[PB37 SPOILER] Coeurva, Bacchus -- Cyrus of Carthage

[Bacchus]

Ok, because we need hunting and BW together and the order doesn't matter so much, what we need to try and do is research Hunting whilst we are making 9 beakers per turn, and switch into BW later, so that we leverage that extra beaker per turn from BW's +20% bonus for having a prereq tech. So try opening with Hunting research, leaving it 1t away from completion and then going to BW. We can then revolt, end turn, whip on the turn coming out of revolt and complete hunting concurrently.

[Bacchus]

If we have to delay the worker because of tech shortage, we can return to a faster build, slower growth scenario on the boat, it somewhat minimizes the delay on the settler to 1 turn, but will further hamper science:

Of course, not much fun in the capital ending at size 2 on turn 38 there. I also tried an alternative with no whips, which seems to work ok:

However, by now I tortured this spreadsheet so much that it may be full of bugs that I'm not seeing.

[Coeurva]

We have 4t at max commerce, which gives 19 bpt into BW after modifiers (lake / oasis / clams; 4*19 = 76); thus 193 - 76 = 117 beakers must have been invested into BW on the turn we grow to size3. The absolute best I can manage so far while also getting 1t Hunting on T20 is... 116 beakers.

Yeah, not kidding.



(I'm still trying.)

[Bacchus]

Don't forget to check out my slowbuild scenario though, if it's right we get everything faster and only need BW on T24.

1 beaker and 1 hammer shortages are supremely annoying. At least in the first 40 turn microplan you have the luxury of spending hours looking for them.

[Coeurva]

That plan seems fine. I think I've made a very similar slowbuild scenario for this in post #23 or thereabouts, which works for the settler ready on T28, cap size 2, 1t into gmine, and good tech (Wheel finished on that turn, Pottery in 9t), with the capital at size 2 and growing in 3t. In fact, it also resembles the "lol 116/117" plan in that it grows to size 3 before whipping, which is an idea I got from your early suggestion to emphasize growth (and commerce!) over hammers while building the boat. -- I didn't make a spreadsheet for it, but I can do so if you want.

Edit: Actually, the plan I posted had T32 growth, but I think I tinkered with it further. Let me leaf through these sheets and saves...

[Bacchus]

Nah, no need for a spreadsheet, I only use them in lieu of having access to the game. Which is why I suggest you check it through in the sim.

[Coeurva]

First scenario in post #52 seems to run into the same problem; not enough commerce to pasture the cow right away, no matter how I juggle BW and Hunting. The 1t saved on requiring BW for Slavery is exactly offset by the six turns of working a 2-commerce tile we lose compared to the other plan.

Second plan works re BW timing, but I think there's a mistake on T22: we'd have to be growing while building a settler at 18/24 food.

Edit: T26 has 61 hammers accumulated, 61 input, the second of which I think is a typo -- we get 16 hpt + 30 from chop.

If the known tech bonus for BW is 2% on total cost, not our own beaker output, that will be just enough to produce it in time with your first plan, provided that we meet someone going BW first.

[Bacchus]

Oh yeah, I forgot to delete the oasis being worked from some earlier version. Should add conditional formatting for blacking out tiles in excess of population. It doesn't change that much -- means the second warrior doesn't complete yet, so we can finish him off on T38 in lieu of starting the granary that turn. Oh maybe we finish him after the worker, so that he can get in place ahead of the settler.

[Bacchus]

There is no known tech bonus for ancient techs, sadly, I also looked it up today

[Coeurva]

At size 2, it's always better to work the gforh for 2t only, then work the oasis for 2t: this nominally delays the boat by 1t, but it's actually made up for by the accumulated food, so growth occurs on T14 regardless -- but with 2 extra raw commerce.

Also, I used my computer to actually compute something, and, well:

Code:

Hunting == 74    BW == 192    deviation == 9    solution: 0 0 3 2
Hunting == 80    BW == 186    deviation == 9    solution: 0 2 4 0
Hunting == 78    BW == 188    deviation == 9    solution: 0 3 3 0
Hunting == 76    BW == 190    deviation == 9    solution: 0 4 2 0
Hunting == 74    BW == 192    deviation == 9    solution: 0 5 1 0
Hunting == 78    BW == 188    deviation == 9    solution: 1 1 4 0
Hunting == 76    BW == 190    deviation == 9    solution: 1 2 3 0
Hunting == 74    BW == 192    deviation == 9    solution: 1 3 2 0

The "solution" vectors refer to turns spent on researching Hunting at 9, 11, 13, or 15 raw bpt (i.e. the lying figure in the interface), all others are assumed to be spent on BW. As can be seen, none of them is an actual solution and we're actually 9 beakers short in the best case, not 1 (1 short of BW, though). We have 1t @ 9 bpt, 9t @ 11, 4t @ 13, 4t @ 15 to spend.

"deviation" is sum[i=1, 2] ( abs ( target_beakers[i] - actual_beakers[i]) ).

I randomly picked the (1 1 4 0) solution to check if the program agrees with the actual game -- not exactly thorough testing, but one passed test case is better than none, right?

It is not possible at all to research BW and Hunting (minus 14 beakers for Hunting, which is our actual bpt after whipping the worker @size3 -- BW costs 193, Hunting costs 96) by T20 with the boat -> grow3 plan, even maximizing commerce. At least we can rest easy knowing that...?

Edit: The program does discount overflow, as I just noticed. Bwah. Might be the silver lining, but as min deviation is 9, I doubt it. -- Checked it with overflow taken into account. No dice.