manually for desired invention tasks.

Akita T
Caldari Navy Volunteer Task Force
Posted - 2007.09.11 16:01:00 - [2]
 

Edited by: Akita T on 02/10/2007 21:24:42

NEEDS

You first need a blueprint COPY of the T1 item you try to invent.
BPOs don't work (at least not yet).
The only thing that matters is number of runs on the BPC, nothing else does (e.g. ME/PE level).

For each inventable T1 item blueprint, in the "Invention" tab, you should (finally) have listed the mandatory skills and items you need to invent : one racial encryption methods skill and two science skills, then a number of datacores from those two science fields plus a data interface.
Only one from any kind of data interface (there's 12 of them, ship/module/rig for each race) is ever needed for any number of invention jobs you can run, as it's only required to start the invention process, but is not tied up into it nor damaged.
Datacores (and the optional decryptor and meta item) will be used up when you start the invention, and not returned in case it fails.

In case two T2 versions of a T1 item exist, one is chosen (seemingly at random) by the system, you have no control over it.
You can't use a T2 item as "optional meta item", and apparently no faction/officer ones either, as you need a meta level below 5.


UPDATE NOTE : Using a T1 item as "optional item" is at best pointless.
T1 items are "metalevel 0", so if they contribute anything at all to the invention chance, it's negligible (and therefore, not profitable).
That's especially true for SHIP invention, where all ships you can use are metalevel-0 AFAIK (not 100% sure on faction ships that are liste as variants, but they are definetely way too expensive to use).




CHANCES

The actual chances to obtain a T2 BPC are still unknown, but we do know what influences them.

We do know ships have a lower base invention chance as modules, for instance.
The three "L1 required" skills grant higher invention chances at higher levels, obviously, but it's not certain if it's a multiplicative or additive effect.
Also, using a higher meta-level (max 4) optional item will increase chance depending on meta level, and again it's not sure wether the effect is additive or multiplicative.
Each optional decryptor also has a chance multiplier, ranging from *0.4 to *1.3. That one is multiplicative for sure.
The decryptors also influence the ME/PE and number of runs on the obtained blueprint (more on that later).

It is uncertain wether number of runs on the T1 BPC influences invention chances, but it apprears to have little influence, if any.
For now, we can safely assume it does not matter at all in invention chances.


There's also a rumour going around that having L5 skills "is broken" with regards to invention (much like Tactical Shield Manipulation L5 was for a long time).

Personally, I can neither confirm nor deny this, as it is possible, however unlikely.
Normally, I would advise everybody to train all related skills to L5 (as the invention chance should be higher that way), but because of this particular issue, I'd say get them all to L4, and only train L5 at own risk, if you want.

UPDATE Confirmed by a dev that L5 skills work correctly. Apparently it is all just a fluke, a RNG hickup, odd unlucky streak, etc.

Akita T
Caldari Navy Volunteer Task Force
Posted - 2007.09.11 16:01:00 - [3]
 

Edited by: Akita T on 07/12/2007 20:20:45

RESULTS


Assuming you got a result, the obtained T2 BPC will ALWAYS be the same (ME/PE/runs) depending on used decryptor (if any) and number of runs on the T1 BPC.

You will always get at least 1 run on your T2 BPC, even if you used a 1-run T1 BPC and no decryptor.
Depending on item type, you might get more runs on the T2 BPC if you used a T1 BPC with more runs (ideally, a max-run one) even if you used no decryptor (exceptions for larger ships, which might still get only 1 run).
The number of "bonus runs" from the decryptor are normally added to the number of runs obtained (exceptions again for things you would not get more than 1 run without using a decryptor, in that case you only get the bonus runs added if you have a max-runs T1 BPO, in other cases the final runs equal the bonus runs on the decryptor).
Well, it sounds more complicated as it is.

There are 5 types of decryptors for each race (I will refer to them by "invention chance multiplier"), so including the "no decryptor" option, you can have the following 6 results (ordered by invention chance):


UPDATED FOR TRINITY (updated on Dec 7th 2007):
Decryptor1 : x1.8 chance, -5 ME, -2 PE, +4 runs
Decryptor2 : x1.2 chance, -2 ME, +1 PE, +1 runs
Decryptor3 : x1.1 chance, -1 ME, -1 PE, no bonus runs
no decrypt : x1.0 chance, -4 ME, -4 PE, no bonus runs
Decryptor4 : x1.0 chance, -3 ME , 0 PE, +2 runs
Decryptor5 : x0.6 chance, -6 ME, -3 PE, +9 runs






So which decryptor is which ?

type : Caldari, Gallente, Minmatar, Amarr

Decryptor1 : Installation Guide, Stolen Formulas, Assembly Instructions, War Strategon
Decryptor2 : Prototype Diagram, Test Reports, Advanced Theories, Classic Doctrine
Decryptor3 : Tuning Instructions, Collision Measurements, Calibration Data, Formation Layout
no decrypt : duuh
Decryptor4 : User Manual, Engagement Plan, Operation Handbook, Sacred Manifesto
Decryptor5 : Alignment Chart, Symbiotic Figures, Circuitry Schematics, Circular Logic


What's up with all the negative ME/PE things ?


