SGOTM 12 - One Short Straw

[Mitchum]

FYI, I'm on a family vacation and will be away from civ until Monday. I'll try to keep up with the thread and comment when possible.

 

[babybluepants]

It might be worth getting currency with the oracle and actually saving gold (maybe even producing gold) for a few turns and using that to get peace with Gandhi with other techs other than currency. Then use math + some gold to improve relations, then maybe trading currency for Alpha.

Something to test, not sure how much gold it would take.

From my 1440 BC Oracle test, I need either Wtg+PH+90g (with no AI knowing PH) or Wtg+Medi+150g (with only DG knowing Medi). Then it took me Math+81g gift to get him to +2 Trade and Cautious, so I can sign OB and trade. He gave me Alpha+40g for Currency - for a net loss of 131g. We can get 29gpt without gold buildage, or 42gpt with Moscow building gold. Not sure how close this is to the actual game, though...

Edit: Trade relations is turn-dependent, so I got it down to Math+40g for +2, by building wealth for 42gpt. This gets me Alpha in 1280 BC, with a net loss of 90g. Not too bad, I think.

 

[babybluepants]

Spoiler :

IIRC it's FAIRTRADE_VALUE / (TURNS_KNOWN*10), where a gifted techs' fair trade value is approx 1.5x beaker cost (varies with how many Civs a Civ knows that know it). We've known him for what, 20 turns? So the threshold for +4 would be 200 fair trade (~ 200 gold), or a 160 beaker tech. PH + Medi give 210 EDIT: plus the map multiplier (1.3x) at least, not 100% sure if handicap multiplier (1.2x) is added, so at least 273, which is cutting it close... Need more calculations to know if those alone would give +4. Math would certainly give it.

Okay, the value of techs for an AI in a trade:

((COST - RESEARCH_DONE) * 1.5) + ( COST/2 * ((CIVS_KNOWN - CIVS_WHO_KNOW_IT) / CIVS_KNOWN)) * AI_TRADE_MODIFIER

If this applies about the same in our current game, the cost of +2 Trade Relations increases at a rate of approx. 10bpt from T63 (1480 BC) to T75 (1000 BC), starting at about 650b. In testing I achieved +2 with Math+40g on T67. IDK how cash gets valued here, but this doesn't seem quite correct. Maybe one of the multipliers is wrong...

If Gandhi meets another AI, the cost immediately drops to 85%, with 2 AI 82% (so more beyond the first one don't make too much of a difference).

If Silu or ZPV are around... Could we get some precision on this?

 

[LowtherCastle]

EDIT: Humm... I now double-checked the fair trade diplo code (functionality of previously given to me/us by bestsss) and actually the turns are only multiplied by five for fair trade and ten for WE hate.[/B]

Does this explain the discrepancy?

EDIT: Btw, I'm pretty sure the *1.5 in that formula is for epic, so you would also leave that out here. No, I don't think so. It got calculated already when he was figuring out the tech costs.

 

[babybluepants]

Yep, those give me 418b for T68 when I got it with Math+40g. Sounds very close.

Edit: Not sure I can figure out the exact cost, but we know what it approximately takes. Math > Currency def. makes sense to me now as a diplomatic move, and we'd just have to be very careful to minimize the costs of DoP/OB in the post-Oracle turns.

 

[Mitchum]

Agree. Math -> Currency makes a lot of sense. Not having to self-research Alphabet is great!!

 

[LowtherCastle]

PPP T48-T58

Research: Medit-Phood-Math
Moscow: worker-Oracle
Fish: wb

wkr1: scrub Fish-cows
wkr2: scrub goldN-road-mine-scrub ivory(?)
wkr3: mine goldN-road horses-roadMoscowE-scrub rice

Unless I'm missing something, the MM is a no brainer this turnset. Moscow works best tiles throughout, giving up the gold to Fish after Phood completes. Or?

warrior1: Continues to explore the north.
warrior2: Explores SE. If land ends, explores south of dG.

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

Question: Where do we want to settle our galley-producing city? Choices:
1. dG-Pigs-E (bbp's site)
2. 2SW-1S of Fish
3. Somewhere to the north

The reason I bring this up here is that wkr2 might go elsewhere than scrub the ivory.

