These forums have been archived and are now read-only.

The new forums are live and can be found at https://forums.eveonline.com/

EVE Information Portal

 
  • Topic is locked indefinitely.
 

Dev blog: Researching, the Future

First post First post First post
Author
Shoogie
Serious Pixels
#601 - 2014-04-30 21:30:08 UTC
You have never researched a blueprint in your life, have you?

Treyan Argund wrote:
I'm only on page 20, but I've been reading several posts saying "But my ME6 BP is already perfect, why do I have to research 4 more levels post patch to get it perfect again?" and the like. The amount of stupid in these posts whining about ME is staggering and it makes me angry. I have not seen a post yet that quashes these misconceptions. This is that post. It is not nice.

Here's the Idiot's Guide to how ME Will (not) Change on Patch Day:

Take a BP, any BP. Let's say it has perfect ME at ME 7. After the patch, your ME will have changed to ME10 (10% reduction in waste). No effective change in how much waste you have, and you even get a bigger number!


No. If you have a BPO that is ME 7 before the patch, you will get handed an ME 9% BPO after the patch. The thing is, if it was perfect at ME7 before the patch, then there will be no difference between the build cost of ME 9% and ME 10% blueprints after the patch... at least until you start factoring in POS assembly arrays and workforce teams that add to ME. Those will probably differentiate the two.

Treyan Argund wrote:
Take that same BP, and some silly person has researched it to something silly like ME200. Silly person thinks that because his number went from ME200 to ME10 he should get something back or that it is somehow worse. No. Wrong. You researched poorly and far past the perfection breakpoint. The cap of ME10 is to prevent people like you from wasting research, although I would love to see Greyscale remove the cap so that you guys can go chasing the Bigger Better Number again while the rest of us who actually know what the hell we're doing point and laugh.


Yeah, we actually have the same opinion there.

Treyan Argund wrote:
Now, Let's take a BP that's perfect at ME20 pre-patch. If you have ME20 (or greater, for you "bigger is better" over-researching silly persons), your BP will have ME10 after the patch. Fear not though! While the smarter beings in this thread have already figured this out, just because your ME number got smaller does not mean you have more waste. HOWEVER (silly persons, read carefully), if you have this BP, and it's only ME4 (remember, the perfect ME for this BP is 20 pre-patch), and you research it to ME10, you're not getting any sort of shortcut to perfect ME because of the patch. Whatever wastage ME10 gets you pre-patch (let's say it gives you 7% wastage) is going to directly translate to what ME you get post-patch (pre-patch wastage of 7% = post-patch ME7)


No. Wrong. If you have a blueprint researched to ME 10 today, that means the wastage on that blueprint is 0.91%. That gets truncated to 0, and you will recieve a perfect ME 10% blueprint after the patch. Just the same as the person who had a previously perfect ME 20 blueprint. You didn't read the dev blog, did you?

Treyan Argund wrote:
TL;DR the current ME on your blueprints has absolutely nothing to do with what the ME will be after the patch. Go back and read that sentence a few more times until it has replaced whatever other silly notion you had. Wastage pre-patch=Wastage post-patch. The ME number prepatch is irrelevant. CCP has even kindly offered generous rounding so that if you have, say, 5.4% wastage pre-patch, they'll give you 6% post-patch due to the less granular 10-step ME progression.


Except that the current ME level on your blueprint directly determines the wastage on that blueprint, and from there you know exactly what you will get post-patch. Have you ever opened up the show info window on a blueprint before?

ME4 blueprints pre-patch all get converted to ME8% blueprints after the patch.
ME5 through ME9 blueprints pre-patch all get converted to ME9% after the patch.
ME10+ blueprints pre-patch all get converted to "Perfect" ME 10% after the patch.