Note that normally, ME 0 would mean 10% waste, ME -4 would mean 50% waste, ME -5 a 60% waste and so on down to ME -8 meaning 90% waste... but for a good deal of time the waste/ME formula was (?still is?) "broken" and ME -4 means 30% waste up to ME -8 meaning 70% waste.

For PE, it's less weird, each negative level seems to increase the manufacture time by an additional 20% of base manufacture time, but starting with *1.4 manufacture time at PE -1 and up to *2.0 at PE -4.


Basically, by using NO decryptor (-4 ME/PE), compared to owning a "high ME/PE T2 BPO", you should manufacture things out of your inventyed BPCs roughly two times slower and at 50% more cost of components ("bugged" version meaning double time and only +30% component cost).
Using any decryptor (except "decryptor4") will make you manufacture items faster, but in most cases at increased component cost.

Akita T
Caldari Navy Volunteer Task Force
Posted - 2007.09.11 16:02:00 - [4]
 

But what does this all mean, and what's it got to do with the market ?!?

Ah, I'm glad you asked.
See, all decryptors are used in ANY invention jobs related to a certain "racial" encryption skill, no matter how big or small.
So, their price should wildly fluctuate according to demand in various sections of invention... yet they don't fluctuate as much as you'd expect.

The reason could either be that
a) inventors seem to be unable to grasp basic math and are unable to choose the "proper" decryptors for their jobs (and yes, in some cases, no decryptor is better as some decryptors)
or
b) that resellers don't realize the value of decryptors and only "play it by ear", manipulating the market the "wrong way" for their own good

Either way, what THIS thread hopes to accomplis is make it clear for both parties which ones are worth using and which ones aren't, depending on the situation... and hopefully, equalize T2 prices even further in the process by streamlining inventor operation for minimal investment.



Any advice ?

I would persuade small-time inventors to stay out of the bussiness (unless it's for personal use in remote areas of the galaxy) for the time being, because they tend to ignore the long-term (average) costs, and usually undersell their "lucky chance" products, only to hit a bad bump when the unlucky streak starts. You're really better off just buying off the market most of the time.

Due to the chance-based nature of invention, I would recommend that you ONLY go into invention at all if your skills are good (L4 each of the 3 relevant skills) and you have a starter capital that will cover (on average) at least two, preferably three succesfull invention jobs per "investment batch".






So how does decryptor/datacore price translate into manufactured goods (T2) prices ?


Given enough invention runs for things to naturally average out, the "cost" formula per invented run is rather easy to figure out (and so is the manufactured end-product value), but the assumed BASE invention chance values are the killer.

C_base = base chance to invent, depending on skills
C_modItem = optional item meta-level modifier

Now, wether "C_modItem" is an additive or a multiplicative value depending on meta-level really doesn't matter for a certain specific invention job and a specific metalevel of item used, because the C_base value is constant.
Example : 25% base chance and *1.2 multiplier is the same as 25% base chance and a +5% modifier... both work out as 30% chance.

So, we'll just use
Cx = chance to invent, depending on skills and optional item of metalevel "x"


Cost_per_try = Decryptor_cost + BPC_cost + datacores_cost + MetalevelX_item_cost + lab_time_cost (usually negligible)
Chance = Cx * decryptor_chancemod
Cost_per_run = Cost_per_try / (Chance * (ExpectedRuns + decryptor_runsmod) )


In an ideal situation, a T2 manufacturer's production lines would ALWAYS be busy, and therefore the PE level of the obtained BPC would also matter, especially since you also have to factor in factory costs.
For mass-producers, it does matter... however, the market is saturated with small-time inventor/manufacturers for which production time is not that relevant (since their "production lines" are mostly empty to begin with).
So, sadly, PE level on obtained T2 BPCs is of little practical value, if any.

Therefore, in the pricing of the end-product, other than the "cost per run" of invention, only ME level is of any practical value.


Assuming identical situations EXCEPT the decryptor (best metalevel optional item used, same maxed-out skills), and also assuming maxruns on the T1 BPC we get:

Cost_per_run = Cost_per_try / (Cx * decryptor_chancemod * (maxruns + decryptor_runsmod))
Cost_per_item = T1_item_cost + Cost_per_run + T2_materials_cost*(1+(1-ME)/10))

Akita T
Caldari Navy Volunteer Task Force
Posted - 2007.09.11 16:06:00 - [5]
 

And what's that depending on decryptor ?



