On Saturday, I had my stream, and crashed
+Ali Aras's stream afterwards along with
Mynnna, Salpun, and a few others I didn't immediately recognize. Also,
+Chella Ranier dropped into my Friday stream (and I will have her segment isolated into its own video soon). It's been exceptionally helpful to get all the various perspectives on the industry changes, and we're still only scratching the surface.
If you're relying purely on the dev blogs for info about the changes, you're going to have a bad time. Mynnna showed me how grossly out of sync I was by missing important information about
CCP Grayscale's BP migration on the forum. Also, I almost missed some of the
POS stuff in the most recent devblog. Lastly, it's worth messing with the UI on all the devices you expect to play EVE on. It's a pretty big window, and I anticipate issues with shorter form factors such as laptops.
Note, none of the following are set in stone yet. Provide feedback if you don't like the changes! It's also worth noting the entire CSM is incredibly active on this Crius front. Feel free to reach out to your personal favorite(s) and include them in your feedback!
Invention Math Changes
There are three big changes for Invention that might have gone unmentioned elsewhere.
- Invention will only consume 1 run off a copy per attempt
- Invention times are far more dynamic to match up with manufacturing times
- Copy + invent = 0.5x T2 build time (roughly research = build for T2)
- Invention results will all be +ME/+TE. No more negative results
- Base materials will be increased to keep prices roughly flat
Both of these changes are pretty big for T2 producers. Reduced copy consumption clears up two problems for most inventors: BPC clutter, and reduce errors with T2 outcomes. Though copy-time vs build-time is being reduced, this results in a big increase for module copies. This ends up being balanced out by the one-run/invent feature. You will still need multiple copies for parallelism, but that supply should last for quite a while.
Personally, I find the invention time balance far more interesting. The goal is noble, to make invention time a far more meaningful choice, but I think this change misses the mark in its current state on Singularity. For periodic gaming, I imagine a few different time groups:
- <=1 hr: multiple per session, up to 6x/day
- < 3 hrs: once per session, up to 3x/day
- < 10 hrs: twice per day
- < 22hrs: once per day
- n-days - 2hrs: once per n days
Lastly, the ME/TE changes should make the T2 BPO tinfoilers jump for joy. Until the material quantities balance out (using a modified refining equation, causing weird material costs) and the SDE comes out to better run the bulk numbers, this will be a very hard thing to judge. BPOs will still have a slight advantage due to the requested runs cost scaling, but I think you will be able to get better ME/TE than most BPOs in the end. I'm particularly interested in how decryptors will drive the margins, since we probably won't see the big savings factors at 6-8% savings like we saw between -ME steps.
What I Don't Like
Personally, I only care if something is >2hrs, because I won't be able to do more than 1/day at that rate (without mobile app). Also, I care that jobs hit at 22hrs rather than 24, because it can be hard to keep up with a strictly 24 or 25 hour cycle time. Though the whole spectrum is interesting if you're going to optimize to share BPs over timezones, as long as jobs are a per-character mechanic, the times can be blocked into the above categories IMO.
Also, I'd rather that the invention time balance be 0.4x or 0.35x to account for specializing in the science skills. I'd like to put a tangible benefit for taking some of the obscure science skills to 5. Lastly, another "between the lines" message here will be wider decryptor-type use, because of the reduced research-time/run.
Currently POS math is broken, so I am not sure how this will wash out in the end. Also, I still need to play with some of the corners to check the balance. Lastly, the decryptor math is going to be something worth checking with a fine toothed comb. On TQ today, decryptors pay for themselves off the ME savings, and the difference between positive steps is much smaller than the difference between negative steps. I have a hard time imagining that better research/build ratios or lowered datacore costs will drive consumption.
POS Changes
There are a bunch of little things that changed between pre-Kronos and now.
- Additional labs/arrays will reduce job install costs
- "Invention" and "Research" labs have been split:
- Design Lab: able to copy/invent
- Research Lab: able to ME/TE research
- Hyasyoda Lab = meta Research Lab
- Lowsec love
- Intensive refining array = mid-grade null outpost refinery
- Lowsec-only Thukker component array
- ME better than outpost
- Built from ghost site materials?
- EDIT: moon mining/reactions will be available in 0.4 sec systems
As I covered previously, the people crying the biggest tears were solo capital producers. With the new POS array changes, the playing field becomes much more even between LS/NS. There's still the risk v reward metric, and a supply chain bottleneck for compressed ore, but the ability to participate in the capital game is still there (even if it's not entirely solo any more).
Also, I really like the flat differentiation between "invention" and "research" labs. This makes POS design far more direct for the amateur. It's still hard to judge the job-cost benefits (in general) because of math and UI elements not being better solidified yet. I'd also like to see the fitting requirements for both arrays be equivalent (or swapped) and a meta Design lab to match the Hyasyoda offering. With the BP@POS rules, I'd like the choice to utilize blingy equipment for both Design and Research labs.
Other Notes
The features are designed for an incredibly wide swath of EVE to participate in industry. Though I am not wholly convinced it's great for high-powered and cooperative industry, it's a big step up from TQ historically. Also, a lot of the bigger scale horsepower will rely on
+Regner Blok-Andersen getting CREST/API/SDE stuff out for developers to start moving some of the every-day calculations outside the client.
Lastly, I have this grandiose idea that job fees will be the bar that separates the men from the boys. Though right now, costs still account for <5% of business. I can see where taxes + fees start to eat your lunch, but it's still a death-by-inches mechanic, and that makes it hard to justify a lot of effort to minimize it. Also, with the job cost UI lacking a lot of the promised breakdown, it's hard to feel like it represents a meaningful choice yet. If the costs are going to be this complex, I need to have a fighting chance to make good decisions.
