Pages: [1] :: one page |
|
Author |
Thread Statistics | Show CCP posts - 0 post(s) |
Salvis Tallan
Gallente The Shadow Order SMASH Alliance
|
Posted - 2007.12.21 14:54:00 -
[1]
As we all know many things were nerfed with the onset of trinity, including that thing many producers hold dear, Perfect ME (Material efficiency Level). To find the level of ME that left us with no waste, It used to be a simple exercise of simply dividing the mineral of the greatest amount by 5, and moving on with what we needed to get done. This was because waste was rounded per run, and any waste less than 0.5 of a unit was considered to be 0. For the intents and purposes of this, we will use the actual ME formula provided by CCP, which is:
Waste = AmountOfMineral * ((BaseWasteFactor/ 100) / (MELevel + 1)) Total Waste = Runs * Math.Round(AmountOfMineral * ((BaseWasteFactor/ 100) / (MELevel + 1)))
Where BaseWasteFactor is a number we happen to get from the data tables (could be 5, 10, whatever). As I said before, pre-Trinity if waste was less than 0.5 it became zero, and then we multiplied waste by runs to get total waste.
But no more. Now the waste is not rounded pre-runs. This means that if you have a waste of 0.4 per run, One run will not see any waste for that material, but 2 runs will. This leaves us with an issue, We can always just produce one item at a time, but what happened to perfect!
Well, for this we need to look back at the original formula. If we plug out numbers into a fancy calculator, assume basewaste and runs and everything is greater than 0, and solve for when waste = 0, it breaks on us . Turns out that perfect ME is infinity (you can check this by taking the limit of the waste equation, no rounding, as ME approaches infinity). I would not like to be in empire waiting for a research slot.
Because of this I propose a new definition of 'Perfect ME'. We know that waste is compounded per run, but we also know that all blueprints have a Maximum amount of runs that they can be produced (300, 30, etc). Because of this, 'Perfect ME' should be the point at which there is no waste for the maximum amount of runs. Secondly, we must assume the rounding still occurs AFTER waste has been multiplied by the runs. This is because you cannot consume 0.4 of a material. Equipped with this information our waste formula goes through the following changes:
Waste = Runs * AmountOfMaterial * ((BaseWasteFactor/100)/(MELevel + 1)) 0.5 = Runs * AmountOfMaterial * ((BaseWasteFactor/100)/(MELevel + 1)) 0.5 = MaximumRuns * AmountOfMaterial * ((BaseWasteFactor/100)/(MELevel + 1))
As you can see, our only unknown here is MELevel, and we can solve for it.
Perfect MELevel = (0.02 * BaseWasteFactor * MaximumRuns * AmountOfMaterial) - 1 This is of course done per material, so the MELevel where ALL materials are perfect is just this result for the material you need the most.
What if you never produce the maximum amount of runs though? You get lucky. Fewer runs means a lower value for the Perfect ME Level, and we simply change the above equation to match.
Perfect MELevel = (0.02 * BaseWasteFactor * Runs * AmountOfMaterial) - 1
So our new equation is not a simple (or low) as it once was, but what can you do. Before you go out and tripple the queue time for empire research, Keep in mind your batch size! If you only make 10 of a rig at once, dont get it to perfect ME at 300 runs. Its a waste.
P.S. If I have screwed up any of the math or whatever, let me know. ------
|
Mona X
Caldari Polish Task Forces C0VEN
|
Posted - 2007.12.21 15:48:00 -
[2]
Originally by: Salvis Tallan
Because of this I propose a new definition of 'Perfect ME'. We know that waste is compounded per run, but we also know that all blueprints have a Maximum amount of runs that they can be produced (300, 30, etc).
Are talking about "production limit"?
|
Salvis Tallan
Gallente The Shadow Order SMASH Alliance
|
Posted - 2007.12.21 15:52:00 -
[3]
Originally by: Mona X
Originally by: Salvis Tallan
Because of this I propose a new definition of 'Perfect ME'. We know that waste is compounded per run, but we also know that all blueprints have a Maximum amount of runs that they can be produced (300, 30, etc).
Are talking about "production limit"?
Yup. Production limit is the most amount of runs you can put the blueprint in for. ------
|
Mona X
Caldari Polish Task Forces C0VEN
|
Posted - 2007.12.21 16:31:00 -
[4]
Originally by: Salvis Tallan
Yup. Production limit is the most amount of runs you can put the blueprint in for.
Actually it's not true. It's only most runs you can have on BPC. Othervise I couldn't do 400 runs with 300 production limit BPO.
|
Pwett
Minmatar QUANT Corp. QUANT Hegemony
|
Posted - 2007.12.21 16:38:00 -
[5]
Can we just use my formula?