Decryptor1 : Installation Guide, Stolen Formulas, Assembly Instructions, War Strategon
Cost_per_run = Cost_per_try / (1.3*Cx * (maxruns+4))
T2cost multiplier is +60% (if fixed, +80%)
Production time is base *1.6


Decryptor2 : Prototype Diagram, Test Reports, Advanced Theories, Classic Doctrine
Cost_per_run = Cost_per_try / (1.1*Cx * (maxruns+1))
T2cost multiplier is +50% (if fixed, +70%)
Production time is base *0.9


Decryptor3 : Tuning Instructions, Collision Measurements, Calibration Data, Formation Layout
Cost_per_run = Cost_per_try / (Cx * maxruns)
T2cost multiplier is +40% (if fixed, +60%)
Production time is base *1.4


No decryptor :
Cost_per_run = Cost_per_try / (Cx * maxruns)
T2cost multiplier is +30% (if fixed, +50%)
Production time is base *2.0


Decryptor4 : User Manual, Engagement Plan, Operation Handbook, Sacred Manifesto
Cost_per_run = Cost_per_try / (0.8*Cx * (maxruns+2))
T2cost multiplier is +30% (if fixed, +50%)
Production time is base *1.0


Decryptor5 : Alignment Chart, Symbiotic Figures, Circuitry Schematics, Circular Logic
x0.4 chance, -8 ME, -3 PE, +9 runs
Cost_per_run = Cost_per_try / (0.4*Cx * maxruns)
T2cost multiplier is +70% (if fixed, +90%)
Production time is base *1.8






Examples ! Numbers ! You only gave me empty talk until now !

Very well then... but be warned, the ACTUAL results vary a lot depending on actual base chances for invention (and on wether the ME/wastage formula was fixed or not).

We will assume max-skills to be used in the following scenarios, so an arbitrary high chance of invention... let's say 60% for modules (or 75% with metalevel4 item) and 20% for ships (or 25% with a T1 ship used as optional item, metalevel 1).

We will assume datacore cost to be around 3.75 mil for the module (most modules), and 30 mil for the ship (tier1 battlecruiser).

We will further assume a 1-run module copy is negligible in cost, but around 250k for a max-run.
For the ship, a 1-run copy at around 200k, and a max-run (15) at around 3 mil.

The "metalevel4" item let's say at around 2 mil ISK, and a T1 ship (tier1 BC) at a cheap 25 mil too.

We will also arbitrarily assume we can get up to 20 maxruns without a decryptor (please correct me if that is not so and you know the exact value) for the module invention, but only 1 maxrun for the ship invention.



I can only repeat this :
ALL VALUES BELOW CHANGES IF VALUES ASSUMED ARE NOT THOSE LISTED.
So, if you are aware any of the "assumed values" below are different, you have to redo the math.


I know, it's a lot of assumptions, and the math might be hard to follow in places, but such is the deal.

Akita T
Caldari Navy Volunteer Task Force
Posted - 2007.09.11 16:07:00 - [6]
 

MODULE INVENTION

* no decryptor, no metaitem, 1-run T1 BPC
get 1-run T2 BPC 60% of times for 3.75 mil
cost per run = 6.25 mil (bad)

* no decryptor, no metaitem, maxrun T1 BPC
get 20-run T2 BPC 60% of times for 4 mil
cost per run = 333k


* no decryptor, meta4item, maxrun T1 BPC
get 20-run T2 BPC 75% of times for 6 mil
cost per run = 400k

* Decryptor1, no metaitem, 1-run BPC
get 4-run T2 BPC 78% of times for 3.75 mil + dec1_cost
cost per run = 1.2 mil + (dec1_cost*0.32)

* Decryptor1, no metaitem, maxrun BPC
get 24-run T2 BPC 78% of times for 4 mil + dec1_cost
cost per run = 214k + (dec1_cost*0.042)


* Decryptor1, meta4item, maxrun BPC
get 24-run T2 BPC 97.5% of times for 6 mil + dec1_cost
cost per run = 256k + (dec1_cost*0.042)

You notice how the cost is much lower with a decryptor, if the decryptor WOULD be "free".
It depends a lot on added chance of metaitem wether it's even feasable to use it at all.
Also, in this particular case, if the decryptor goes over (roughly) 3 mil, you're better off without one, as cost-per-run goes (besides, you're also better off as ME/PE anyway).
Yes, better off without using one, *if* those are the numbers (chances, prices, etc).


* Decryptor2, no metaitem, maxrun T1 BPC
get 21-run T2 BPC 66% of times for 4 mil + dec2_cost
cost per run = 288k + (dec2_cost*0.072)

* Decryptor3, no metaitem, maxrun T1 BPC
get 20-run T2 BPC 60% of times for 4 mil + dec3_cost
cost per run = 333k + (dec3_cost*0.083)

* Decryptor4, no metaitem, maxrun T1 BPC
get 22-run T2 BPC 48% of times for 4 mil + dec4_cost
cost per run = 378k + (dec4_cost*0.094)

* Decryptor5, no metaitem, maxrun T1 BPC
get 29-run T2 BPC 24% of times for 4 mil + dec5_cost
cost per run = 574k + (dec5_cost*0.143)


You can see here, the +9 run decryptor is probably the WORST possible ideea, as you actually lose money compared to "no decryptor"... not only is the cost per run higher even with a "free decryptor", but also you get the worst ME and PE possible.

