Next Patch Suggestion

[Tomgs]

Drake I did it the way you said for the tournament but normally I don't bother with monilith facilities because I don't usually use an organic race so my consumption of anything other than minerals is very low. It gets me more minerals to build mineral miner III's so I go that way until I can research the resource converter at least. I very rarely run out of minerals before I have conquered the galaxy. But after the patch I might have a few longer games and need the extra minerals.

Also I have shipyards on 99% of my planets in a normal game so mineral miners III only take one turn to build normally. Once I research level III facilities a shipyard is the first thing I build on the planet. The 5 turns it takes to build are paid for after building the first 5 facilities then no upgrades are needed.

[This message has been edited by Tomgs (edited 09 February 2001).]

[Drake]

Good point. I've always taken organic, and I often get temporal too, so my resource need is a little more balanced. Mineral usage is still predominate, just not as much perhaps.

-Drake

[raynor]

quote:

---

Originally posted by Tomgs:
Also I have shipyards on 99% of my planets in a normal game so mineral miners III only take one turn to build normally. Once I research level III facilities a shipyard is the first thing I build on the planet. The 5 turns it takes to build are paid for after building the first 5 facilities then no upgrades are needed.

---

If you are going to leave the space yard on the planet, you *almost* break even by building the space yard first. But if you plan on scrapping it after you've built most of your mining facilities, then you will definitely lose resources along the way.

You just don't get enough resources from the Miner III vs. the Miner II facility.

[Tomgs]

I always leave the space yards on the planets. What use is a planet with nothing to build? Why not use those planets to pump out ships to use. Producing a lot of resources is useless if it just gets wasted (except in the tournament ).

And on a large or huge world you will definately more than break even. You get 100 per facility then when you add up the population and happiness bonus and any facility bonuses you come out a lot ahead.

Another point if you have minerals going into storage, or even worse being wasted, every turn anyway the expense of what you are building really "costs" you nothing.

[This message has been edited by Tomgs (edited 09 February 2001).]

[DirectorTsaarx]

Warning: you asked for a numerical analysis, so here it is...

First, the assumptions:
2000/2000/2000 Min/Org/Rad build rate
No population bonuses
No happiness bonuses
Upgrade costs 50% of new facility

Next, the simple part of the analysis: building the Mineral Miner Facility III's (MMF3's) directly, without mucking around in building lower-tech facilities first (and faster) and upgrading later.

We build N MMF3's, at two turns per facility (cost=2500 minerals, just a bit higher than the build rate), we get:

1000*[(2N-1)+(2N-3)+(2N-5)+...+5 + 3 + 1] minerals produced, from the time building starts until the turn the Last facility starts producing. This simplifies to:

1000*N*N

(standard mathematical formula: the sum of a series of numbers is equal to (n/2)(a+m), where n is the number of elements in the series, a is the value of the first element and m is the value of the Last element).

We then subtract the cost of the facilities (2500*N) to get the net gain. Pretty straightforward; net gain from beginning of build cycle through the end of the first turn all facilities can produce is:

N*N*1000 - N*2500.

Now, if we instead build N Mineral Miner Facility II's (MMF2's), at one turn per facility (cost=2000 minerals, exactly the build rate), we get (N-1)*900 + (N-2)*900 +... + 1*900 minerals produced, from the time building starts until the Last building turn; note that at this point, the Last facility has not started producing. I'll explain why in a moment. This simplifies to:

900*N*(N-1)/2

We then subtract the cost of the facilities (2000*N) to get the net gain so far. Now, if we upgrade those facilities (to MMF3's), it costs (1250*N) for the upgrade. Everyone still with me?

Obviously, we still produce minerals during the upgrade cycle; this amounts to:

1250*N/2000 [cost divided by build rate]

Technically, that number should be rounded up to the nearest integer to get actual number of turns, but we'll ignore that for a moment. In addition, this figure includes that first turn of production for the Last facility; that's why I didn't include it in the previous formula.

Now that the upgrade is finished, we can produce at MMF3 rates. In order to compare the "upgrade" strategy to the "build once" strategy, we include enough turns of production to equal the amount of time it takes to finish building the MMF3s from scratch. This amounts to:

[(2*N)+1] - N - (1250/2000)*N

Obviously, (2*N)+1 is the number of turns required to build the MMF3's; N is the number of turns required to build the MMF2's; and (1250/2000)*N is the number of turns required to upgrade MMF2's to MMF3's. Again, that Last number should be rounded up; however, in the interest of simplifying the algebra, I've avoiding the rounding. Which really gives a slight overestimate in the amount of minerals produced in the "upgrade" strategy, since we're now calculating mineral production as 900/turn for part of a turn, and 1000/turn for the rest of that turn. After combining the above formulas and doing some algebra, we come up with the following calculation for the upgrade strategy:

450*N*N - 2762.5*N + 1000

for net gain. In the same amount of time, the "build once" strategy gets:

N*N*1000 - N*2500

for net gain.

If we set the two formulas equal to each other, N comes out to slightly more than 1. Solving both equations for values of N ranging from 1 to 25 shows that the upgrade strategy is better if we build 1 facility and upgrade it. The "build once" strategy is better if we build more than 1 facility. The attached spreadsheet shows the calculations (if anyone's interested).

Comments are welcome - especially if someone finds an error in the math...

[Zanthis]

What, no analysis of building MMF1's and upgrading to MMF3!?

[raynor]

DirectorTsaarx:

Your numbers are off by so much, I can only assume that you do statistics for a living.

'nuf said.