SGOTM 12 - One Short Straw

[bcool]

Getting Mao down to 3 cities in time to steal Communism means a macemen/treb war I would think.

If the plan is to beeline Scientific Method, we wouldn't have steel, and thus wouldn't be able to use cannons to conquer Mao.

 

[bcool]

Stealing Steam Power would also be nice... Capturing Mao cities might mean we get an intelligence agency or 2...

 

[babybluepants]

Obviously the epps are outrageously better, but there are two constraints:
1) We have to produce enough epps in time to get the tech when we want it.
2) We have to have Mao down to 3 cities in time to gift him Moscow-2S1SE.
3) (In other games, there is an additional constraint: some AI has to have the tech you want. Mao already has it for us. )

Fair enough. That's about what I said above. Calculating the opportunity cost of constraint #2 is quite difficult at this stage, though, as our cities are underdeveloped (IMHO) for significant war effort. Their rate of development and subsequent gains are dramatically reduced, which is why I've argued so much against taking on Mao pre-Steel. Overall, the problem with making a move like this on constraint #1 is the level of commitment required and the associated lack of flexibility in game play (specifically our ability to use/develop cities as we see fit at any given point). I don't accept that we can pre-calculate the relative benefits/costs accurately enough to make that determination just yet. Which is why I suggested focusing on DG or Churchill for the steal, as that gets rid of both those constraints.

The loss in, say, Libbing Communism is that we'd need to then self-research Bio, while gaining in ideally not self-researching Nat or Steam (mind you, I don't understand this ep business enough to know whether this is a real possibility - Steam would happen quite late, for instance, after we crush Churchill with cannons and when we have the ability to accumulate more ep without sacrificing much else - it's also on the Mining Inc. beeline, btw). We do need Liberalism first to unlock Communism, so it's technically impossible to actually get it faster via espionage. Relatively speaking, State Prop is almost infinitely bigger for us than Bio, since we have few farms to work. Really, I think Steel is much bigger than Bio in mid-game development.

I'm not saying you're wrong at all, just would like to understand the options.

Edit: x-post(s) with bcool Edit2: and mdy

 

[bcool]

Why won't they trade construction to us? Is it just that enough other AI don't know it yet? Or maybe it has something to with us having Ivory? If we gave away all of our Ivory would they be willing to trade it to us?

 

[babybluepants]

@PPP

• Courthouse in Moscow? Can we just build an extra Mace? (This after all the EP talk )
• Moai in Cuba? It's fine, but I thought we should at least consider some alternatives.
• Whipping the WB in Cuba asa pop3 (to not waste turns working seafood) might be ok, too, as long as it helps Bahamas with the Uni build timeline (maybe it doesn't?). I think I'd send the WB there, instead of Bermuda, as we have sheep + pig over there.

 

[babybluepants]

Why won't they trade construction to us? Is it just that enough other AI don't know it yet? Or maybe it has something to with us having Ivory? If we gave away all of our Ivory would they be willing to trade it to us?

I doubt that enters the calculation. Not enough AI, as: We are not ready... We probably need 1 more AI with it, I would think.

 

[LowtherCastle]

@bbp: I feel like this is a moment to be frank, so we don't go back and forth about this too much. Listen, I wouldn't put all this effort into explaining it, if it weren't something big. This is huge. Put another way, if we don't do it, we'll get creamed by any team that does. Believe me.

1. The ch's absolutely pay off, both in epps and gpt, and we can easily whip about 8 of them in the next 25 turns. We need them for Corporations anyway, so there's every reason to NOT delay (after the unis, I mean). We don't need to discuss this, but rather just do it.

2. We will steal every tech: Philo-Communism-Steam Power-Nationalism-Constitution-MilSci*. And then some, such as Guilds, Banking, Feudalism, if we don't get them in trade. This will also be no problem. They'll get progressively cheaper as we build more and more ch's in preparation for corporations, and as we gradually conquer the world. In fact the cost will eventually be about 7 times cheaper than beakers. We don't need to discuss this, but rather just do it.

