SGOTM 12 - One Short Straw

[LowtherCastle]

mdy, what's your fastest Oracle-CS slingshot without Sailing? Have you tried building a Silver City also? I'm wondering whether we should build a wkr+settler or a couple of wkrs. Also, the worker turns are affected by whether we want a Silver CIty or different city.

I can get 960BC with and without Sivler CIty. I think I got a 1000bc without. Maybe even 1040bc is possible.

 

[babybluepants]

I highly doubt there is anything in the code stopping the AIs from reseearching PH and building the Oracle if we beat them to PH or put a hammer in the Oracle. My hunch is that that's a fairy tale.

Well, I did test this and get a pretty excellent result. I doubt there's anything in the code preventing it (Silu has told me that predicting AI research is basically impossible), but the odds of PH on the RNG roll must drop by a significant amount. From playing around with our test save, I know that availability of marble increases the odds substantially, but gets overridden by the 1h invested in Oracle, for instance.

Anyway, I'm ok with pursuing the fastest Oracle CS and foregoing GLH, though I think there's a huge possibility we'll have to abort to Currency (or even just lose it altogether).

 

[LowtherCastle]

bbp, explain how you tested the 1h thing again, please

 

[mdy]

mdy, what's your fastest Oracle-CS slingshot without Sailing? Have you tried building a Silver City also? I'm wondering whether we should build a wkr+settler or a couple of wkrs. Also, the worker turns are affected by whether we want a Silver CIty or different city.

I can get 960BC with and without Sivler CIty. I think I got a 1000bc without. Maybe even 1040bc is possible.

I've been settling south by the gems not the silver. With a very early third city the best I can get is 975B.C. With a later city my best date is 1000B.C.

 

[babybluepants]

bbp, explain how you tested the 1h thing again, please

I was previously using a file bcool prepared for 2400 BC. In it DG consistently built Oracle in 1520 BC (got this result 10 times over, in different test runs - I think you guys did, too).

(1) I gave myself Phood and put 1t into Oracle circa 1880 BC (don't remember exact date - it was before he completed PH, but after he started it). The result was 1520 BC Oracle.

(2) I gave myself Phood and put 1t into Oracle circa 2200-2000 BC (don't remember exact date again - this time before he started PH). He didn't complete PHood as of 1520 BC. Didn't take this one further, but I do have the save somewhere...

I've been settling south by the gems not the silver.

There are no gems in the real game, btw - we should fix this aera in the test save.

 

[LowtherCastle]

@bbp: You see, that can just be variation by random seed (I think). We'd need to find a test that test the mechanics without changin the randomness.

 

[mdy]

There are no gems in the real game, btw - we should fix this aera in the test save.

O.K. This won't make any difference to my best date as the city was founded too late to make a difference.

 

[babybluepants]

@LC,

Well, I said somewhere above that there must be a random aspect to it. Like I mentioned, I had quite a few runs where he built the Oracle, and only two where he didn't. The one described above and one where I took away his marble earlier on (in an unrelated test). Both seem like fairly strong correlations. I could run it more times and see if I come up with variations and/or some rough approximation of the odds involved?

How would you design the test? I don't know how random seed on reload works. Is it even on in our test game?

Or, maybe this is irrelevant, since we're talking about aborting to Currency as an option if/when either DG or Church complete a suspicious tech?

 

[LowtherCastle]

You're right that it's not crucial to know, I guess, but at the same time, it would be nice to konw if it's really a factor.

I understand that there's a random factor in the tech choice mechanics. What I was referring to was the random factor related to just changing moves. I think you get that, though. I'm also not sure about the random seed thing. I'd just like to know how to run a test that only tests the random aspect related to the tech choice mechanics. I don't konw how to do it.

---------------

Anyway, I guess our focus should be on how to go forward. Should the worker keep mining MOscow-S? It seems like a relatively useless tile (in the short term) if we're going to switch over to spawning a GS. If we succeed in getting CS, then we'll want to grow asap and get Calendar for the silks, I assume. Not sure how much we'd work that mine.

 

[bcool]

I agree that mine is largely useless now, sorry. I think clearing the rice would be the next priority or even moving the 3rd worker to help with fish city/silver city.

 

[babybluepants]

Skip the PH mine.

I think the biggest question right now is where to settle next. Doesn't seem like we have an obvious favourite site. I would prefer getting a settler and a worker pre-Oracle, and having the next city growing asap so it can contribute quickly. If we're just giving giving up on GLH, we may as well build a quick library in the fish city and work out a GS from there.

 

[babybluepants]

Ok, on my first attempt I settled the next city 3E+1N of the fish city, built an explorer WB after the first one, then library, got to 5 workers total and was on pace for 975 BC Oracle CS and 525 BC GS. I'm going oracle(grow to 5)-worker-settler-worker-oracle. Sounds maybe a bit worse than mdy's... The trade-off is GS for 2 workers?

Edit: moving that worker to 1S of Moscow was kinda awful, now that we're not following through with the original plan.

 

[LowtherCastle]

Okay. If we're going to abandon the hill mine for now, then I think I'd send that worker to the Fish City area. We can decide where to settle a little later. In particular, we need to see if dG finishes Stonehenge next turn, because that will expand his borders quickly and in one of my tests that border manages to keep creeping despite all the water its crossing.

 

[babybluepants]

Yeah, the border creeps in the test save. It's only relevant if we chose to settle the site I previously proposed (which we shouldn't now). Your GP farm would only lose a couple of hills and probably regain them quickly.

The worker on the desert tile should go straight to cow and help pasture it. The PH worker cancels actions, goes to road the desert tile, then starts on the ivory.

