EVE Industry - All you want to know
This blog is going to go over everything we're doing with research and costs in Crius, to explain the "final form" of what we're shipping so you can see everything that's happening in one place, and to act as long-form release documentation (the patchnotes of course go into significant bullet-pointed detail).
If something relating to cost or research is not listed here, then it isn't happening, even if it was mentioned in a previous blog or post.
M.E. AND T.E.
- PE (Production Efficiency) is now called TE (Time Efficiency) because it's more obvious what it means. ME (Material Efficiency) keeps the same name. In both cases we are now displaying both values in the UI as negative numbers, as they are reductions; this is possible because of the invention changes mentioned later on!
- Wastage as a concept is gone. All material-affecting bonuses simply reduce from the "full price" input materials. Material needs for T1 items have all been adjusted upwards by 1/0.9 to balance this out. We no longer need to guard so vigilantly against "material printing" from ME bonuses as reprocessing-built items now only return 50% of the materials.
- ME and TE research now operates on a 10-step scale. Each step of ME reduces materials by 1% (always rounded up). Each step of TE reduces build time by 2%. These are expressed as simple percentage values, so a blueprint might be ME-3%, TE-12%.
Two specific notes: firstly, we are now rounding materials per-job rather than per-run, so savings of less than one unit can often be realized with a long enough production run. Secondly, to prevent weirdness with T2 builds in particular (for example building 10 Paladins from 9 Apocalypses), every run requires a minimum of one unit of every listed material.
In order to keep a non-linear progression, each subsequent level takes longer to research than the previously one.
Here is the time for a T1 ammo blueprint:
|Approximate readable time
(If you look at the even-numbered levels, the values may be familiar; this is intentional!)
In order to keep blueprints differentiated, we have a system of ranks, all of the above times are multiplied by a blueprint's rank. T1 ammo is rank 1; jump freighters are rank 4000. For a better outline of what falls under what rank, see Appendix 1.
Additionally, the old Material Efficiency skill has been renamed to Advanced Industry, and gives a build time reduction of 1% per level. We are not totally happy with the reduction in skill value that’s happened here, and we are committing to revisiting this skill post-Crius to evaluate how to meet our overall goals here in a less dramatic fashion.
Obviously we need to ensure that pre-Crius blueprints still work post-Crius, so we will be altering the values on existing blueprints for the new system. The guiding goal with this translation is "no blueprint gets functionally worse" i.e. the materials needed for a single run should be the same or less after the transition. Do note that, because we are calculating percentage reductions on the job as a whole rather than per-run, there will be cases where multiple runs can realize hidden savings that weren't previously accessible.
Here is the conversion chart for old to new:
|5, 6, 7, 8, 9
|5, 6, 7, 8, 9
(For negative ME/TE, see the next section!)
COPYING AND INVENTION/REVERSE ENGINEERING
Team Game of Drones will be working on a more comprehensive overhaul to Invention and Reverse Engineering mechanics immediately after Crius, so the changes here are merely to keep these systems functional while we invest further work in them.
- Copying times are now expressed per-run, and are always 80% of base build time for the blueprint (thus generally shorter for ships, longer for modules)
- Invention only requires one run from a blueprint copy, and deducts it exactly as if you'd built from the blueprint
- Successful invention always outputs a maximum-run copy as a baseline, which is then modified by decryptors; max runs have been adjusted to be 1 for T2 ships and Rigs and 10 for all other T2 products
- Successful invention without decryptors now creates an ME-2%/TE-4% blueprint; material requirements for T2 items have been increased by 50% to balance this out as it means we’re no longer adding +50% materials due to negative ME, so that decryptor-less invention now requires 2% less materials than previously.
Decryptors have been adjusted to the new system with the TE bonuses doubling to reflect the fact that TE goes up to 20%. This leaves them in a somewhat unbalanced state, which will be addressed with the aforementioned upcoming Invention/Reverse Engineering changes coming after Crius.
To transition old blueprints, we'll be adding 6 to ME and TE for all T2 BPCs, and then converting them according to the above scheme. Thus an un-decrypted blueprint at ME-4, PE-4 will first become ME2, PE2 and then be converted to ME-7%, PE-14%. This is a one-time windfall to make the DB scripts for the transition simpler and more robust, so enjoy it while it lasts!
Slots are gone. There is no longer any limit on how much work you can do in a given facility. Anywhere you can start an industry job of any sort, both stations and starbases, is a facility).
Instead, we have dynamic pricing to give an incentive to not just build everything in Jita 4-4. This is explained as the cost to hire necessary workforce in a given system.
The base component of pricing is the system activity index. This value is tracked per activity (Manufacturing, Copying, ME research, TE research, Invention, Reverse Engineering), per system, so Jita will have a separate index for each of the six job types.
The index for a given activity-system pair is equal to the number of job-hours of that activity type in that system over the last 28 days, divided by the number of job-hours of that activity type in the universe as a whole over the last 28 days, and then square-rooted to give better numbers. This is then expressed as a percentage, which will tend to be in the low single digits, and can be all the way down to 0 at one end and (in theory) all the way up to 100 at the other.
system_activity_job_hours is the total duration of all jobs of that activity type in that system over the last 28 days, and:
global_activity_job_hours is the total duration of all jobs of that activity type in the universe over the last 28 day, and:
system_activity_index is the index we’re creating, the formula is:
This index is then used in conjunction with job value to create a base cost. (For those of you who want to do the actual math, we suggest taking a read of this devblog.)
