I have been quite busy with a new promotion, travel, and well... making a new profitability spreadsheet for my T2 calculations in Eve. I know I haven't posted recently, but here's to making up for it.
I have been deeply looking into both the T2 invention calculator and the Eve Blueprint Information site http://bp.kiwi.frubar.net/ and I have had difficulties with a couple of assumptions. The T2 invention calculator assumes that you buy everything off of the market, and it calculates for maximum income, not maximum return. (I know the first can be corrected with settings). The Blueprint information drawback is that you have to set all component BPOs at the same ME level, and can't mix and match. This is troublesome for ship invention, as ships won't tend to have 100ME. That, and I want to ensure I REALLY understand the profitability mechanics... I've been making my own spreadsheet.
I am almost done, and there hasn't been any grand epiphanies, but I do have a clearer understanding of what could potentially make more isk. So, I am spending tonight traveling around in my Viator and picking up a few BPOs.
This has also helped in prioritizing my science skill learning as I can now get a maximum increase for my lvl4 learnings. (lvl5 just doesn't do enough for the time spent. 2 lvl 5 skills give an additional 0.025% chance AT BEST more than lvl4s... assuming you're using a meta 4 item for invention. with no item, it's 0.005% more)
For those not familiar with the Eve invention formula, it's argued to be: (1 + 0.01 * racial encryption)*(1 + (Skill1 + Skill2)*0.1/(5-meta item lvl)) * base %
Base % is 40 for modules, 30 for frigates, 20 for other ships. In general, modules & ammo are going to be the most profitable, but very time intensive for the 1hr research times. Ships are slower, but not as profitable. I do a mix of everything.
Not all inventions are profitable, however. And, as in T1 construction, people don't always value their raw materials. The big culprit here is datacores. In order to obtain datacores, you need to research with an appropriate agent. However, some people see these as free, and don't value the opportunity cost of just selling them.
This means that some inventions actually make more isk if you sell the datacores and raw materials rather than inventing the BPCs.
Ex: Miner II invention
Datacore cost : 905k per attempt
% success with lvl4s: (1.04)*(1.16)*0.40 = 48.25%
# of attempts for 1 success: 2.07
average cost per attempt: 2.07*905k = 1.875m isk
Cost to make 10 Miner II @-4 ME = 2.85m
Sale of 10 Miner II = 2.24m
Profit: 2.24m - 2.85m - 1.875m = -2.5m (ouch)
This is where people tend to complain that T2 BPOs from the old days have an innate advantage and invention can't win. Which is true, but if you have enough isk, one could be yours as well, and people have paid dearly for their BPOs. Reality is, demand isn't high enough to overcome the BPOs... so just look elsewhere.