4. In all likelihood, this also tells us our order of conquest: dG-Mao-Churchill-Roosy-Gandhi-Stalin

We may have to build a second city for stealing purposes, though. I haven't thought about that yet.

It's all very simple, workable, and fun.

* @bcool: Remember those CRII-promoted Riflemen you were talking about, think Grenadiers... .

 

[mdy]

Courthouse in Moscow? Can we just build an extra Mace? (This after all the EP talk )

We could easily build a mace instead of a courthouse (we could actually fit 2 in) if we were willing to slow down growth a little. I just put a courthouse in there because we temporairily had some free production and we seemed to need a lot of espionage for our future plans.

Moai in Cuba? It's fine, but I thought we should at least consider some alternatives. Whipping the WB in Cuba asa pop3 (to not waste turns working seafood) might be ok, too, as long as it helps Bahamas with the Uni build timeline. I think I'd send the WB there, instead of Bermuda, as we have sheep + pig over there.

With only 1 land tile it seems highly unlikely we will find a site which will produce more hammers. True it would take a while to get built there even with the help of the whip, but in one of our existing sites will produce nearly as many hammers, and I can't see us building it in a new city before we can build it in Cuba. On second thoughts sending the workboat to Bahamas will improve the efficiency of the uni build there so I will amend the plan accordingly.

Not only is the difference overwhelming, but 2 maces are 40h more expensive than 2 cats, which is precisely a Confucian missionary. To make a long story short, let's just research Construction. We capture cities faster, gain gold faster, rebuild cities faster, lose fewer units, and build missionaries with the difference, thus getting our unis faster, thus making up part or all of the 500. Meanwhile, it seems that were going to finish Education too soon anyway.

It looks like Gandhi could research construction for us as well. Given the uni building timetable I can't see how spreading religion will speed up Oxford, and we would waste hammers on spreading our religion when it would probably spread to most of our cities naturally if we waited a while.

I am currently predicting that Paris will fall on T+12, hopefully another AI should have construction by then (in my experience they almost always research construction before feudalism). If not we could always research construction after education if we really need to. I don't think losing an extra mace taking Paris is worth delaying education by up to 6 turns.

 

[bcool]

I think we should look at researching construction and building the Heroic Epic in Fish or Moscow. Surely a Heroic Epic would be better than a courthouse in Moscow.

Although missionaries in Moscow instead of a courthouse or barracks is also tempting.

I think it would be nice to have an updated test game to test out how a Heroic Epic build (either in Fish or Moscow) and researching Construction compares to whipping out a bunch of Macemen in Fish

 

[LowtherCastle]

Updated test save.

The beakers are way off. The settler in Siberia has 1h too many. OTherwise, I think it's pretty much on. OF course, the AIs are all wacky...

EDIT: Changed the test save, because I forgot to add the GG.

 

[bcool]

Thanks for the test game LC!

Did a quick test

Self Researched construction in 3 turns (could have done it in 2 turns but real game only can do it in 3 turns)

Build missionary in Fish (interrupting the macemen build), switched Moscow to build HS in 2 turns and grow in 2 turns by working plains farm, whipped forge in Pigs (so it can run a engineer next turn and still beat Moscow even if Moscow builds the HS in 2 turns)

Moscow finishes HS in 2 turns, then on 3rd turn of researching construction, builds a barracks wth just enough hammers (some MM with tiles that I don't remember exactly, Fish and Moscow flipping gold back and forth I think), starts catapult (1 turn build)

Fish finishes missionary in 2 turns, then finishes macemen, then starts catapult (2 turn build)

loose warrior, macemen GG hang out on Grass hill 1NE of Galley City... Macemen and 2 catapults join them, attach GG to warrior, giving 4 xp to all units there

(not sure this was the best call--could have attached GG to catapult with a macemen and the other catapult there, then we get 10 xp macemen 10xp catapult with GG, 9xp catapult) The GG catapult could be extra movement with accuracy bombardment promotion (free upgrade to cannon later on too), the macemen with 10xp could be city raider x3) Unlocks the Heroic Epic...)