Job value is based on the item or blueprint that represents the primary output of the job. For manufacturing, this is the value of the inputs required to build that output item, so for pretty much all build jobs it's the value of the job's input.
For copying and ME/TE research, this is fairly straightforward: it's 2% of the inputs for a job using that blueprint.
For invention and reverse-engineering, it's 2% of the average value of the inputs for the thing(s) built by the blueprint(s) you're /hoping/ to get out of the job. For T2 ships and T3 stuff generally, the output can be several things, so we average across all possibilities.
With this value (input costs for manufacturing, 2% of output's input costs for research), we then multiply by the activity index for the activity in question to get the base build cost. Example: let's say the manufacturing index in Jita is 5%, and you want to build an Apocalypse whose base materials cost 200m ISK. Base build cost is 200m*5% = 10m.
For ME and TE research, there's a further wrinkle, which is that the higher level jobs both take longer (as above) and also cost commensurately more. There is a cost multiplier for each level of research, equal to the time it takes to research to that level divided by the time it takes to research to level 1.
There are a range of modifiers that further affect the final price.
1. A reduction in price based on the stations and upgrades present in a system. Stations, Factional Warfare upgrades and certain outpost upgrades provide beneficial multipliers to costs. For any applicable job in a given system, the multipliers for all the stations and upgrades in that system are multiplied together with the base price. See Appendix 2 for a full list.
Also note: nullsec outpost upgrades which previously added manufacturing lines to the station, will instead offer a 1% material use reduction (i.e. an additional ME-1%) per upgrade installed.
2. Teams will increase job costs when they're used. This cost is a simple percentage increase, and will be listed in the team's information.
3. Facilities can impose a tax on jobs within that facility. For NPC-owned facilities, this is a flat 10%, meaning a 1.1x multiplier to the job price. For player-run outposts, this can be configured by the owners. For starbases, setting taxes won't be supported in the initial release but we are investigating supporting this in future after player feedback.
These multiply together with the base price to create a final per-run price for that job, which needs to be paid up front.
We've done a fairly comprehensive overhaul of blueprint stats, to try and bring them into a more consistent and interesting pattern.
Firstly, we've assigned all blueprints a behind-the-scenes rank, which dictates a lot of the values. You can see examples in Appendix 1, and here are the rules we used to determine them:
- Modules: Sized 3/6/9/120, multiplied by 2.6 for T2; unsized modules are sized S/M/L by low/medium/high slot
- Some modules have had their ranks adjusted to be more "interesting", principally cloaks, warfare links, mining modules and "specialist" modules that are generally intended for use on specific ship groups, such as interdiction launchers, doomsdays and so on. Still multiplied by 2.6 for T2 variants.
- Ships: Sized 20/30/40/50/60 for frigate/destroyer/cruiser/battlecruiser/battleship, with industrials sized as cruisers, and multiplied by 20 for T2. Capitals are 200, except for supercarriers at 400 and titans at 600.
- Charges: Sized 1/2/3/4, multiplied by 1.6 for T2
- Drones: Sized 1/2/3/40, multiplied by 5 for T2
- Starbase structures: Sized 20/30/40 for S/M/L turrets, 100/200/300 for towers, 40/60/80 for other structures based on CPU usage (40 is below 200, 60 is below 750, 80 is above that)
- Rigs: sized 5/10/15/20, multiplied by 10 for T2
- Everything else: assigned on a case-by-case basis
With a few exceptions outlined below, the combination of rank, classification and tech level determine a blueprint's stats.
Build times are generally rank * 300 seconds.
- T2 ammo is multiplied by a further x10, keeping their current batch sizes
- T1 capital ships are multiplied by a further x25
- T2 capital ships are multiplied by a further x4 (not cumulative with the x25 multiplier above)
- Components have not had their build times changed at all, as they're balanced around specific build chains
- Special edition ships (including all the skins) have their build time set to 10 seconds
Note: Special edition ships also have an attribute set so that they never cost any ISK to build - The New Eden Painters’ Union still isn’t on speaking terms with the Interstellar Industrial Workers Union.
Copy times are, as noted above, always 80% of base build time. Due to the interaction of various skills, copy and build end up having the same max throughput. We may revisit this in future to give copying an additional edge.
Research times are always rank * 105 for the first level, with scaling for additional levels as described earlier.
Invention time is slightly more complex. We wanted to line the base stats up so that, assuming a 50% success rate, research time would match up roughly with build time, so a max-skilled character can supply itself with enough T2 BPCs to fill its manufacturing slots. Yes, we know invention job length is problematic. See previous point about tackling invention more thoroughly once we're done with Crius!
Base invention time is thus:
- Time to build a max-run BPC of the possible job output(s)
- Divided by 2 due to the assumed 50% success rate
- Minus time to make the T1 copy
Thus, 2*(T1 copy + invent)=T2 build.