ME 39 = perfect-enough. _______________ Pwett CEO, Founder, & Executor <Q> QUANT Hegemony
|
Salvis Tallan
Gallente The Shadow Order SMASH Alliance
|
Posted - 2007.12.21 16:46:00 -
[6]
Originally by: Mona X
Originally by: Salvis Tallan
Yup. Production limit is the most amount of runs you can put the blueprint in for.
Actually it's not true. It's only most runs you can have on BPC. Othervise I couldn't do 400 runs with 300 production limit BPO.
I really need to build things smaller than battleships. After asking around and going over data, turns out you cant produce for more than 30 days. Thank you for pointing this out to me, Ill update the post and math. ------
|
Jennine Tyler
NewDawn
|
Posted - 2007.12.22 00:09:00 -
[7]
so parallelism is key then?
I need 5000 Widgets at 800 Trit/Widget. To make the BPO Perfect for 5000 Runs, would require ME 7999, I believe?
7999 =(0.02*0.1*5000*800)-1 PerfectME =(0.02 * 10% Waste * 5000 Runs * 800 Trit)-1
However, if I were to obtain 5x Blueprints. 1599 =(0.02*0.1*1000*800)-1 PerfectME =(0.02 * 10% Waste * 1000 Runs * 800 Trit)-1
Now, this admitedly has extra cost, but I have access to labs & blueprints arent (nessecarily) expensive. The other downside is that if i run multiple jobs, or smaller batches, I either run out of factory slots, require more alts, or more interaction.
Is this CCPs intent? When will we see non-rounded data on blueprints?
|
Buildius Maximus
|
Posted - 2007.12.22 05:11:00 -
[8]
It's silly, of course, for CCP to do this new rounding...as they are now advocating increasing slot useage, which I hope is not their intent.
I mean seriously, the amount of materials to make 10x1-run items should not be less than the amount of materials to make 1x10-run items...just doesn't make any sense. Even if it were to add more overall, it makes sense to round the base materials after Material Level is applied, even if the new formula is Math.ceil(AmountOfMaterial * (BaseWasteFactor/100)/(MELevel + 1)))
Please CCP, change it so it is at least consistent across the number of runs.
|
Trivas
NQX Innovations Southern Cross Alliance
|
Posted - 2007.12.22 06:50:00 -
[9]
What would really be nice is to see how much my BP actually uses for the job with all the decimals included. How the heck am I supposed to know how many minerals to buy for a multi run job now?? Even if it was just when you rolled over the amount with your mouse would be good.
|
Quutar
Ars ex Discordia
|
Posted - 2007.12.23 19:42:00 -
[10]
for reference, the production limit is for BPCs only.
The Gravimetric sensor cluster blueprint has a 1,000 production limit, but I am currently building 6,341 of them from a single BPO
Quutar Research Services Amarr Outpost BPCs ME:10 |
|
Trayk
Minmatar The Dead Parrot Shoppe Inc. Brutally Clever Empire
|
Posted - 2007.12.23 20:34:00 -
[11]
Originally by: Buildius Maximus It's silly, of course, for CCP to do this new rounding...as they are now advocating increasing slot useage, which I hope is not their intent.
This has been their game ever since the ammo production nerf. You cant tell me it takes just as long to build 1 Bullet (or even box of bullets) as it does to produce 1 of the gun that fires it.
More alts = more accounts!!
Please resize image to a maximum of 400 x 120, not exceeding 24000 bytes, ty. If you would like further details please mail [email protected] - Cortes
|
Pwett
Minmatar QUANT Corp. QUANT Hegemony
|
Posted - 2007.12.24 02:25:00 -
[12]
Originally by: Trivas What would really be nice is to see how much my BP actually uses for the job with all the decimals included. How the heck am I supposed to know how many minerals to buy for a multi run job now?? Even if it was just when you rolled over the amount with your mouse would be good.
Easiest way is to basically take your perfect build for one item, and add 1 of every mineral per run. _______________ Pwett CEO, Founder, & Executor <Q> QUANT Hegemony
|
ShadowRat
|
Posted - 2007.12.24 08:08:00 -
[13]
CCP are starts to 'acumulate' your production waste at their account. So like in early e-banking they just cut some digits after dot and store it at another account. Your waste help make toooonnnsss of 'free' items and support rat's factions at belts.
So DONT do perfect ME it is pointles... Without 'waste' there will be nothing at belts.
Have nice day.
|
Stefan F
Enrave Knights Of the Southerncross
|
Posted - 2007.12.24 11:09:00 -
[14]
As i see your reasoning behind the "perfect ME at 30 days production queue size" I dont think this will be a reasonable way of putting it, as it would take like 100 years to bring a decent bpo set to the enormous ME levels required, as someone before me already calculated.
