Turn 2135-2137, Simulation data

Kody · September 1, 2003, 08:32

Using the new hive simulator, I did some testing playing our turn several years into the future. (much easier than mircomanaging every turn)

The main subject of this experiment was to gather data on whether we are better of building the VW immediately or delaying for a short period before attempting to build it.

My findings are as follows.

If the VW is to be built in 2137, at the start of turn 2138 we have the following.
_____________________________________________
Gross production - 13+13+4+6+7+14+9+12+9+4 = 91
Minerals in build queue - 7+13+4+6+5+14+17+16+4 = 86
Crawlers active = 11
Minerals in crawlers = 11*21 = 231
Total minerals (queue+crawlers) = 317
ECs = 40
_____________________________________________

With the VW built in 2138, at the start of turn 2138 we have the following
_____________________________________________
Gross production - 13+6+4+6+7+21+11+12+9+4 = 93
Minerals in build queue - 10+4+12+7+1+18+8 = 60
Crawlers active = 13
Minerals in crawlers = 13*21 = 273
Total minerals (queue+crawlers) = 333
ECs = 57
_____________________________________________

My estimate is by delaying the building of the VW for 1 turn we get a 8% industry advantage. I think I can increase that to a 10% advantage as I only spent 20 minutes doing each turn.

The raw data can be found here.

http://apolyton.net/upload/files/Kody/SPsimulations.zip

Jamski · September 1, 2003, 10:08

Its not worth waiting one turn for such a small advantage, is it?

-Jam

Kody · September 1, 2003, 10:19

10% isn't a small advantage.

Just think of it this way. We've been spending hours getting an extra 1% or 2% every turn by all the planning we're doing. Being able to get 10% in 1 turn is a big deal.

However, the real question is do the risks justify the benifits. The main reason we're building the VW is to screw up the university. It's not like we actually need it with our police rating and our industry. Rather we're building it more for a denial strategy.

PS: HongHu you did some good theorising on when you think the university will be able to build the VW. Can you write those ideas in here. I forgot to save all your ideas.

Jamski · September 1, 2003, 11:08

We have to compare 10% extra industry for us, against the Uni getting 2 less drones in every base, assume that gives them 1 mineral for each extra citizen that can work, over say 8-10 bases, then we're talking about giving the Uni 16-18 minerals per turn extra. Not to mention the extra energy, food...

And we will be building nodes eventually, and since we can't run Demo, we want every edge to get the GAs running to pop-boom.

Also, it only takes a turn to get out two more crawlers and get the minerals up to that level again. So as early as possible, reduce risks, hurt the Uni, and hopefully give use the benefits of the VW in the future.

-Jam

Kody · September 1, 2003, 19:08

That 10% extra industry means EVERY subsequent SP can be built 1 turn earlier.

The reason why I'm against it isn't only because we can get more minerals from the crawlers. It's also because the turn before I have to move crawlers around to squares that interfere with city workers so we're already significant losing resources the turn before we build the VW. Waiting 1 more turn means that we don't lose resources the turn before and we gain extra resources from those crawlers.

Edit: As for likelyhood of the university getting the VW, you should talk to HongHu.

Kody · September 1, 2003, 19:52

Jamski run the poll, I suppose.

Oh yeah just remember that due to pre-planning I'm actually making the moves to strip our industry over 3 turns. So compared to what you're normally used in respect to rushing SPs the "Can we build it now?" happens a turn or two earlier than what you are usually able to do with the same amount of resources.

Also delaying the VW by 1 turn opens up the possibility of getting either the ME or CN in 2139.

Anyway to restate the important facts.
- current turn 2133
- We can just build VW in 2136 (final preparations for build start in 2134)
- Or we can build VW in 2137 (final preparations for build start in 2135)

Anyway next turn (2134) we need to make a binding decision. Next turn's moves decide whether we will be able to build the VW in 2136, or whether we keep our resources invested in optimal gathering of resources for an extra turn.

Snowflake · September 2, 2003, 10:58

Oops, I wrote down my good theorising somewhere else. Or at least what I remembered. Hopefully they are still good ideas.

But, again, we can't let the univ get VW. So if we are not 100% or at least 95% sure, maybe just go ahead and do it without the 1 turn delay.

Jamski · September 2, 2003, 11:28

Quote:

Edit: As for likelyhood of the university getting the VW, you should talk to HongHu.

