Building new vs. expanding

What's going on in Kapi Regnum? This section is for everything concerning the game itself

Moderator: moderators

Guest

Post by Guest »

telyni wrote:What did the optimize function do in kapilands?
Create a mathematical aberration to avoid hitting the point of inflection for an exponential function with a fractional base.

Say you need a formula to represent how efficient your building is. We know it gets less efficient as it gets larger, and it would be nice to have something simple like this:

Productivity = employees * 0.99<sup>employees</sup>

So with 4 employees, it would be:

4 * 0.99<sup>4</sup> = 3.84

and then with 5 employees:

5 * 0.99<sup>5</sup> = 4.75

So each employee adds productivity, but not as much as the one before him.

Though the numbers are different, that's basically how it scaled in Kapilands (I have no idea here yet.) But here's where it gets weird:

100 * 0.99<sup>100</sup> = 36.60
101 * 0.99<sup>101</sup> = 36.59

As you get higher and higher, each employee does less to the point of doing almost nothing. And then if that's not bad enough, if you keep going the formula makes your building less productive with more employees.

Now you can push that point off by making the base closer to 1, using .999 or .9999. But eventually it will always happen.

What the "optimize" button did was reset the formula. Every 400 employees would then be worth a fixed value, and then you would add to it the value for the next 1-399 employees using that function.

As a result your production would really jump. I don't have time to draw a graph for you, but picture a set of steps that don't have right angles but rather kind of curve down. So it curves from steep to flat, then jumps up straight, then curves steep to flat again, then jumps up, etc.
Guest

Post by Guest »

cool i want that
Guest

Post by Guest »

So if that happens at 100 employees at that resolution, then that means that employees 101-400 aren't doing too much, even if they're not decreasing the production. Three hundred employees seems like a huge cost to pay to get to a point where adding more employees helps again. Is it really worth it to have buildings larger than 100 employees then?
Guest

Post by Guest »

Well, you NEED to have buildings with more than 100 employees, 100 buildigns with 100 employees are reached sooner or later.
Guest

Post by Guest »

telyni wrote:So if that happens at 100 employees at that resolution, then that means that employees 101-400 aren't doing too much
Did you miss the part where I said that was a hypothetical number to demonstrate the effect, not the real figure? I don't want anyone to draw the wrong conclusion.

It's actually closer to:
employees*(0.995112^(employees-1))

at least for lower levels.

However your conclusion still has some validity. Here's something I wrote in the Kapilands forum not too far back:
Knolls wrote:Here's an example taken from Kapitools. I am using a green power plant, and picked Merchant for no apparent reason..

398 employees (non-optimized) 17.251.103,68 per day
399 employees (non-optimized) 17.265.575,96 per day
400 employees (optimized) 18.189.473,68

As you can see, moving from 398 to 399 gained you 14.472,28. You would expect that moving from 399 to 400 would gain you around the same amount, say 14.450.

Which means optimizing alone gained you 909.450, or about 5.2%. Or to put it another way, the act of optimizing alone is worth more than 60 new levels (around that size).

So, if you have any buildings in the 350 employee range, I would think very hard about moving them all the way to 400 in a single shot rather than leaving them in that midrange for a while. Same goes for the late 700s, late 1100s, etc.
Again though I'm not sure yet how this translates into Regnum.
Guest

Post by Guest »

I just calculated how much is costs to expand a victuals market compared to building a new one. Very expensive.

Including opportunity cost for me it's over 288,000 for 30sqm. I hope I will make up for it in higher productivity with advertising but I have my doubts.
Post Reply