So if you have an ME4 blueprint today, try to research it to ME10. If it takes too long before patch day to get it researched to 10, try to research it to 5. If you have an ME5 blueprint today, but cannot get it to level 10 before patch day, don't bother researching it anymore.
MailDeadDrop
Archon Industries
#602 - 2014-04-30 21:41:52 UTC
Shoogie wrote:
Except that the current ME level on your blueprint directly determines the wastage on that blueprint, and from there you know exactly what you will get post-patch. Have you ever opened up the show info window on a blueprint before?

ME4 blueprints pre-patch all get converted to ME8% blueprints after the patch.
ME5 through ME9 blueprints pre-patch all get converted to ME9% after the patch.
ME10+ blueprints pre-patch all get converted to "Perfect" ME 10% after the patch.

So if you have an ME4 blueprint today, try to research it to ME10. If it takes too long before patch day to get it researched to 10, try to research it to 5. If you have an ME5 blueprint today, but cannot get it to level 10 before patch day, don't bother researching it anymore.

Agree, expect there's a small asterisk. I have heard of, but have no direct knowledge of, some blueprints which have a base waste factor that is not 10% (like the vast majority of blueprints); I think I heard those blueprints have a base waste factor 5%. So for those blueprints, I really don't know what CCP is going to do. Presumably they are going to do some :math: to fit them into the ME0%..ME10% scale of the new system.

Anyways, if you possess one or more of the "5% base waste" blueprints you should probably do some careful thinking.

MDD
DireNecessity
Mayhem-Industries
#603 - 2014-04-30 23:01:35 UTC  |  Edited by: DireNecessity
Seith Kali wrote:
With all the other ME bonuses we are getting through teams and pos bonus changes, 10% cap looks far more reasonable.


Far more reasonable compared to what? Optimal BPs in the old system reduced required material by 10%. Optimal BPs in the new system will also reduce required material by 10%. That 10% will soon be calculated differently and the currently proposed transition is creating winners and losers. In many cases the wins and losses are minimal but in many other cases the wins and losses are astoundingly large. To date CCP hasn't commented much on their creation of winners and losers beyond an initial "we're aware this will happen" statement.
Flyinghotpocket
Doomheim
#604 - 2014-04-30 23:57:03 UTC
CCP Greyscale wrote:


Mal Nina wrote:

Will we see the ability to research invented BPCs to improve the ME/TE? if not are you finally going to get rid of BPOs and give us nuggs a chance to compete and actually make money?



We're not planning on letting any BPCs be researched right now, and we don't have a concrete plan in place regarding the future of T2 BPOs right now.


******* get rid of them. they are bullshit and you know it. nobody will take this industry **** seriously until you remove those damn t2 bpo's

Amarr Militia Representative - A jar of nitro

ALI Virgo
Dreddit
Test Alliance Please Ignore
#605 - 2014-05-01 00:08:28 UTC
remove t2 bpo instead
Krystyn
Serenity Rising LLC
#606 - 2014-05-01 02:57:37 UTC
ALI Virgo wrote:
remove t2 bpo instead

That would be nice...

My 'over' researched BPOs will be nerfed into uselessness thanks a lot CCP.
Now selling BPCs will be a waste of time mostly.
Well since I can't use my POS to safely research expensive BPOs thats out too.
2 major hits to my corps isk generation from industry related things
-1 corp doing any real industry
probably - multiple subs
Loraine Gess
Confedeferate Union of Tax Legalists
#607 - 2014-05-01 03:12:05 UTC
Krystyn wrote:
ALI Virgo wrote:
remove t2 bpo instead

That would be nice...

My 'over' researched BPOs will be nerfed into uselessness thanks a lot CCP.
Now selling BPCs will be a waste of time mostly.
Well since I can't use my POS to safely research expensive BPOs thats out too.
2 major hits to my corps isk generation from industry related things
-1 corp doing any real industry
probably - multiple subs




Use your perfect blueprints to make shittons of money??