The +2 runs ("decryptor4") is only feasable if you NEED churn out huge numbers of modules with very drastic T2 component costs compared to markup, but even here the "no decryptor" option might be more feasable... yes, it takes twice as long to manufacture them, but you manufacture them at the same material cost and much lower overall cost... so it really depends on the amount you can move.



Final module invention conclusions ?

* Using expensive metaitems is a no-go.
* NEVER use "Decryptor5", under any circumstances.
* If you can get "Decryptor1" cheap and the extra wastage doesn't matter to you, use it.
* In most cases, not using any decryptor at all is best, both as manufacturing costs and invention costs.

Akita T
Caldari Navy Volunteer Task Force
Posted - 2007.09.11 16:08:00 - [7]
 

SHIP INVENTION


* no decryptor, no metaitem, 1-run T1 BPC
get 1-run T2 BPC 20% of times for 30.2 mil
cost per run = 151 mil (bad)

* no decryptor, T1 ship, maxrun T1 BPC
get 1-run T2 BPC 25% of times for 58 mil
cost per run = 232 mil (the worst possible, I hope)



* Decryptor1, no metaitem, 1-run T1 BPC
get 4-run T2 BPC 26% of times for 30.2 mil + dec1_cost
cost per run = 29 mil + (dec1_cost*0.961)

* Decryptor1, T1 ship, 1-run T1 BPC
get 4-run T2 BPC 32.5% of times for 55.2 mil + dec1_cost
cost per run = 42 mil + (dec1_cost*0.769)

* Decryptor1, no metaitem, maxrun T1 BPC
get 5-run T2 BPC 26% of times for 33 mil + dec1_cost
cost per run = 25 mil + (dec1_cost*0.769)

* Decryptor1, T1 ship, maxrun T1 BPC
get 5-run T2 BPC 32.5% of times for 58 mil + dec1_cost
cost per run = 35 mil + (dec1_cost*0.615)


We notice that the combo with a non-maxruns BPC, a ship and a decryptor1 can't possibly be best, ever (maxrun, no ship is always better).
Depending on dec1_cost however, EITHER one of the other three variants might be better.

Granted, the version with a T1 ship and maxruns T1 BPC used is only more viable if the decryptor cost goes OVER 65 mil.ISK, but it's still a feasable option.
The option without ship and with a minimum-run BPC is only feasable if the decryptor is under 21 mil ISK.
For any values of the decryptor in between 21 and 65 mil, the no-ship, max-run BPC option is the most feasable.

It's also quite obvious that the results of a "no decryptor" invention job would be pointless, unless for some reason "Decryptor1" price goes over 188 mil ISK.
Also, I hope I don't have to explain why "Decryptor1" is the BEST possible for ship invention.

Let's take next the other decryptor that seems to be quite popular with budding inventors, namely the +9 runs "Decryptor5".


* Decryptor5, no metaitem, 1-run T1 BPC
get 9-run T2 BPC 8% of times for 30.2 mil + dec5_cost
cost per run = 42 mil + (dec5_cost*1.388)

* Decryptor5, T1 ship, 1-run T1 BPC
get 9-run T2 BPC 10% of times for 55.2 mil + dec5_cost
cost per run = 61 mil + (dec5_cost*1.111)

* Decryptor5, no metaitem, maxrun T1 BPC
get 10-run T2 BPC 8% of times for 33 mil + dec5_cost
cost per run = 41 mil + (dec5_cost*1.25)

* Decryptor5, T1 ship, maxrun T1 BPC
get 10-run T2 BPC 10% of times for 58 mil + dec5_cost
cost per run = 58 mil + dec5_cost



We notice that the combo with a non-maxruns BPC, NO ship and a decryptor5 can't possibly be best, ever (no ship, maxruns BPC is always better).
Depending on dec1_cost, EITHER one of the other three variants might be better... but even from the first glance, you already notice how the base costs for obtained invention run are significantly worse as the ones obtained with "Decryptor1".
However, they are still a lot better as the ones obtainable without a decryptor at all.

In order for "Decryptor5" to yield better OVERALL results as with "Decryptor1", you'd have to have very high prices for both decryptors, AND you'd have to have Decryptor1 cost 59 mil ISK or more extra compared to a Decryptor5 (but not over 188 mil, because then "no decryptor" invention becomes more profitable).
Needless to say, that's highly unlikely.

Akita T
Caldari Navy Volunteer Task Force
Posted - 2007.09.11 16:09:00 - [8]
 


FINAL CONCLUSIONS


If you are inventing modules, don't bother with any decryptors at all unless they're practically free, use cheap metaitems, and always use max-run BPCs.

If you are inventing ships, always use "Decryptor1" if possible.
If not affordable, then "Decryptor5".
Only use "Decryptor2" or "Decryptor4" as a last resort, and only if cheap.
Don't EVER bother at all with "Decryptor3" unless all others are insanely expensive, and even then if you're more interested in getting your ships faster rather than cheaper... it's cheaper to invent without a decryptor, and without one, it's expensive as hell anyway.