Anyways either way, gets the attack force to Paris's doorstep and sets it up to fall by T143-T144

T+9 Paris falls (and gives 120+ gold) to continue our research, and gives us Paris and every future city 3+ turns faster.

I think it is worth researching construction ourselves.

 

[babybluepants]

Listen, I wouldn't put all this effort into explaining it, if it weren't something big. This is huge. Put another way, if we don't do it, we'll get creamed by any team that does. Believe me.

Excellent points. All right, let's not argue.

 

[mdy]

whipped forge in Pigs (so it can run a engineer next turn and still beat Moscow even if Moscow builds the HS in 2 turns)

Don't we want a 100% GS next for the education bulb?

(not sure this was the best call--could have attached GG to catapult with a macemen and the other catapult there, then we get 10 xp macemen 10xp catapult with GG, 9xp catapult) The GG catapult could be extra movement with accuracy bombardment promotion (free upgrade to cannon later on too), the macemen with 10xp could be city raider x3) Unlocks the Heroic Epic...)

LC's analysis showed that if we had catapults then we are almost certain to win in Paris just with CR2, so we wouldn't gain too much from CR3 immediately, and if we settled the GG we could build a lot more level 2 units. We should be able to get a level 4 unit out of the French war for the Heroic Epic without spending a GG on it.

The real problem I have with researching construction next is that it delays Oxford/unis by several turns when we can probably trade for it in the not too distant future. I don't think capturing a few French cities 3 turns earlier is worth putting our global research back by around 8T.

 

[LowtherCastle]

Further research on Tech Stealing

One issue that has been nagging at me was this, in bcool's test, Roosy was able to help with a number of techs:

Spoiler :

Iron Working
Polytheism
Horseback riding
Monotheism
Paper
Monarchy
Banking
Compass
Drama
Optics
Astronomy
Printing Press
Scientific Method
Physics
(last two might have been mistakes since now I'm fighting the advanced Rosey)
And yes I was beaten to Physics by Rosey, I did the railroad beeline.

The problem is, we can't easily get Roosy to Friendly and we want his monopoly techs, so we need to steal them, but he's too far away, so they're REALLY EXPENSIVE to steal. Now, he's got too many cities, so we can't gift him Moscow-2S1SW, but we don't want to bash him down to 3 cities, because that nerfs his ability to research useful techs. Luckily, Silu gave us the answer in SG11:

Spoiler :

An AI is willing to accept a gift city from a human if ANY of the following apply:

...

If none of those apply, AIs only accept a gift city if ALL the following applies
i) They are not in financial trouble
ii) They have a city within a plot distance 9 (a diagonal step is 1.5)

I tested this. On the very same turn, we can gift Roosy a chain of cities, one by one, from his nearest city to some city close to us that we don't care about and bingo, we can steal techs cheaply form him, all we want.

Below are two test examples. Open the first one up, gift Roosy Novgorod, then gift him St. Petersburg, then steal 4680 Constitution for 514. In the second test, I tested two things: 1) Gifting Stalin an iceball city (Rostov) with Buddhism, so he'd adopt that religion, which he did. Then I gifted him Yaroslavl' to steal 4680 Mil Sci for 421 ; and the other thing I tested was 2) to see if he'd teleport his nukes. He didn't. My plan, though was to monitor Rostov each turn, and if he transported the nuke, capture the city with my chariot, before he could launch it...

And while you're at it, keep in mind that a sci gives us 6, while a spyspec gives us 4 + 4 (+ 3.5gpt and increasing for the ch in that city...)

 

[LowtherCastle]

Provisional Pre-Play Plan for the next 15 turns:

Rheims:
Start granary once out of revolt.