We've then played with the numbers a bit further, multiplying invention time for rank 3 or less blueprints by 0.9, and by 1.1 for blueprints above rank 7. This means "larger" items are more invention-bound while "smaller" items are more build-bound. Obviously, various bonuses available will mess with this balance; dealing with those is deliberately left as an exercise for the player :)
Max runs have been adjusted to be more consistent, based around build times:
- Most T1 items are set to ~2 days' production time
- Components are set to ~7 days' production time
- T2 ships and rigs are set to 1 run
- All other T2 is set to 10 runs
We then round, to the nearest 100 if the build time is above 1400, and to the nearest 10 otherwise.
Finally, build costs have (as noted earlier) been multiplied by 1/0.9 for T1, and 1.5 for T2, to balance out the removal of wastage.
THINGS WE AREN'T DOING
After considerable discussion, both internally and on the forums, we have decided that we are not going to award any additional compensation for blueprints currently researched past ME/PE 10. There are a lot of things feeding into this decision, including the strong precedent it sets, the fact that no functional value is lost, and the work involved in a one-time compensation deal that could be spent on polishing up the features we're shipping. We understand that some people will be unhappy about this, and we empathize with that, but we have to weigh everyone's interests equally and we believe in this case that the best thing for the game as a whole is to convert blueprints to the new system as previously described but not make any additional changes in this area.
We had intended to implement a bonus for having multiple similar facilities of the same type at a given starbase. This was in fact implemented in a draft state, but there was no easy way to make it both user-friendly and well-performing (it was only updating once an hour), and given the relatively limited upsides it provided, we elected to cut it and focus the time that would've been spent bringing it up to an acceptable standard on polishing the rest of the release instead.
The original design had a batch discount, whereby longer jobs had their job costs reduced. After further consideration we have also decided to cut this on the grounds that it offered fairly marginal benefit in the best case, had some counterintuitive implications (researching TE would reduce your batch discount), was challenging to explain clearly, and was at the end of the day adding complexity that we decided we didn't want or need.
We had discussed, in previous blog feedback threads, making adjustments to Gallente outpost copy speed bonuses. After slowly walking through the scenario, this turns out not to actually cause a problem. The concern revolves around the total throughput of T2 BPOs, and the impact of an increase in BPO throughput on the invention market given that total cost from a BPO undercuts total cost from invention. Currently they can get a 0.4x build speed bonus in Amarr outposts, so switching this out for a 0.4x copy speed bonus doesn’t impact potential throughput at all given that a blueprint can only be doing one thing at a time. Even if you copy in a Gallente outpost and then build in an Amarr one, you’re still bottlenecked by the copy speed. In practice, being able to take advantage of the speed bonus without having to manufacture in nullsec may increase the usage of this bonus, but balanced against that in practical terms is the need for a T2 BPO sitting in a Gallente outpost upgraded for copy speed, which is believed to be a somewhat rare occurrence.
This should comprise everything that is changing in these two areas (blueprints and pricing) in the Crius release scheduled for July 22nd. We will of course be monitoring and responding to this blog’s feedback thread, the issues and feedback threads for the release and the way use of industry features evolves on Tranquility in the months after release.
If you have questions, comments etc please raise them in the feedback thread!
That’s all for now,
-Greyscale, on behalf of Team Super Friends and Team Game of Drones
APPENDIX 1: common ranks
|All T1 small ammo, T1 small drones T1 mining crystals
|All T2 small ammo, T2 mining crystals
|All T1 med ammo, T1 med drones, subsystems
|All T1 large ammo, T1 large drones, T1 lowslot/frigate modules
|T2 med ammo
|T1 cap ammo, capital T2 components
|T2 large ammo
|T2 small drones, T1 small rigs
|T1 medium/cruiser modules
|T1 deep core/strip/ice miners
|T2 small/frigate modules
|T1 large modules
|T2 med drones, T1 med rigs
|T1 Warfare links
|T2 large drones, proto cloaks, T1 large rigs
|T2 med/cruiser modules
|T2 strip/deep core/ice miners
|T1 frigates, T1 capital rigs
|T2 high/bs modules
|T2 warfare links, Improved cloaks
|T1 cruisers, fighters/bombers
|T1 battlecruisers, T2 small rigs
|T1 battleships, capital components
|Medium T2 rigs
|T1 Capital modules, T2 large rigs
|Fancy capital modules
|Non-super capitals, station components, T2 capital rigs
|Supercarriers, T2 frigates
|Titans, station eggs, T2 destroyers
APPENDIX 2: Station cost modifiers
|Manufacturing output multiplier
|Research output multiplier
|Amarr Factory Outpost
|Manufacturing (Nullsec conquerable)
|Caldari Research Outpost
|Gallente Administrative, Minmatar Service Outposts
|Cloning (Nullsec conquerable)
|Factory, Shipyard, Assembly Plant, Foundry, Construction Plant, Biotech Production
|Warehouse, Chemical Storage, Academy, School
|Testing Facilities, Reprocessing Facility, Chemical Refinery
|Biotech Research Center, Research Center, Biohazard Containment Facility