In a perfect market, Decryptor1 would be the most expensive, Decryptor5 at around half the price or even less (both used exclusively in ship invention), and the rest of the decryptors would be worth next to nothing (used exclusively in module invention, if at all).

Akita T
Caldari Navy Volunteer Task Force
Posted - 2007.09.11 16:12:00 - [9]
 

There, posting complete, you can start bashing away Wink
Also, space reserved for future possible usage

Kazuo Ishiguro
House of Marbles
Posted - 2007.09.11 16:41:00 - [10]
 

You really need a spreadsheet to deal with all the calculations. I made one some time ago that I've been using to do rough estimates of expected profitability:

http://dl.eve-files.com/media/corp/KzIg/invention-2.zip

There are some important exceptions to the principles in your conclusion, particularly Cloaks and modules that are just entering the market. It takes 9 days to produce a max-run cloak BPC, so although the profit per job is higher with max-run BPCs (and decrytor 1 is best), the profit per hour is better with low run BPCs (and decryptor 5).

With the recent addition of the new mining laser & ice harvester upgrades, there was a brief opportunity to sell at a very high profit. However, each invention cycle had a base time of 32 hours, and it takes a significant amount of time to build the upgrades - over 2 hours each iirc. In those circumstances, decryptor 2 (as well as being very cheap) was the best one to use.

Akita T
Caldari Navy Volunteer Task Force
Posted - 2007.09.11 17:12:00 - [11]
 

Nice observations, and nifty spreadsheet (once you get used to it and reformat a few fields).

Sir Scorpion
Black Banners
Posted - 2007.09.11 17:28:00 - [12]
 

wow looks intresting will read it and comment on it asap

Benvie
Benvie Enterprises
Posted - 2007.09.11 18:44:00 - [13]
 

Couple things. First, isn't the maxruns on modules 10 without a decryptor? Second, you say that it's best to use max run ship bpcs but that's using the assumption that bpcs have no cost. The copy time on ships is usually extensive and I believe in most cases a 1 run ship bpc is the much better option, even if you get 1 less run on the final product. Copy slots are not infinite and neither are character lab operation slots.

Otherwise this is pretty spot on. Not that I want people to learn how to invent better though. =D Most people do stupid things that increase their invention cost by a large margin, meaning I make a much higher profit when I go to sell my invented wares.

Akita T
Caldari Navy Volunteer Task Force
Posted - 2007.09.11 19:38:00 - [14]
 

Edited by: Akita T on 11/09/2007 19:45:05

Originally by: Benvie
First, isn't the maxruns on modules 10 without a decryptor?

Quite probable. That's why I listed the 20 as an assumption, not a certainty.
And I also said right afterwards "please correct me if that is not so and you know the exact value". Wink

Originally by: Benvie
Second, you say that it's best to use max run ship bpcs

No, I said in *module* invention, you should always use a max-runs blueprint.

Quote:
but that's using the assumption that bpcs have no cost. The copy time on ships is usually extensive and I believe in most cases a 1 run ship bpc is the much better option, even if you get 1 less run on the final product. Copy slots are not infinite and neither are character lab operation slots.


I have listed BPC costs too under "assumptions".

Namely, "negligible" for 1-run module, 250k for max-run module, and in the case of a BC invention, 200k per single run ship and 3 mil per max(15)-run ship.
Note that the costs might not be an accurate value of actual 0 ME, 0 PE, T1 BPCs... but the math is the same, you just have to adjust the prices to reflect the actual value of the BPC.


Yes, in some cases, it might be better to use a 1-run BPC instead of a 15-run BPC, but that depends on actual invention base chance and decryptor price mostly (as shown above).


As for the char lab slots... well, because the BPO's ME/PE levels you use to create the T1 BPCs have absolutely no influence over the output BPC, and because it's T1 blueprints we're talking about here... you can pretty much create an alt char capable of running 6 copy jobs "out of the box", if only you have enough T1 BPOs available.

The other problem would be with installation slots, but a small empire POS is able to sustain 3 mobile labs for minimal monthly cost, and that's plenty of copy slots right there (even with speed bonuses compared to normal installation slots).

Quote:
Otherwise this is pretty spot on.
Not that I want people to learn how to invent better though. =D
Most people do stupid things that increase their invention cost by a large margin, meaning I make a much higher profit when I go to sell my invented wares.


Oh, I wouldn't worry.
Most people don't even bother reading this forum in the first place.
And even if they did, they probably couldn't calculate things right, and still make big mistakes.

tinuvin
Posted - 2007.09.11 19:44:00 - [15]
 

Two questions:

1. Are you talking about 10 run BPC's for "max run BPC's"?

2. Do the same calculations apply to rigs as for modules?

Akita T
Caldari Navy Volunteer Task Force
Posted - 2007.09.11 19:54:00 - [16]
 

Originally by: tinuvin
Two questions:
1. Are you talking about 10 run BPC's for "max run BPC's"?
2. Do the same calculations apply to rigs as for modules?