I guess that's too hard for you, even though all you have to do is run an application and it will do it all the math for you...
Valterra Craven
#608 - 2014-05-01 05:11:59 UTC
I understand why CCP is doing what they are doing (even if I dont like it) in terms of the push vs pull mechanic in terms of industry, but one thing I don't understand is why the "optimal" ME levels are become so curvey instead of so linear? It seems alittle insane for a game thats only been going on 10 years to require someone to research a bpo for 6 years (even if a high end one) to reach level 10 ME. What exactly is the "gameplay" benefit or the reasoning behind doing this.... I just don't understand why you'd want to do this vs the old system where each level required the same amount of research time.


In any case, I have an idea for keeping numerous pos modules relevant.

Parallelism.

In other words the "killer feature" of POS would be to allow you to break up research jobs in parallel to complete them faster.
You have two mods, you can break them up to complete twice as fast and so and so forth.

I did some rough estimates, and it looks like with a dread gur tower and assuming labs cpu cost of 500 cpu would allow you to have 15 labs at once if that's all you put on it. So to balance this you could either hugely increase the cpu cost so that it wouldn't be wise to go over 3-4 labs, or limit the amount of jobs that you could run in parallel. (I'd say balance it on the average number of labs people run now). I'd also mess with the current numbers that labs give bonuses to so that they are closer to NPC stations. In this way, POS don't compete with NPC/Null stations in the same way.

To be fair this could also be adapted to production jobs as well.
DK Anaroth
Border Industrial Limited
#609 - 2014-05-01 06:37:06 UTC  |  Edited by: DK Anaroth
CCP Greyscale wrote:
...


I was thinking about the research changes and realised that it can be split up into a smaller changes which may help analyze the situation (incoming wall of text).

1) Recast the material and processing (time) efficiencies in terms of percentage savings
Each existing blueprint has specified ME and PE levels, which are really just a count of how of how many times it has been researched, and it takes some arithmetic to determine how it effects the materials used, and time taken to manufacture using the blueprint. For ME, the percentage of waste is currently reported, and there is no additional information reported for the processing time efficiency.

In the next expansion the blueprints will instead report the percentage reductions over the base unresearched blueprint. Reporting of the waste percentage will also be removed.

You can calculate the percentage reduction corresponding to the levels of research. For both materials and processing time there is a limit value that you can't go below. For material efficiency the limit is a 9.09% reduction and corresponds to zero waste. For time efficiency it is a 20% reduction. At each level the time reduction is 2.2 times the material reduction.

The step makes a lot of sense. The new system will be much easier for most people to understand, as percentage reductions are used in many other places in Eve.

2) Adjust the base values stored on blueprints that are used for calculating the requirements for each of the source materials.
In the static database, currently, the material requirements are calculated by adding the waste percentage of the blueprint to the base numbers.which are set for the so-called perfect efficency (without waste)

In the expansion all the blueprints will be modified to have the unresearched numbers as the new base, and apply a reduction based off those. Doing so also gives the opportunity to apply an overall adjustment to materials or even finer grade adjustments at the blueprint or even material level.

One issue with this is that the due to the calculations having to round values to integers, there will inevitably be small differences in the material requirements for some material efficiencies before and after this step even if there no deliberate change. In that sense it may be better for the material reduction to be display only and stll use the old way of calculating the requirements internally, you can always convert between the material reduction and waste percentages.

A bigger issue is that the current adjustment from the perfect base point to the unresearch base point adds 11.1% rather than the expected 10%, which means that the unresearched blueprints and thus all blueprints will effective have 1.1% increased requirements.

3) Simplify the research process and make it more like training skills.
Currently, research works by doing a fixed period of research no matter what the current efficiency level of the blueprint is. The level is then increased by 1 and the efficiency of the blueprint is adjusted. There is no theoretical limit to amount of research you can do but after a while there will be little difference, and eventually any additional research will make no difference to the efficiency of the blueprints. 90% of what can be achieved is available at the an efficiency level of 9, and 95% at a level of 19. The shape of percentage reduction curve over research time is a negative hyperbola, with a horizontal asymptote at the previously mentioned limits.