Choice #1. Like bbp, my first thought was #1. But then I saw the +16f GP Farm site at North-Suger-1NE. This site also has tons of hammers for building infrastructure, including the NE. It is true that this will take some time to develop, but #1 is no better in this respect because it probably won't get the pigs before we capture Paris anyway. There's no guarantee whatsoever we'll find another 16f site. None of the AI capitals appear to have that much food, including Paris, judging from the Demographics screen.

Choice #2. This is a garbage city, its purpose is just to build a couple galleys asap. The benefits are that it can use the ivory tile immediately, while waiting for the plains hill 2S to get mined. It would also enable us to farm the rice in advance of settling the GP Farm site. Another benefit of this garbage site is that it doesn't stop us from settling any other good sites. If we settle to the north before we've fog-busted thoroughly, we might regret it. Eventually, we can gift this garbage city to some AI to keep him alive or for whatever other reason.

It may seem wasteful to build a garbage city just to build galleys, but imo, capturing Paris is that important and since we imperilialistic, a settler is our cheapest option (cheaper than workers + worker turns to road around), and cheaper in the long run than ruining a good city site.

Thoughts?

 

[babybluepants]

The pig's in his third ring, and we're CRE. Plus, we can scrub it before it culture-flips.

I definitely wouldn't settle the north anytime soon.

Regardless of where we settle, we want ivory for the happy and for prod in Fish. It's on the way to most probable sites anyway. If we go for a settler right after Oracle, we are a touch short on worker turns, though.

I'd road the horse before heading to rice. Doesn't cost any extra turns.

warrior2: Explores SE. If land ends, explores south of dG.

According to my test from this morning, we might have Alphabet by T+20. From current position, this warrior would need 9t to get within 2 tiles of Paris for DoP.

 

[LowtherCastle]

Alternative PPP T48-T58

Research: Medit-Phood-Sailing-Math
Moscow: worker-Oracle(to pop5)-worker(partial)-Oracle (not part of this turnset anymore)
Fish: wb-lh

This gets the Oracle done 3t later, 1320BC, but keeps open the faint possibility of getting the GLH. I say faint because Mao is exploding (he already has his second city btw). But maybe he'll start the GLH in some tiny city. On the other hand, my indicators are that Stalin is putzing along at 1hpt so he's not likely to succeed with the GLH even if he tries. So we give it a shot, I guess. Since we now know better how to get trading going with Gandhi, maybe we can be a little flexible on the ORacle date

More data:

dG will need about 5t to complete the Oracle, best-case scenario after reseraching Phood. Churchill would need 7-8t. So we'll need to keep an eye on that.

 

[LowtherCastle]

The pig's in his third ring, and we're CRE. Plus, we can scrub it before it culture-flips.

He's been putting 7cpt per turn into his piglets since T29. If we settle it on T75, we're already at least 322 behind. Then on T86 he adds 20cpt. Hmmm...

I'd road the horse before heading to rice. Doesn't cost any extra turns.

Okay. But how does it not cost 2 turns?

According to my test from this morning, we might have Alphabet by T+20. From current position, this warrior would need 9t to get within 2 tiles of Paris for DoP.

Wouldn't we rather send a new unit to dG and let this guy go find Gandhi?

 

[babybluepants]

He's been putting 7cpt per turn into his piglets since T29. If we settle it on T75, we're already at least 322 behind. Then on T86 he adds 20cpt. Hmmm...

Hmmm...
Yeah, you're obviously right... I generally like to do this with CRE, but I forgot it would be 1000 BC already, against a holy city.

I'm not sure in that case. I kinda like Rice/Deer/Ivory NW of Fish, for sharing a good tile and potentially getting galleys out right across from Paris later.

Okay. But how does it not cost 2 turns?

Sure, but we won't have Agri in time for rice anyway. Depends on whether we might want a chariot at some point before that worker's done all its actions, I guess (like as an option to grow to 6).

Wouldn't we rather send a new unit to dG and let this guy go find Gandhi?

Ok... Well, you should play that by ear. Any exploring is obviously good.

 

[LowtherCastle]

I'm not sure in that case. I kinda like Rice/Deer/Ivory NW of Fish, for sharing a good tile and potentially getting galleys out right across from Paris later.

That's also on my dotmap of possibles. I see two drawbacks for now: 1) No plains hill--at pop1, all these cities are going to be running -2f from unhealthiness, so we really want to be working a plains hill mine. That's down in the range of #1 and #2; and 2) we haven't completely defogged tiles near Fish-NW. ANyway, what you said was right, we need that ivory camped asap anyway, so that answers my concern. We'll cross the bridge of settling our next city later on, right?

