# Finding Optimum

Thursday, 11 May, Year 9 d.Tr. | Author: Mircea Popescu

Experience provides datapoints.

Datapoints can then be interpolated, thereby producing a function (on the left, blurred).

The function can then be drawn, and its inflection points thereby found / calculated.

This process -- the turning of functions into interesting statements -- is called functional analysis.

Displayed on the right, such an inflection point, which happens to also be a (global) minimum.

The interpolation-analysis binome often yields practical results of marked industrial importance. The case here is no exception :

mircea_popescu diana_coman hey listen... i've been doing a ton of math... turns out there's no dispute possible, best efficiency has me making q ~50 ish cfti. is that outright unusable for you ?
diana_coman ugh, q50 out of grass q50 or what? I can certainly use *any* cft but it kind of makes it even harder/more expensive to get reasonable q on non-basics

mircea_popescu nope. you will get a much better deal, both for bvii and %iii. lemme finish the run i'll give you details.
diana_coman the issue re non-basics remains anyway. basically you don't have bps high q enough for getting the same gain with q50 or what

mircea_popescu which was that ? they expensive ?
diana_coman on non-basics for one thing I can't use as MANY cft as on basics

mircea_popescu ah there is that yes.
diana_coman for the other yes, any reasonable q on those will be more expensive, yes, and their bv is higher to start with, so I suspect it will be *anyway* harder to get same q. you know, piling on the hard basically.

mircea_popescu though honestly, i been doing the math, and it seems to me i actually want q 1 non-basics.
diana_coman lolol

mircea_popescu ikr x.x
diana_coman sounds weird though because they also pack more bv so kind of pushing up the q of output; I suppose if you go however of undercraft anyway then yeah

mircea_popescu this is the problem, you put the finger on it : because i have high skill and i lift q's, it is VERY hard for natural q lifters to compete with me. so many crafts are actually most productive with lower q inputs than i had been using.
diana_coman so you say you are going to go for undercraft mainly?

mircea_popescu so it appears.
diana_coman basically pushing q up on skill + bp q mostly. hm. I mean: I KNOW it is essentially money printing, guaranteed etc but I suspect that overall it brings all qs down pretty much.

mircea_popescu CERTAINLY on grass this is the better play, i even drew the function on graph paper.
diana_coman ahahahah

mircea_popescu picture me surrounded by piles of graph paper. now, other things, disputable, esp as you get way out to tool bps etc.
diana_coman grassy graph paper, lol

mircea_popescu lol.
diana_coman actually noob ate all the cft already; any of the new one ready?

mircea_popescu o.O ok gimme a moment here to maff.
jhvh1 mircea_popescu: 6*9999*60*0.5 = 1799820iv

mircea_popescu you can either have 55413 q 36 or 34398 q 62. which do you prefer ?
diana_coman I suppose at this stage I'll take the q36 - to last longer at least, supposedly

mircea_popescu alright. i have two stacks ready, rest you can hit me up for in about 10 hours. here's the beauty though : 142%.v
diana_coman nice; will have to re-calc and see anyway wtf this gets me

mircea_popescu welcome to eulora, the sad land of recalcs. anyway, in bv terms, which is what we think matters, or is it ? this is actually larger than half a stack q 120
diana_coman kind of curious what does the noob get with this anyway

mircea_popescu for srsly. oh, also in the news : i want grass q 15 from now on.

In other words : familiarity with math of some of the people involved significantly reduced the costs of technological processes to the benefit of everyone playing the game. La ce imi serveste mie radicalurile ?

———
1. I had been making q 125 before. []
2. Base value. The concern is that if you have an economic cycle whereby you get out less base value than you put in, your total available supplies will dwindle to nothing over time. In order for a cycle to maintain itself, it has to have some means of increasing BV. []
3. Items in Eulora have a Base Value (ie, what Minigame pays for them, in any quantity) and a player-negotiated value, which, considering utility, rarity and so forth essentially represents a premium paid over base value. Blueprints, for instance, being difficult to get and central to all production carry 1`000% or thereabouts -- meaning that to get 1 ECu's "worth" of blueprint you can expect having to spend 10 ECu in cash. Other items, such as basic harvestable resources, carry a lower premium, maybe 200%. The most recent official auction saw goods with a base value of 6.244 mn sell for 63.4mn - about 1`015.37%. []
4. She had sold me ~1.8mn worth of grass BV. For the discussed reasons, it behooves me to push back her way at the very least that much. As it works out, I have available almost four times as much, which is why she gets her choice of either stack below. The first one's worth 55413 ×.36×180 = 3.59 mn ECu base value, the second 34398 ×.62×180 = 3.83 mn ECu base value. Either is a viable choice for her, depending what quality / quantity mix she prefers.

Formerly, I'd have turned the same base value of grass into about 20k CFT q 125, for a total base value of only ~4.5mn. The significant increase in base value output allows me to both put some threads on the auction, which is good for independent miners and the economy generally, and also reduce asking price, which ditto. []

5. CFT prices took an utter nosedive, from 250% to 200ish to under 200 to 142% now. It's been driving a slow but certain drop in CDG prices, from 300% to 215% and perhaps even lower once this latest price decrease propagates through the technological pipeline. []
Category: Trolloludens