For T1 BPCs, the "max runs" listed is the absolute maximum number of runs you can have on a blueprint copy.
Rigs normally have 1000 maxruns, modules around 300 maxruns, frigates 30 runs, battleships 10 runs.

For T2 BPCs, the "max runs" refers to the number of runs you get when you try to invent from a T1 BPC with maxruns on it, without using a decryptor (or using the "Decryptor3", the one without a number of runs bonus).
Apparently, T2 modules mostly have 10 runs, but I've heard versions with 20, so it might depend on the module... as for T2 ship BPCs, there you always have 1 as "maxrun" value (as explained above).


Yes, same calculations apply for all inventable items, be it modules, drones, ammo, rigs or ships.
The only variables are base invention chances and the "T2 maxrun" value.

Benvie
Benvie Enterprises
Posted - 2007.09.11 20:03:00 - [17]
 

I guess the main thing about ship bpcs is that the prices vary wildly depending on the size of the ship. At the lower end, at frigate or cruiser level, it's probably worth it to do max runs. At the top end, with things like a covetor, the bpc cost is usually 1.5 mil+ and the copy time is like 3 hours per run at a POS. With max runs on a covetor being 10 you're talking 15 mil for the bpc which does end up being a significant cost. I imagine this will be similar with tech2 battleships.

Shadarle
Posted - 2007.09.11 20:28:00 - [18]
 

Wow, very interesting and useful post. I knew absolutely nothing about invention before. I probably still won't bother to try it... but if I do this thread is going to be my starting point Very Happy

Kazuo Ishiguro
House of Marbles
Posted - 2007.09.11 21:21:00 - [19]
 

Originally by: Shadarle
Wow, very interesting and useful post. I knew absolutely nothing about invention before. I probably still won't bother to try it... but if I do this thread is going to be my starting point Very Happy


You've missed the boat - or at least most of the boats. The only time an enormous profit was ever made via inventing most items was while prices were crashing from the levels previously set by the BPO cartels. Since then, the market has balanced out to the point where you might as well stick to T1 production.

These days, you're doing extremely well if you can rake in more than 150m of profit per day per character. Many people settle for a lot less.

Two step
Aperture Harmonics
K162
Posted - 2007.09.11 22:13:00 - [20]
 

You are making way too many assumptions to be able to say what decryptor is better for ship invention. You are also using a single ship and datacore cost to draw your conclusions, which is not really helpful for all types of ship invention.

If you really want to write a guide for this, you should run some experiments on sisi or get some real numbers from someone that has run a lot of invention jobs.

Shadarle
Posted - 2007.09.11 22:17:00 - [21]
 

Originally by: Kazuo Ishiguro
Originally by: Shadarle
Wow, very interesting and useful post. I knew absolutely nothing about invention before. I probably still won't bother to try it... but if I do this thread is going to be my starting point Very Happy


You've missed the boat - or at least most of the boats. The only time an enormous profit was ever made via inventing most items was while prices were crashing from the levels previously set by the BPO cartels. Since then, the market has balanced out to the point where you might as well stick to T1 production.

These days, you're doing extremely well if you can rake in more than 150m of profit per day per character. Many people settle for a lot less.



T1 production is worthless in most cases.

Invention is worthless now, it won't be when a whole boatload of new t2 items are added to the game in the future. It also will constantly be worth it depending on the prices of T2 items and invention components. You just have to find the right items. If invention wasn't worth doing no one would do it. It's like T2 BPO owners saying that T2 BPO's aren't worth buying anymore since you can't make a 2000% markup, you're stuck with a mere 300% markup.

I don't see why 150m/day is any sort of amazing feat. It has become easier to make money in the past months from my perspective. But then I've never been a big T2 owner... those who are would find things harder I suppose, but I find things far easier.


Akita T
Caldari Navy Volunteer Task Force
Posted - 2007.09.11 23:28:00 - [22]
 

Edited by: Akita T on 11/09/2007 23:41:23

Originally by: Two step
You are making way too many assumptions to be able to say what decryptor is better for ship invention.

All my assumptions change is the ISK cost for which a different approach towards the same decryptor is more viable in ship manufacture.

Quote:
You are also using a single ship and datacore cost to draw your conclusions, which is not really helpful for all types of ship invention.

The higher you go in unit price, the more effective using the higher-chance decryptor becomes... so only cruiser and above get decent returns, for now.
Even given the heavily-invented Hulk ("cruiser-sized") and the corresponding decryptors, looking at the pricetags (which will be representative of how costs for all decryptors will line up to) you can safely conclude "Decryptor1" is best for mass production there followed not very close by "Decryptor5", and that goes double for higher classes.

Quote:
If you really want to write a guide for this, you should run some experiments on sisi or get some real numbers from someone that has run a lot of invention jobs.