I think she's underestimating the number of bases the Uni have. Her estimate is based on the Uni having 6 bases of size 2, while I suppose that 8-10 would be closer to the truth, and we can assume that Googlie has seeded some mineral resources in the area, as he did for us.

Quote:

But, again, we can't let the univ get VW. So if we are not 100% or at least 95% sure, maybe just go ahead and do it without the 1 turn delay.

We're not even 50% sure, so that's why I'm sticking with my plan to build the VW as soon as possible.

BTW, what happened to proposals about getting the CN instead of / as well as the VW? Eh?

I'll make a poll, shall I, but I need to know if we want to consider the CN.

-Jam

Kody · September 2, 2003, 11:37

I feel like tossing a coin on this subject. In my mind the risks and gains are fairly evenly balanced.

If we weight the risks and the gains together for each plan they come out fairly even.

As for how useful the VW will be for us. The industry gain can easily cover the need to build drone reducing facilities. The part where we lose out is mainly in the denial strategy against the university.

I would say it's 10% chance that we lose it. However, the 10% industry gain is much bigger than people seem to realise. From someone that mircomanages our industry every turn you haven't see those minerals exponentially grow with futher mircomanagement.

To explain it with the extra 1% I've been scourging every turn for the last 10 turns I think we've increased our industry by 50% beyond what we would have if I hadn't. That might not make sense, but it can be best explained like a bank account.

With every day expensives you need to keep money in a normal bank account. However when you manage to get spare money you can put it into a term deposit that gives you 5 times the interest rate on what you usually get. The extra 1 percents I've been collecting have been going into term deposits and now we have a significant surplus of production.

Compared to the drones our current industry is actually 2 times stronger. And we're making much more energy credits than they are, and they're running freemarket. That 10% industry will have a lasting effect on our future growth which is just as big as if we had gotten the VW.

So really the only reason why we're building the VW is to deny the university it, and maybe for the prestige.

Kody · September 2, 2003, 11:38

Get online Jamski and remove the umm reference that has been carefully avoided.

Snowflake · September 2, 2003, 11:44

Well Univ only has 11 votes so it's unlikely they have 9 pop 1 bases?

Anyway I personally will weigh safty heavier just because I'm not good at micromanagement.

What is umm reference?

Kody · September 2, 2003, 11:45

Actually I'll probably talk to you tomorrow. Time for me to sleep.

I just don't like the idea of inserting significant inefficiencies to build a SP only 1 turn earlier.

Snowflake · September 2, 2003, 11:49

See how much worries can be eleviated if we've got our probe foil earlier? (Of course I'm just joking.

)

Kody · September 2, 2003, 11:55

Maybe I should clarify that the +10% difference is partly due to not having crawlers on our +7 mines. That is my main issue, losing minerals from not having the crawlers on those rocky mined minerals.

Jamski · September 2, 2003, 12:27

OK, poll is here

-Jam

Kody · September 7, 2003, 08:59

With strong mircomanagement I managed to improve on the 2138 outcome.

With the VW built in 2137, at the start of turn 2138 we have the following
_____________________________________________
Gross production - = 13+20+6+7+9+8+9+9+18+3 = 104
Minerals in build queue - 0+1+5+5+7+7+9+2+5+7 = 48
Crawlers active = 14
Minerals in crawlers = 14*21 = 294
ECs = 92
Potential rush buy = 92/2.3 = 40
Total minerals (queue+crawlers+ECs) = 382
_____________________________________________

With 2136 build of VW. Using old data
_____________________________________________
Gross production - 13+13+4+6+7+14+9+12+9+4 = 98
Minerals in build queue - 7+13+4+6+5+14+17+16+4 = 86
Crawlers active = 11
Minerals in crawlers = 11*21 = 231
ECs = 40
Potential rush buy = 40/2.3 = 17.4
Total minerals (queue+crawlers+ECs) = 334
_____________________________________________

So that's a difference of 14% between the two options.

However, if I re-mircomanaged the earlier VW option the value will probably reduce to 13% extra industry.

The previous industry estimates were improved because recently I figured out a way to get further turn advantage by constantly moving and rehoming crawlers in turns 2135, 2136 and 2136.

Basically the idea was to rehome crawlers so minerals were channeled like a lens. The rehoming was balanced so the amount of minerals tied up in our build queues were reduced and allowed for earlier crawler productions and increased stockpile energy.