Sure, but we won't have Agri in time for rice anyway. Depends on whether we might want a chariot at some point before that worker's done all its actions, I guess (like as an option to grow to 6).

True on Agri. My latest thinking is maybe to send the third worker over to Fish after roading the horses. Then the next worker or two can worry about the rice tile. If we're serious about the GLH, Fish has to get scrubbed fast so it can grow fast. This also allows us to let one worker road to and scrub Galley City, if we want to. I ran one attempt and it was timed perfectly with the Sailing+Oracle-wkr-settler variant.

On the other hand, there's an entirely different possibility which is pretty good, I think. The third worker roads horses, then mines the grass hill at Moscow-1S. Moscow could work this tile instead of the goldmine, after Math. Then Fish would have more hammers at pop3 and grow a little slower, which may be the right move till it gets scrubbed more.

 

[bcool]

Unless I'm missing something, the MM is a no brainer this turnset. Moscow works best tiles throughout, giving up the gold to Fish after Phood completes. Or?

In my earlier oracle tests I was working the silk tile with Moscow 2 turns after the 1N gold mine is finished so it grows to pop 5 the same turn fish city would give that tile back to the capital in favor of either the fish or the cows 1N of fish city.

 

[LowtherCastle]

Btw, I think I've cracked the code. If I had all of the saves (EDIT: actually, I think I could do it even without the saves form teh first 12t), I'm pretty sure I could tell you just about everything there is to know about the AIs MM. Obviously not what on their tiles, but when they made workboats, warriors, other units, how much they're putting into research, epps, the whole lot.

The key is that on the demo screen we can not just adjust our own GNP but also our hammers, food, etc. So before we thought we didn't know exactly how many AIs were running 1, 2 or 3, hammers, but we could have known exactly. FOr example, the T48 save shows this:

GNP     Hammers  Food
----    -------  ----
1ai@18  1ai@1h   1ai@8f
1ai@21  1ai@7h   1ai@12f
1ai@22  1ai@9h   3ai@14-17f
1ai@23  1ai@10h  1ai@17f
1ai@24  2ai@12h  
1ai@27

The one factor that we wouldn't have enough info about is food, because our max food is less than 4 of the AIs. This means we would have some trouble homing in on the food-hammers AIs were putting into workers and settlers.

Now, the use of this would be that at this point we would have a fairly good idea of what AIs have built so far, where they're at in their general development, and maybe an idea of where they're going. We'd also have a decent idea of the seafood they have, how much food they have, maybe even stuff like whether they got marble or not.

 

[LowtherCastle]

In my earlier oracle tests I was working the silk tile with Moscow 2 turns after the 1N gold mine is finished so it grows to pop 5 the same turn fish city would give that tile back to the capital in favor of either the fish or the cows 1N of fish city.

Did FIsh give it back the turn the wb was finished or the turn the nets were made? In mine, I start working the mine immediately and Moscow grows to pop5 the same turn the nets are made, which is a turn later now that we work the lake fortwo turns to get Phood done sooner.

 

[babybluepants]

The key is that on the demo screen we can not just adjust our own GNP but also our hammers, food, etc. So before we thought we didn't know exactly how many AIs were running 1, 2 or 3, hammers, but we could have known exactly.

You're willing to track this for 48t?
Sure, I'll post my saves later tonight...

 

[LowtherCastle]

You're willing to track this for 48t?
Sure, I'll post my saves later tonight...

Not promising anything...

You can just pm them to me, to reduce the clutter on the thread.

 

[bcool]

Quote:
Originally Posted by bcool
In my earlier oracle tests I was working the silk tile with Moscow 2 turns after the 1N gold mine is finished so it grows to pop 5 the same turn fish city would give that tile back to the capital in favor of either the fish or the cows 1N of fish city.
Did FIsh give it back the turn the wb was finished or the turn the nets were made? In mine, I start working the mine immediately and Moscow grows to pop5 the same turn the nets are made, which is a turn later now that we work the lake fortwo turns to get Phood done sooner.

Sorry wasn't taking into account the lake turns.



autosaves (would have had to post them somewhere to pm them to you anyways)

 

[babybluepants]

Saves attached...

 

[Mitchum]

I'll try to post mine tomorrow if you need them. But aren't the demo screenshots enough for what you're trying to do?