No, it wasn't my intention to make a guide, although it came out sounding like one.
Would it have been an attempt at a guide, it would have been titled accordingly, say, for instance, "Guide&FAQ: Invention".


What this thread is, is an explanation on how to figure out your actual invention costs, how to properly determine the "base invention chance" based on your personal experience, and last but not least on how to calculate which decryptor and which approach (no item, cheap item, best meta item... 1-run BPC, in-between or max-run BPC) is the most profitable.

It's also a thread suggesting/recommending decryptor prices, or rather said, quick ways of determining which are worth shortselling and which stocking up considering the approaching T2 ship roll-in (ships which will ALL be "invention-only").


So, bottom line, it's not a guide nor an instruction manual spoon-feeding you actual in-game data, but rather a mostly-theoretical "class" which teaches you how to calculate for yourself according to your own experience and current in-game status.


Originally by: Shadarle
T1 production is worthless in most cases.
Invention is worthless now, it won't be when a whole boatload of new t2 items are added to the game in the future. It also will constantly be worth it depending on the prices of T2 items and invention components. You just have to find the right items. If invention wasn't worth doing no one would do it.


T1 production is only "worthless" because it has too little variables (7, actually, and inter-connected), has hard-coded limits both ways, on the upper and lower boundary (NPC goods reprocessing and platinum insurance payouts).
Due to this fact, and the very low (skill-wise) barrier of entry, it's insanely competitive, and there's just too many people who fail to take into account their own costs ("minerals I mine myself are free", sounds familiar?) and they flood the market with way-too-cheap goods.


With T2 invention/manufacture, the skillpoint entry barrier is much higher, the needed capital reserves to "smooth bumps over" are also insanely higher for consumables alone, and don't even get me started on the ISK cost of skills (sunk cost) and interfaces (recoverable, but still tied-up capital).
Also, the volatility of each individual product is quite high, and those who can predict where an unfulfilled demand will show up and "invent" accordingly stand to gain a lot more as the rest.

You will unavoidably get the "datacores I research myself are free" types here too, and that can be a problem, especially for ship invention/manufacture. However, if you know when to trade or invent, you can capitalize faster on other people's ISK.
Not only that, but even as a pure T2 trader, knowing how exactly T2 prices will shift by following the datacore and decryptor market is a valuable enough tool.

Benvie
Benvie Enterprises
Posted - 2007.09.12 00:34:00 - [23]
 

Originally by: Kazuo Ishiguro
You really need a spreadsheet to deal with all the calculations. I made one some time ago that I've been using to do rough estimates of expected profitability:

http://dl.eve-files.com/media/corp/KzIg/invention-2.zip

There are some important exceptions to the principles in your conclusion, particularly Cloaks and modules that are just entering the market. It takes 9 days to produce a max-run cloak BPC, so although the profit per job is higher with max-run BPCs (and decrytor 1 is best), the profit per hour is better with low run BPCs (and decryptor 5).

With the recent addition of the new mining laser & ice harvester upgrades, there was a brief opportunity to sell at a very high profit. However, each invention cycle had a base time of 32 hours, and it takes a significant amount of time to build the upgrades - over 2 hours each iirc. In those circumstances, decryptor 2 (as well as being very cheap) was the best one to use.


Very nice spreadsheet. I've been playing with it quite a bit. I noticed you have ship base chance at 15% and module base chance at 35% and was wondering why that is. Most sources I've read put them at 20% and 40%. That 5% makes a large difference in the expected profitability.

Benvie
Benvie Enterprises
Posted - 2007.09.12 01:39:00 - [24]
 

Also for the spreadsheet, I'm kind of confused. You figure out the total profit per invention attempt and then figure out the profit per unit from that. The formula I had come up with does it sort of the other way around, it figures out the average invention cost per unit and substract that from the sell price (along with mat cost). The numbers are very very different.

The formula you appear to be using to calculate per unit profit is:
(Chance * DecryptorMultiplier * Runs * (SellPrice - BuildCost) - DecryptorCost - InventionCosts) / Runs

The formula I used was:
SellPrice - BuildPrice - ((1 / (Chance * DecryptorMultiplier)) * (DecryptorCost + InventionCosts)) / Runs

Chance is your combined chance before decryptor multipler, that includes base chance and skills and meta level of input item. In your formula you use Runs twice. So the question is, which equation is correct?

Kazuo Ishiguro
House of Marbles
Posted - 2007.09.12 09:44:00 - [25]
 

Edited by: Kazuo Ishiguro on 12/09/2007 09:59:51
Since the two formulas are different and only one can be correct, I'll attempt to demonstate that yours is wrong by showing that mine is right Razz.

All I've done is to add up probability * consequence for all possible outcomes of each attempt; this, by definition, is the expected value of the profit per invention cycle. I've then divided this by the # of runs from each successful cycle.

The # of runs features twice because the costs of failed invention attempts (which produce no runs) have to be subtracted from the profit per run of the successful ones. If you were only interested in the profit per run of successful jobs, the number of runs would cancel.

