Friday 23 May 2014

Balancing Freighters - why a nerf is very hard to avoid, the rule of three

Hello

A quick post about the freighter rebalance:

A lot of folks have called the proposed change a nerf, and with good reason.  These ships *have* been nerfed... but I'm not sure it was avoidable.

First, the goal of CCP was to make these ships more customizable, without significantly boosting their power - they didn't want to see a 500K ehp freighter or one with 3 million cubic metres of cargo.   This seems reasonable.  So a nerf accorss the board was needed to reduced the post-rigged stats.  But what stats?  A freighter has 3 important stats:

- Cargo space
- Tank
- Align time

(There is also warp speed but let's ignore that one for now as I believe the above 3 are more important.  You'll see my argument is still valid with those.)

So in theory, if you had one cargo rig, one tank rig and one agility (ie align time) rig, you would have about the same results as before... but that can't work!  If that was the case, having (using cargo as an example) a ship with 2 cargo rigs would have had a fair bit more of cargo, and 3 cargo rigs would have been massive.  So the only way this was tolerable for CCP was to have the nerf so bad that only by putting 3 rigs in one stat would you exceed previous capacity, at a significant cost to the other 2. 

This design space of "3 stats, 3 slots, don't buff too much" leads to an almost inescapable nerf. 

You can see the thread here:  https://forums.eveonline.com/default.aspx?g=posts&t=345280  I'll note that Fozzie wisely chose to go away from rigs and toward low slot modules, and I haven't had time to crunch the numbers yet... but I think the fundamental challenge remains.

Ironically, I'll note that having *less* slots to work with might make balancing easier... 

Sunday 11 May 2014

Industrial Griefing - number crunching

In my last post, I mused if it would be possible to intentionally ramp up installation costs in a system to then ask the local industrialists for "donations" to make you go away.  Let's crunch some numbers and see if this makes sense!

First, we need data.  There are about 7500 systems in EVE, but only 1212 are (based on my research) High-Sec systems.  However, based on a recent Dev Blog, about 85% of jobs are in high sec.  We also know there are about 50 000 industrialists (see here).   A big unknown in this analysis is how "lazy" industrialists are - are they running at 75% capacity?  50%?  95%?  I don't know.  Because of this, I'm just going to assume that everyone involved is equally productive.

If we assume that the average job takes about a day, and that the average industrialist character has a 10 job capacity, this means that about 500 000 jobs are run per day, and of those 425 000 are in highsec.  This results, if the job were spread out evenly, in about 351 jobs per system.... meaning there are about 35 industrialist toons per HS system.  That's not an insignificant number.

The complete job instalation cost equation is quite complex, but I will focus on this part:

Instalation cost  = cost of product * square root of fraction of global work hours.

Here I am again forced to take a short cut, assuming that all jobs have the same number of hours.  Clearly this is not true, but we will have to tolerate the approximation for now.  This means that the formulae can be changed to

Instalation cost  = cost of product * square root of fraction of jobs.  

So let's take our perfectly average system with 35 industrialists.  and do the math.  The cost I calculate is 2.6%... which isn't far off from the 2% cost noted in the dev blog that would happen if all HS jobs were distributed equally - I'm on the right path!

So how much is that?  Well if an industrialist is making simple t2 modules that sell for say, 0.5 million, and he makes 100 per day... that is about 18 million per week.  Of course, people who make larger items may be paying more.  But it doesn't seem that big a cost.  If you were doing t2 manufacturer and an extra 18 millions is going to sink your business... you're doing it wrong!

Of course, an average player may have more than 1 toon doing industry.  For ease of math, let's assume that the average is 3.5 (seems reasonable no?) and thus we have 10 industrialists *players* (not toons) in our very average HS system.  This installation cost is now adding up to 63.7 per week, or about 250 million a month... that is starting to add up isn't it? 

Ok ok, so many numbers... what about the evil griefer?   Well let's say he comes in, 2 accounts, 6 toons, doesn't need a POS and starts building cheap crap to drive up the activity.  It's on a 28 day average, so he's going to have to do this for a while for people to notice.  How much damage can he do?  60 jobs a days, the activity in system goes from 351 to over 400.   The installation cost has gone up to 2.9%... meaning an additional 29 million cost per player.   Incidentally, the busier the system is the smaller the change a small griefer can accomplish.  In a system with 1800 jobs/day, the griever can barely increase the installation cost % by 0.1%

So could someone get small sums from several players to go away?  Sure.  But the amounts are small and this doesn't really seem viable.  Then again, the miner bumpers/gankers of the New Order sell mining permits for 10 millions isk, so perhaps a new form of griefing will indeed be born...

Will I be responsible?  Maybe, but someone else would have thought of it eventually...

P.S. Lastly, I'll take a moment to express my frustration that CCP has *still* not fully released the information about how POSes will work now.  Given this number crunching, I'm not really sure POS industry will be very viable (vs just using a station) in the future...