For the expansion, the system will be changed to resemble gaining levels in the skill system. This means that there will be a maximum limit to the amount of research you can do on one blueprint, there will be a limited number of steps (currently 10), and the research time between the steps (and the total time for each step) will increase as a power series. The shape of the percentage reduction curve over time is thus logorithmic (although with a limit). The material reduction will go up in steps 10 of 1% and the time reduction in 10 steps of 2%.

One issue with this is the long research times at the high-end of the scale. With skill training you can switch around partly through gaining an increment. Introducing more steps would ease the situation somewhat, but the power relationship means you'd have many small extra steps at the start. As the relationship between research time is known for the scheme you could in theory have a floating point material reduction and chose whatever research periods you like. Another way of implementing that would be to keep track of the total researched time and only update the percentage reduction when it crosses the threshold. That would make things more skill like.

A second issue is that the ratio of time reduction to material reduction has been reduced to 2 from 2.2, which means that material research gets better faster than it used to.

The big issue is how to transfer existing blueprints to this scheme, in fact, it's the major point in contention and so belongs in a section of its own. In fact it'll need a post of its own.
DK Anaroth
Border Industrial Limited
#610 - 2014-05-01 06:39:35 UTC  |  Edited by: DK Anaroth
DK Anaroth wrote:
[quote=CCP Greyscale]..[./quote]

I was thinking about the research changes and realised that it can be split up into a smaller changes which may help analyze the situation (incoming wall of text).


Wall of text continues...

4) Convert the existing blueprints to work in the new framework.
The plan seems to be to convert every existing blueprint so that no currently researched blueprint gets any worse in terms of the bonus it provides. There are also some details on the conversions, level 1s will become 5% material reduction and 10% time reduction, level 5-9 will become 9%/18% and above level 9 will become the maximum 10/20. There are also some details for the negative level invention blueprints. Going off the time reduction information that I calculated level 2 would be 7%/14% and level 3&4 8% / 16% in accordance with the above scheme.

While this plan has a good intent, all existing blueprints are lumped into only 6 of the 11 pecentage categories. This is because the difference between level 0 and level 1 is a 5%/10% difference. Adding finer subdivisions would help with this, and moving to a continous model (or the skill point equivalent) would eliminate this problem.

Unfortunately even in the continous model there is an issue with the research time. For processing time efficiency as you move towards a 17.5% reduction (around level 7) it will take over twice as long to train to that level post expansion as before expansion, which is hardly ideal when you're trying to attract new players to industry. From there it slowly settles down, at 19.0% (level 19) it's about 1.46 times as long. A better option maybe to transition based on time taken, in which case the actual efficiency of the blueprint will change with the patch, but it's effect isn't as severe. The peak is earlier (at around level 6) and at that point the time reduction is about 90% of what it should be, however that only means that actual processing takes 1% longer than it would. Using coarser groupings of levels greatly magnifies the effects as others have reported. I have yet to finish the analysis of the material efficiency but it will be more extreme, as there is an additional factor due to changing the ratio of time reduction to material reduction.

Related to the last point, the listed rates for invention would widen the advantages that the T2 BPOs have over invented products. This would be relatively easy to fix though as they're just constants.

Conclusion

Changing to using percentage reductions seems worth it to me. I'm not so sure about modifying the way the research works is worth it especially if the conversion of existing blueprints is done as clumsily as originally suggested. With some tweaks, such as focussing on converting the time investment rather than the actual change in values then it might be a good idea. The other thing to bear in mind is that with the changes to the other parts of manufacturing, it's quite likely that those effects will be far more important to the bottom line than small changes to the material or time efficiency of the blueprints.
Mohenna
Knights of the Dark
#611 - 2014-05-01 09:22:11 UTC
Sorry if I don't read 32 pages of thread... Has anyone mentioned the issue of low MR small cost bpos that are already perfect? Say for example all the small ammo bpos.