Regarding the base invention chance estimates, I take those with a pinch of salt. I usually assume a 40% success rate when inventing modules (I might as well be pessimistic when long streaks of failures are quite common).

Originally by: Shadarle
I don't see why 150m/day is any sort of amazing feat. It has become easier to make money in the past months from my perspective. But then I've never been a big T2 owner... those who are would find things harder I suppose, but I find things far easier.


I meant via invention alone; I'm sure there are much more profitable acitivities for those so inclined. Cool

Quote:
Invention is worthless now, it won't be when a whole boatload of new t2 items are added to the game in the future. It also will constantly be worth it depending on the prices of T2 items and invention components. You just have to find the right items. If invention wasn't worth doing no one would do it. It's like T2 BPO owners saying that T2 BPO's aren't worth buying anymore since you can't make a 2000% markup, you're stuck with a mere 300% markup.


The market for most T2 modules has reached a state resembling simple harmonic motion near an equilibrium point determined largely by datacore prices. Some items are occasionally profitable for a while, but then lots of inventors jump on them and saturate the market, and everyone forgets about them until prices rise (or datacore prices fall) and the cycle repeats.

New items are only really profitable for a short, transient period while the market settles. I'm going to try my hand at a few of the new ships for a while, but I don't think they'll do well in the long run.


Akita T
Caldari Navy Volunteer Task Force
Posted - 2007.09.12 12:17:00 - [26]
 

Originally by: Benvie
(Chance * DecryptorMultiplier * Runs * (SellPrice - BuildCost) - DecryptorCost - InventionCosts) / Runs
[...]vs[...]
SellPrice - BuildCost - ((1 / (Chance * DecryptorMultiplier)) * (DecryptorCost + InventionCosts)) / Runs

(C*DM*R(SP-BC)-DC-IC)/R
=C*DM*(SP-BC) - (DC+IC)/R
=[SP-BC - (1/(C*DM))*(DC+IC)/R] * C*DM
The two formulae are nearly identical, short of the *C*DM part at the end.

As for which one is correct... well, I already DO have a "cost per run" formula listed several posts above, so all you have to do is "integrate" it into profit as simple "extra build cost", no ? Twisted Evil

Cost_per_run = Cost_per_try / (Cx * decryptor_chancemod * (maxruns + decryptor_runsmod))
using your above notation, it turns into
Cost_per_run = (DecryptorCost + InventionCosts) / (Chance * DecryptorMultiplier * Runs)
This is obviously correct, as it's total costs / average obtained copies.

So, profit PER UNIT is

PPU = SellPrice - BuildCost - cost_per_run
= SellPrice - BuildCost - [(DecryptorCost + InventionCosts) / (Chance * DecryptorMultiplier * Runs)]

Which would mean that the second listed formula is right (even if written a bit weird).

Kazuo Ishiguro
House of Marbles
Posted - 2007.09.12 16:36:00 - [27]
 

Edited by: Kazuo Ishiguro on 12/09/2007 16:36:20
Ok, so I was nearly right. As I was dealing with the average profit, I should have divided by the average number of runs achieved, not the number achieved with a successful job.

I wasn't using that column very much - I tended to look at the profitability of the whole job, rather than individual runs. So invention is not any more profitable overall than I thought it was - the predictions made for whole jobs have turned out to be quite accurate. As you'd expect, in fact Laughing

Akita T
Caldari Navy Volunteer Task Force
Posted - 2007.09.12 16:59:00 - [28]
 

You were merely underestimating the profitability by a factor of actual chance to invent Twisted Evil
Doesn't change the fact some jobs would be profitable or would result in losses, it merely distorted the amount of profits/losses.

So, if the "per unit" profit was, say 4 mil, and you had a 25% chance of success (post-decryptor), your spreadsheet would tell you the potential profit would have only been 1 mil.
Same thing with losses, it was underestimating them too, in the same case of 25% chance per job, a potential 4 mil loss would have been underestimated as a 1 mil loss.

Kazuo Ishiguro
House of Marbles
Posted - 2007.09.12 18:29:00 - [29]
 

Originally by: Akita T
You were merely underestimating the profitability per run by a factor of actual chance to invent Twisted Evil


Fixed that for you.

I've corrected the spreadsheet, cleaned up some of the formulas and added an additional 'break-even point' column:

http://dl.eve-files.com/media/corp/KzIg/invention-3.zip

Benvie
Benvie Enterprises
Posted - 2007.09.12 18:36:00 - [30]
 

Edited by: Benvie on 12/09/2007 18:38:19
Yep that's what I figured out too. Your formula was correct, it was just describing something else than what I thought it was. The formula describes profit per invention job. Your profit per unit, as said before, was incorrect though. I believe my formula is correct to derive that number though I'm sure you can use another method to come up with the number as well.

The profit per run was most interesting for me because that's what a lot of people look at...how much isk can be made on each of these items. The numbers are, well let's just say invention is more profitable than inventors usually like to let on. =D


Pages: [1] 2 3 4

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