Micromanagement Sims/Planning

[Fintourist]

I agree there are many ways to think about these things and you can arrive at the same answers through different ways of reasoning. Personally I don't understand the distinction between food and surplus food. In my mind, each unit of food is equal to any other unit of food. Usually you have some food in the box too, so you are only loosely constrained by how much food you need to avoid starvation.

Just quickly. Assume you are working five 2-food tiles and a 2-food city tile. Your food surplus/growth rate is 2 food per turn. However if you are working five 3-food tiles. Now your surplus/growth rate is 7 food per turn.
--> A ~50 % increase in the absolute value of food gave you a ~250 % increase in your food surplus/growth rate.

Well basically what I'm saying is that food surplus and growth speed go hand in hand. And small increases in the absolute value of our food production can have a significant impact on our growth rate. That's why we are so interested in microing those extra bits of food into our cities.

[zakalwe]

Right, I see what you're saying. I guess I just tend to make growth an explicit part of the objectives I'm optimizing for, or just prioritize growth because it's the fastest way to reach my other objectives, anyway. When comparing game states to see which one is better, you can usually just look at the amount of food in the box. But what you're saying is obviously important to keep in mind when planning out tile improvements.

There is also the happy cap to consider. In many cases, a city is in practice stagnant, although it could be on a whip cycle. So you cannot just grow indefinitely. In short, the value of food can be quite variable, depending on the circumstances.

In general, and on a slight tangent, having a solid understanding of things like this is important so you can compare game states without actually simming ahead. It's like having a good evaluation function in a chess engine; that can compensate for several levels of lookahead, or so I've heard.

[DaveV]

Actual Edit: C'mon, really? I spent like 5 minutes formatting that with spaces because tab won't work and it just ignores them?

I always use [ code] [ /code] when spaces are important:

Code:

.         Column 1              Column 2
              1                     1
              2                     4
              3                     9

[kjn]

Updated sandbox.

While doing the updating, I noted that our in-game state doesn't match Seven's notes for T100. Horse Feathers should work the newly completed grass mine instead of a coast. The sandbox linked above matches the in-game state.

[kjn]

T101 sandbox

Should match the game state for all units and workers within our borders, except that we've missed the galley movement between HF and BbB, probably 2t back. Not that I think it will matter much. We can always WB in another galley and delete the first if needed.

[kjn]

Final T101 sandbox

Should match the game-state for all units within our borders, but for the galley outside Horse Feathers.

[Sullla]

I spent quite a few hours working on this today, and I'm fairly happy with this result:

https://docs.google.com/spreadsheet/ccc?...sp=sharing

Zipped save files for each turn.

Quick explanation/thoughts:

* I had us run a couple of turns of min science, then went Construction -> Code of Laws. Even if we don't get involved in a fight with CivPlayers, Construction is a pretty useful tech for crossing over the many river bends in the east, and if we're settling on the ivory, we surely want Construction soonish anyway.

* Assumed that we would win the 90% dice roll to spread Hinduism into Seven Tribes. If not, the Starfall missionary will head there instead.

* Built all three wonders in production on their planned finishing dates:

- Hanging Gardens: eot 104
- Great Wall: eot 105
- Mausoleum: eot 107

We will get a massive score jump from all of this, so now is a good time to work out any long term arrangements with other teams.

* A lot of cities are just barely keeping up with improved tiles, especially after landing Hanging Gardens. It's rather amusing to have so many workers still scrambling just to make ends meet. When I didn't have enough improved tiles on hand, I had cities whip pop away rather than work basic tiles. Examples:

- Forbidden Fruit swaps to a missionary and double-whips it for Eastern Dealers right after we finish HG, dropping off two unimproved tiles.

- Tree Huggers needs more workers badly; it grows onto two Scientists at size 8 and then immediately double-whips a worker to clear its own jungle.

- Gourmet Menu waits for religion to arrive in the form of a missionary, then swaps to lighthouse and double-whips it, overflowing into (and completing) a library.

- Eastern Dealers is the one place where we don't keep up with the rate of growth. Just too fast. As such, I have it grow to size 6 and then triple-whip a settler, headed for the spot to the west of The Covenant.

In other words, plan has us settling the ivory T108 and the German border city with rice/cows on T111.

* Left the earlier worker notations in place with red text, my changes and new stuff are in black. Most of the old stuff from Seven was still good, and I kept it in place. Main alteration was shifting a few actions near Forbidden Fruit to get the third grassland spice plantation (3/0/5 tile) done faster.

I'm sure that more optimization can be done here, but someone else is going to have to be the one to do it.

[kjn]

Going for construction next is highly approved by the C&D team.

[kjn]

Sullla, a simple request for your micro document: can you freeze the first column and the first row (View > Freeze rows/columns)? It makes it much easier to follow along on who does what when.

[kjn]

One small optimization in tile assignments for this turn (T103).

Gourmet Menu is finishing a worker, while Seven Tribes finishes the aqueduct. I think ST should work one of the FP cottages, while GM works the shared grass cottage. -1H in GM, +1F in ST.