It would be a truly major PITA to go and research all of these back up to the new "perfect" status. Is it possible to have perfect status bpos moved to perfect status in the new system?
Tatjana Jouhinen
Doomheim
#612 - 2014-05-01 11:21:33 UTC  |  Edited by: Tatjana Jouhinen
I really do not want to whine over the existance of T2 BPO, but this expansion would have been the chance to slightly adjust the difference between T2 BPO and invented BPO. In the new system, a T2 BPO can be researched to ME10%/TE20% in order to have zero waste.

What if you would reset invented BPC in a way, that the negative ME values are set to zero. Without any changes to decryptors, max ME levels for invented T2 BPC would be 7%. This would still retain a benefit for T2 BPO, but would close the gap a bit without completely destroying the value of T2 BPO, especially as I would not change the system for TE values. I f the remaining advantage is not big enough, the decryptors could be adjusted, or the team bonuses could be altered depending on whether production is done with an original or a copy.

Another option would have been to give a general ME/TE malus on production based on copies (this for both T2 and T1), so the decision to be made would be to either put the BPO on risk (= put it in a POS) or accept worse ME values.
Darin Vanar
Federal Navy Academy
Gallente Federation
#613 - 2014-05-01 11:51:51 UTC
Valterra Craven wrote:
I understand why CCP is doing what they are doing (even if I dont like it) in terms of the push vs pull mechanic in terms of industry, but one thing I don't understand is why the "optimal" ME levels are become so curvey instead of so linear? It seems alittle insane for a game thats only been going on 10 years to require someone to research a bpo for 6 years (even if a high end one) to reach level 10 ME. What exactly is the "gameplay" benefit or the reasoning behind doing this.... I just don't understand why you'd want to do this vs the old system where each level required the same amount of research time.


I don't even know if I'll be ALIVE for the next 6 years. But my character's research will still require to be ongoing. I don't even know what to say about that.
Darin Vanar
Federal Navy Academy
Gallente Federation
#614 - 2014-05-01 12:19:03 UTC
Actually I do have more informative feedback for this thread.

CCP, it would be in your best interest to adapt a more granular research system, and one where players still require to be alive in order to benefit from their investment. I would rather pay a higher tax, while still being alive, and purchase a higher research BPC from someone else, even if they have to PLEX to gain that capital to do so, while still being alive. I'm sure they would appreciate that as well.

Also, it would benefit you as a business, to see some of that investment in profit, as you are running a business, not a retirement fund. Or are you planning to transition into a retirement facility housing old people who may not even see a dime on their in game investment, before they, you know, die?

This research time system is certainly not dynamic in any way. Granular would be better, a good investment for player capital, and would make more sense from a business point of view on your end. We all want to accomplish our goals while we are still alive. I'm sure you would benefit from having living players playing your game. You know, as opposed to players who have passed away, but still have research projects ongoing that will complete, in say the next 3 years.

I'm ballparking it here. But you know, just sayin'.
Shoogie
Serious Pixels
#615 - 2014-05-01 12:39:06 UTC
Mohenna wrote:
Sorry if I don't read 32 pages of thread... Has anyone mentioned the issue of low MR small cost bpos that are already perfect? Say for example all the small ammo bpos.

It would be a truly major PITA to go and research all of these back up to the new "perfect" status. Is it possible to have perfect status bpos moved to perfect status in the new system?


Yes, the small cost BPOs have been mentioned. For these blueprints, when they multiply everything by 1.11111 and then round to the nearest whole number they will end up right where they started. So your small rig BPO that was effectively perfect at ME0 is still effectively perfect at ME0.

But is it? They are adding teams that can reduce material costs by 7.5%. They are also adding 2% ME for building in a POS. But we don't know how those bonuses stack with blueprint bonuses. If they add, you can get 9.5% ME on an unresearched Blueprint. This yields nothing on the small cost blueprint in question. But if you research that blueprint to 10% and it adds to 19.5% the cost savings could be significant.


With infinite slots suddenly coming open in NPC stations, I think there will be a ton of ME research done in the first couple weeks of the expansion, followed by not much.
Mohenna
Knights of the Dark
#616 - 2014-05-01 13:15:33 UTC
Shoogie wrote:
Mohenna wrote:
Sorry if I don't read 32 pages of thread... Has anyone mentioned the issue of low MR small cost bpos that are already perfect? Say for example all the small ammo bpos.