What would be a more usable standard would be the following: Assume the average player can log in once a day to refresh their jobs Assume producers dont want any waste.
If we take these two assumptions into account perfect ME would be meaning: "ME level which causes zero mineral loss on a n-run job covering 24 hours, rounded down"
This would still require a lot of BPO's to be researched a whole lot higher in their ME, but it wouldnt take you to ME1000+, let alone ME7999 mentioned before. Do the math yourself and input plz.
|
UK Glum
Gallente BTTF Enterprises
|
Posted - 2007.12.24 13:24:00 -
[15]
lets take gamma M as an example
The pre-Trinity perfect ME was 8. Which took just under 8 hours. Lets call it 8 for ease of use ie 1 hour per ME (I think its 56 minutes but lets keep it simple).
The manufacturing time for 1 batch of these is 120 seconds. So in the proposed 30 days maximum production run we could produce 21,600 batches.
Now multiplying the max usage of minerals (41 units of mexallon) and the base wastefactor (0.1) etc using your proposed formula we get
(0.02 * BaseWasteFactor * MaximumRuns * AmountOfMaterial) - 1 (0.02 * 0.1 * 21600 * 41 ) - 1
So the perfect ME level is now 1770.2 instead of 8 prevously. This will take 220 times longer (now nearly 74 days) to research :(
Now I dont think anybody would do a whole month of Gamma M but where do you draw the line?
We would never get ships such as bs etc to perfect anyway so going from a million ME for perfect to 400 million is irrelevent, its more components and equipment where the manufacturing time is quite short.
I tried plugging your propsed changes into my manufacturing spreadsheet but I'm getting new ME figures of 100 times more than you are. Guess my formulas needs some work to integrate this. Once thats done I'll need to plug in an expected batch run and determine the ME for that.
I already have an perfect ME and Ideal ME column. I've found that the difference between ideal and perfect can be huge in terms of research time but may save only a few units of trit or pyrite. Eg Valkyrie perfect is 490 (7d 15h 45m), but ideal is only 98 (1d 12h 45m of research) where I lose 2 units of trit. Looks like I need a perfect and ideal columns for production run size too.
|
Salvis Tallan
Gallente The Shadow Order SMASH Alliance
|
Posted - 2007.12.24 14:01:00 -
[16]
Originally by: Stefan F As i see your reasoning behind the "perfect ME at 30 days production queue size" I dont think this will be a reasonable way of putting it, as it would take like 100 years to bring a decent bpo set to the enormous ME levels required, as someone before me already calculated.
What would be a more usable standard would be the following: Assume the average player can log in once a day to refresh their jobs Assume producers dont want any waste.
If we take these two assumptions into account perfect ME would be meaning: "ME level which causes zero mineral loss on a n-run job covering 24 hours, rounded down"
This would still require a lot of BPO's to be researched a whole lot higher in their ME, but it wouldnt take you to ME1000+, let alone ME7999 mentioned before. Do the math yourself and input plz.
I did point out that you should only consider the most runs you will do at once. However, this is exactly what CCP wants to do, create an economy of scale. Before there were certain blueprints that anyone could get to perfect, and that gave the market of that item a sort of price floor (since it cost everyone the same value to build). Now there is an advantage to using multiple BPOs at once. ------
|
UK Glum
Gallente BTTF Enterprises
|
Posted - 2007.12.24 15:17:00 -
[17]
Edited by: UK Glum on 24/12/2007 15:22:04 I've been playing further on this.
If we take your formula for a 1 run job. Perfect MELevel = (0.02 * BaseWasteFactor * Runs * AmountOfMaterial) - 1
plugging in the gamma m stats again.....
(0.02 * 0.1 * 1 * 41) -1 (0.0082) -1 = -0.9918. Rounded down would give -1
Also, taking the afore mentioned Valkyrie. The new max run ME works out at 105,925 from the pre-trinity of ME490, but the 1 run ME works out at 4.
that just looks wrong
|
Alusha
Caldari Lobster of Babel
|
Posted - 2007.12.30 11:34:00 -
[18]
Edited by: Alusha on 30/12/2007 11:45:59 Ok so reasonable "Perfect ME" is the amount of ME needed for 1-7 days of production. While theoretical "Perfect ME" is the amount for 30 days of production with good PE research and industry 5.
so a normal module takes 6,4 minutes to build. in 24 hours that equals 225 runs. in 1 week it equals 1575 runs in 30 days it equals 6750 runs
with a material need of 100 units for a single run, ME 0 = 110 units ME 20 = 100 units = "pre-Trinity Perfect ME" (actual fractional waste is 0,4762)
24-hour Perfect ME is now: ME 0 = 24750 units (2250 unit waste) ME 20 = 22607 units (107 unit waste) ME 50 = 22544 units (44 unit waste) ME 100 = 22522 units (22 unit waste) ME 500 = 22504 units (4 unit waste) ME 4500 = 22500 units (no waste on 24 hour builds)
Still quite horrible.