Rheims comes equipped with a granary.

Speaking of which, I updated test save, because I forgot to add the granary to Rheims. bcool or anyone, if you've already added lots of tiles to my other T135 test save, just add the granary in Rheims.

 

[bcool]

LC's analysis showed that if we had catapults then we are almost certain to win in Paris just with CR2, so we wouldn't gain too much from CR3 immediately, and if we settled the GG we could build a lot more level 2 units. We should be able to get a level 4 unit out of the French war for the Heroic Epic without spending a GG on it.

The real problem I have with researching construction next is that it delays Oxford/unis by several turns when we can probably trade for it in the not too distant future. I don't think capturing a few French cities 3 turns earlier is worth putting our global research back by around 8T.

I don't see how researching construction delays Oxford/Univ by 8 turns. We can't whip the Universities we need in a timely manner, so researching construction would not significantly delay Oxford. (there are very good odds we will get a decent gold trade to power through the rest of education after the bulb)

The gold from capturing Paris would power our research as would the gold from the capture of his other cities as well. Keeping us afloat research wise longer into the path of getting Communism. It would give us more cities to whip courthouses in faster, giving us the we need to steal more techs faster.

I thought attaching the GG is the most time effective way to get Paris captured. (And maybe getting the Heroic Epic earlier would mean less hammers invested in the units which means faster Universities ?)
Maybe settling the GG is worth the delay of 1? 2? turns of getting 2 5 xp catapults to his doorstep.
Edit: I'll take a look at this maybe we can do it better long term with settling the GG, maybe whipping the catapults in Fish. The settled GG does produce 3 too...

Delaying the conquest of DeG has risks as well. It is more hammer efficient to hit him when he doesn't have War Elephants or (unlikely) crossbows, or possibly longbows? Would Mao gift him feudalism?. So a turn or 2 savings in the start of the war could mean a significantly better survival rate or the difference between a war that bogs down and a war that goes smoothly.

Don't we want a 100% GS next for the education bulb?

I was willing to accept a 3% risk of getting a great engineer. I can see how this might not be the wisest course of action. Getting the HS 1 turn earlier was nice not only so Moscow could build a barracks but also getting the benefits of the HS for all our workers. Is that worth a 3% risk of a GE instead of a GS? Maybe.

edit: I also thought there might be some risk we lose the HS race too.

 

[LowtherCastle]

edit: I also thought there might be some risk we lose the HS race too.

No chance. We now see the sabotage production of every city on the globe. No one has close to the production of a 500h (350h? for AIs) build. In retrospect, if we had beelined Paper before Machinery, we could have known what was happening with the Pyramids and the Colossus too. Alas...

 

[bcool]

Settling the GG in Fish only delays Paris conquest by 1/2 to 1 turn. Settling the GG in Fish is a better option in combination with researching construction than my first suggestion of attaching it to a unit. (gives up on unlocking Heroic Epic immediately, but every city wants to build Universities anyways)

(Whip the first catapult in Fish for 1 pop, gets 1 catapult and 2 macemen there and the next catapult 1 turn later)

 

[mdy]

I don't see how researching construction delays Oxford/Univ by 8 turns.

Oxford is not delayed that long. This is the number of beakers lost due to haveing to research construction, and delaying Oxford/unis by a few turns/by our current research rate. In the short term we can use captured/traded gold to power through to education, however we can't keep that up indeffineately .The cost is that the next tech we researched after education would be delayed.

 

[bcool]

If we knew how soon and how likely a construction trade would come than I would more open to waiting to trade for it. Right now, I think the risk of delaying the conquest of DeG is more significant than the beaker savings of trading for it.

We have the pyramids and we are running representation. I think we can and should push to capture and expand our civilization even if that means in the short run we delay the next tech after education.

The costs of more cities will be mitigated soon enough by whipping courthouses and the ability to run specialists that will keep our research going through the lean times.