It would be a truly major PITA to go and research all of these back up to the new "perfect" status. Is it possible to have perfect status bpos moved to perfect status in the new system?


Yes, the small cost BPOs have been mentioned. For these blueprints, when they multiply everything by 1.11111 and then round to the nearest whole number they will end up right where they started. So your small rig BPO that was effectively perfect at ME0 is still effectively perfect at ME0.

But is it? They are adding teams that can reduce material costs by 7.5%. They are also adding 2% ME for building in a POS. But we don't know how those bonuses stack with blueprint bonuses. If they add, you can get 9.5% ME on an unresearched Blueprint. This yields nothing on the small cost blueprint in question. But if you research that blueprint to 10% and it adds to 19.5% the cost savings could be significant.


With infinite slots suddenly coming open in NPC stations, I think there will be a ton of ME research done in the first couple weeks of the expansion, followed by not much.


Sounds like a case could be made to move perfect bpos to 10%...

That'd be a nice QOL bonus for OCD players (the majority, me incl.)
Sigras
Conglomo
#617 - 2014-05-01 15:42:31 UTC
Mohenna wrote:
Sorry if I don't read 32 pages of thread... Has anyone mentioned the issue of low MR small cost bpos that are already perfect? Say for example all the small ammo bpos.

It would be a truly major PITA to go and research all of these back up to the new "perfect" status. Is it possible to have perfect status bpos moved to perfect status in the new system?

I understand not reading the whole thread, but at least read the blue posts...

and yes, grayscale said the way rounding works that if a BPO was perfect before the change it will be perfect after.

IMHO they should do it on time and not waste but thats a different issue.
Chi'Nane T'Kal
Interminatus
#618 - 2014-05-01 20:40:01 UTC
In my opinion, you should simplify the ME/TE system even further - and at the same time address the worst issues people have with the transition.

I propose:

Blueprint LEVEL - that is: no more differing between ME lvl and TE lvl

Blueprint levels would go from 1 to 10. 1% material efficiency for every level, 2% time efficiency, but both are always at the SAME level

That means, at the price of reducing flexibility slightly by removing a choice that was IMO rather underused (TE research was AFAIK always insignificant compared to ME research), you solve a lot of problems.


- instead of researching level directly, you could research blueprint experience, using the level as a calculated value
- partial research would not be an issue anymore, as every research second translates directly into experience
- conversion of old BPOs would be absolutely FAIR, since you could directly convert from accumulated research time to experience
- FULL market integration would be a lot easier with only 11 different blueprint levels

and on the technical side:
- you would only have to store 1 value (experience) for every BPO
MyHaula
Wages Of Sin
#619 - 2014-05-01 22:10:05 UTC
I would have preferred something a little more player driven. Keep the slot limitations in the stations and the ability to remotely copy and research BPOs. In place of these changes it would've been nice to see a finer granularity for players to rent out corporate POS slots to non corporate or alliance members. That would be very much in keeping with the sandbox player generated nature of EVE while allowing competition to drive down the prices of copy and research slots through providing additional supply.
Drone 16
Holy Horde
#620 - 2014-05-01 23:00:17 UTC
MyHaula wrote:
I would have preferred something a little more player driven. Keep the slot limitations in the stations and the ability to remotely copy and research BPOs. In place of these changes it would've been nice to see a finer granularity for players to rent out corporate POS slots to non corporate or alliance members. That would be very much in keeping with the sandbox player generated nature of EVE while allowing competition to drive down the prices of copy and research slots through providing additional supply.


I tend to agree. The lack of remote research seems like it will funnel players into limited research stations. No one sane is going to risk expensive BPOs in their POS. I thought the idea as to get everyone to spread out.

It puts the peanutbutter on itself or it leaves the bonus round... - E1's greatest Hits