INNOVATION AND RESEARCH services
|
Ki Shodan
Gallente deep blue
|
Posted - 2007.12.30 13:34:00 -
[19]
If you use the max. run in 30 days limit to calculate perfect ML. You got to ask, 30 days in a normal factory or in an POS assemvly array and in what kind of assembly array? Base time multiplier: 0.75 or 0.65?
For most items the new rounding does not make a real difference, you only got to update your spreadsheets and pass the waste on to the consumer.
On Ammo, Missles, Lenses, Probes and Drones it does. For those, i think, the production limit on a BPC is the criteria for most producers. It is a reasonable batchsize and a perfect ME for this amount of runs is reachable in good time.
Again another case are rigs, while I doubt anyone will produce a batch of 300 CCC rigs at once. More than one rig per batch is quite usual, I guess. So most of the rig BPOs have to be re-researched to a new ML, depending on the usual batch size of the producer. --
Evemail me, if my name is used as guarantor! |
Ki Shodan
Gallente deep blue
|
Posted - 2007.12.30 13:55:00 -
[20]
Another thing speaking against the "30days/time for 1 run" is PL. What level do you use 0, 50, perfect PL? (I use perfect PL BPOs, normally ...) With perfect PL you got a another factor to put into the equasion.
To keep it practical calculate with the production limit of a BPO, that eliminates PL and the factory time multiplier on 99% of the BPOs. --
Evemail me, if my name is used as guarantor! |
|
Salvis Tallan
Gallente The Shadow Order SMASH Alliance
|
Posted - 2007.12.30 15:04:00 -
[21]
The point I was trying to make was that ME waste is now a function of runs, and so those people obsessed with perfect ME need to pick a maximum point for their runs value. Its not a pretty or low number, and it certainly gets the economy of scale thing across. ------
|
Ki Shodan
Gallente deep blue
|
Posted - 2008.01.08 00:06:00 -
[22]
There is no real perfect ML any more above 1 run for most BPO.
To optimize an Antimatter L to be perfect at 2 runs you need to research to 639, at 1500 runs to be prefect you need a ML of 479999, which is, well, not practical.
As rule of thumb: If you already have the perfect ML for one BPO simply multiply it with the runs you have in mind, for that specific BPO and you get an estimate of how useless it is now to research even Ammo BPOs to their productionlimit batch size ... --
Evemail me, if my name is used as guarantor! |
Ki Shodan
Gallente deep blue
|
Posted - 2008.01.08 00:15:00 -
[23]
Originally by: UK Glum If we take your formula for a 1 run job. Perfect MELevel = (0.02 * BaseWasteFactor * Runs * AmountOfMaterial) - 1
plugging in the gamma m stats again.....
(0.02 * 0.1 * 1 * 41) -1 (0.0082) -1 = -0.9918. Rounded down would give -1
Also, taking the afore mentioned Valkyrie. The new max run ME works out at 105,925 from the pre-trinity of ME490, but the 1 run ME works out at 4.
that just looks wrong
You have to use 10% not 0.1 as BaseWasteFactor. (5% on eg. drones.) --
Evemail me, if my name is used as guarantor! |
Brixer
Dai Dai Hai
|
Posted - 2008.01.08 11:46:00 -
[24]
You guys are crazy... hehe
Perfect ME has to go. Some of the BPOs can *never* be research to perfect ME because of time until earth get absorbed by our own sun.
For me perfect ME on *any* BPO is 40ish.. Some easy ones I bring to 100, that goes for bombs / rigs / frigs mostly, where there *might* be anything to gain in isk vs research time.
|
Ki Shodan
Gallente deep blue
|
Posted - 2008.01.08 14:58:00 -
[25]
Originally by: Brixer You guys are crazy... hehe
No arguing there, but if you and you research Alts have nothing else to do ...
Having run the calculation on several BPOs now, mods, ships are not really worth bringing over 30 (and that is still an overkill for most...). Bringing Ammo to prefect for one run still viable now, but more than that it is academic and only looks good, but is not really worth your time, shaving of the last round of trit not really worth it.
Although, researching rigs to a perfect level for 5 or 10 runs will be worth it, i think. --
Evemail me, if my name is used as guarantor! |
|
|
|
Pages: [1] :: one page |
First page | Previous page | Next page | Last page |