I think we should have a clear idea of settling pattern. I'm building two more workers before the Oracle and the sixth relatively quickly after, but still having trouble keeping up with settlers and growth. We'll need a tight worker schedule, with pre-roading and pre-clearing as needed.

 

[babybluepants]

Potential future cities:

(a) LC's GP farm (pig/rice/sugar/sugar/banana)
(b) Rice/Deer/Ivory coastal city (3E+1N of Fish city)
(c) Something with silver and deer
(d) Far south with rice and pig

My proposed sequence would be: (b) - (a) - (c) - (d)

The other longer term question is whether we wanna hunt for marble. That's a lot of effort and cost, but we should have a decent chance at the wonders in this game.

 

[LowtherCastle]

I think we should have a clear idea of settling pattern.

I absolutely agree. This determines whether ou upcoming worker turns will be very useful or totally useless.

Potential future cities:

(a) LC's GP farm (pig/rice/sugar/sugar/banana)
(b) Rice/Deer/Ivory coastal city (3E+1N of Fish city)
(c) Something with silver and deer
(d) Far south with rice and pig

My proposed sequence would be: (b) - (a) - (c) - (d)

How to maximize our REX in this fallout-rich scenario: 1) Work foodless hill mines at pop1 in cities starting with deficit health (so no food is wasted). 2) Acquire max health resources asap. 3) Build a galley+chariot asap and worker steal dG from the sea. 4) Get CS to spread irrigation to rice tile in Moscow.

1) Foodless hill mines: Silver City, Southern Pigs, GP Farm/Fish-2SW1S.
2) New health resources: Silver City, Southern Pigs.
3) Worker stealing galley: Fish-3W1N, Fish-2SW1S, Elsewhere.

A. Sailing next, I'd go: Fish-2SW1S - Silver1S - Southern Pigs
B. CS beeline, I'd go: Silver1S, Fish-2SW1S - Southern Pigs

Problems:
1. dG's culture growth may soon make Fish-2SW1S temporarily impossible.
2. bbp's Fish-3W1N makes Silver1S and 3 possible sites for a N/NW city forever impossible.

Thoughts:
1. I would seriously consider having the worker at Moscow-1S road to Southern Pigs (starting with Moscow-1NW) and not stop till its time for him to scrub the city site.

2. We need to unburden our workers: 1) roading, 2) scrubbing city sites, 3) scrubbing and improving resource tiles, and 4) scrubbing other tiles against . Our REX vitally requires 1), 2), and 3). But 4) can be done by connecting new health resources. Furthermore, connecting one new health resource replaces scrubbing two tiles in every single city. So our early REX must focus on acquring new health resources: deer, clams, rice, pigs, (later: bananas, corn).

3. We could build a scout immediately to better scout the nearby northern area. Our workers desperately need to know what to do, based on where we're going to settle up there. The scout would later be useful spying on Paris from Paris-2SE without causing dG to build defenders.

Edit:
4. If we build Fish-2SW1S, then we could build a fort on the hill at Fish-1S1SW and Fish can build galleys that can go from Fish straight to dG.
.

 

[LowtherCastle]

All right. This is taking forever. I'm going to play my last three turns and see what happens. I'm sticking with Math for the CS attempt. I'll be watching all the indicators.

 

[LowtherCastle]

Turnset T55-T57

Turn 55, 1800 BC
Conitnue researching math. Cancel worker actions.

Turn 56, 1760 BC
dG completes Stonehenge
Mao -6, 1pops his second city evidently
dG +16
Turn 57, 1720 BC
Southern warrior arrives at SE Land's End and spots clams next to the island and more fish to the west.
Gandhi (+1 Cautious) meets Mao (+1 Annoyed)
Churchill +5, London grows to pop6
Gandhi has connected copper and threatens with a spearman

Notes:
1. I'm ending my turnset here (only played 2 more turns). Over to Mysty or mdy. I've left two workers with work to do, depending on where we're going next. dG's borders will expand on Edit: T73 T74, I think. (Paris also gets 1 from the free monument)
2. That Gandhi is already Cautious with Mao is a bit of bads news for us. That means their hidden random values aren't making them dislike each other as much as we had hoped. This more or less guarantees that we'll be Gandhi's WE until he's at Cautious with us. The good news, though, is that Mao knows a lot of the same techs as Gandhi, so Gandhi will trade them to us at Cautious.

Here is your Session Turn Log from 1800 BC to 1720 BC:

Spoiler :

Turn 55, 1800 BC: You are the worst enemy of Churchill, De Gaulle, Roosevelt, Gandhi, Mao Zedong, Stalin the Despicable.
Turn 55, 1800 BC: Gandhi will trade Agriculture, Alphabet, Archery, Bronze Working, Masonry, Pottery
Turn 55, 1800 BC: Will Sign Peace Treaty: Churchill, De Gaulle, Gandhi, Mao Zedong, Roosevelt, Stalin the Despicable
Turn 55, 1800 BC: You are the worst enemy of Churchill, De Gaulle, Roosevelt, Gandhi, Mao Zedong, Stalin the Despicable.
Turn 55, 1800 BC: Gandhi will trade Agriculture, Alphabet, Archery, Bronze Working, Masonry, Pottery
Turn 55, 1800 BC: Will Sign Peace Treaty: Churchill, De Gaulle, Gandhi, Mao Zedong, Roosevelt, Stalin the Despicable
Turn 55, 1800 BC: De Gaulle has completed Stonehenge!

Screenshots and saves for T56-T57:

 

[babybluepants]

Ok, go ahead with 3t. No time to comment on settling stuff right now... We can deal with it after.

Edit: x-post

 

[babybluepants]

LC

Can't look at the save now. One minor note: why are we not in binary mode? The overall rounddown difference is pretty minimal, but still...

I'm assuming mysty can't play. mdy, are you available?