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

[Bacchus]

Prioritising islands is an interesting task -- as with land cities, you want to both claim territory, but not over-extend, work good tiles, but also build a solid strategic position, except the decision space is much larger and the trade-offs starker. Overall the naval game seems much more interesting and less explored than the land one. And here Mardoc added the choice of which wonder-doubler you'd rather have.

I don't like the dry rice settlement at all, it seems to achieve very little on any front, even "tiles worked", and it's not at all urgent. Even the canal filler is a better move, as it can take more food from Lifeblood and pay back for itself faster, brings in the horses tile to be worked, with easier worker access, doesn't require the use of the galley to settle, and can be zone-defended without commitment of extra garrison. In fact, because of the abundance of food and the fact that this city can double-whip a settler pretty quickly, I think we should throw it down as soon as our galley is at capacity and can't transfer settlers anywhere more meaningful.

I would suspend the decision of where to settle right up until we have to make it, just to avoid tunnelling into opportunities because we happen to reveal them earlier. Also, I'm sure there will be other settles happening in the meantime, which would affect everything.

When are we going 100% science? Is it next turn to complete it in 2 turns with full-on Research building?

[Coeurva]

You're right, canal filler looks stronger as long as Lifeblood doesn't require the food. I'm hoping for that seaboard to hold food yet unseen, though; it seems to do so for everyone else.

T92 will be the first of 100% science, to benefit from the Academy that we'll install right then (1t of travel time for the GSci). We're still taking ~2.5t to pay for 1t of 100% sci; civics and cities cost us a lot (but after Currency, every city at least pays for itself with trade routes + city centre on 0% sci turns).

[Bacchus]

T92 will be the first of 100% science, to benefit from the Academy that we'll install right then

Ooff, I can see that GNP spike. And then it will continue spiking because Currency. And then we can build wealth to stay at 100%! For some time at least. You reckon we'll be able to knock out CoL in one sitting? Maybe even Mysticism-Meditation-Priesthood-CoL? The 20% extra from Priesthood on a 340 base beaker tech saves a whole bunch of beakers.

[Coeurva]

We'll have to decide if this should be X10 or X11. I'm still in favour of X10 as the island west of Styx, as we'll need a border pop at either city and the island is better set-up to do it, but this city would not need a garrison, which means more "we're prepared" signalling at Savant in the north. Island X10 also has workers in the area that will be easy (and desirable) to ferry over after chopping/improving the western horse. In any case, our food problem might be solvable after all. (We're ranked #4 at 130 CY, best is 140)

Calendar becomes ever more desirable. I'm in favour of skipping CoL at this point, all that we'd really get from it is Confucianism. If we want to boost happiness, an earlier IW is actually more efficient, and the Holy City will almost assuredly be Sicil or Lifeblood. Neither needs the culture.

Joey has built the Great Wall (for the GSpy points, I assume), and has a bunch of barracks throughout his cities (he's PRO/CHM).

Savant is still adding soldiers, and has a new-built galley in Harmondale. If there's another one 8 of Fountain Head (my guess is he wouldn't have built the other galley if there was), he can sink one of our work boats this turn, but I dare him to do it -- if he's planning to go to war against us, he'll be forced to show his hand early this way and raise our alarm at once, which might induce him not to because a city would be a much larger prize. I'm concerned that Savant will look at GNP graphs (we've peaked for a new maximum there) and conclude we'll "run away", more so if he's not mindful of how RtR makes the KTB so much more meaningful and beelines less practical.

Yuri's soldier graph remains flat. He's settled another city, though, and he's building granaries.

The Khatunate does have one library inside Oghul Qaimish, on the border with Joey.

Three cities will grow eot.

[Coeurva]

It's probably more accurate to think Currency + wealth builds will put us at 50% breakeven, because we'll have to build some more military now (that's the catch with GNP spiking) and there's another issue -- Rhapsody cannot really avoid growth until we build a settler for the hill 44 of clams (a city that will just run clams + cottages forever; pure commerce), probably X12 (X13 western wheat island?). The Academy adds ~20bpt, Currency as well (if X10 is island, otherwise 10bpt -- maybe adjusted for Khatunate trade routes)

Confucianism is nice, but a) it paints an even huger target onto us, b) we don't have that much time to tech Calendar, IW and HBR, all of which are extremely important, c) the benefits from religion are all locked behind even more tech and at least 1t of anarchy. The other unlocks at CoL are ignorable for now: we'll benefit from Caste much more if we have 30 more CY from banana plantations and can afford not to run Slavery for emergency whips (because, say, we have NumCav running around at the borders).

EDIT: I'm thinking Maths next, in fact. We still have cities to get up to speed and more forests than I had expected, and we can probably complete it within 4t at breakeven.

[Bacchus]

I assume getting Maths before Currency for the 20% bonus doesn't work out as actually beneficial when accounting for commerce lost from trade routes?

I'm not against Calendar, that's a lot of bananas around the place. I would still get mysticism though, reliance on libraries for border pops is getting impractical for the islands.

EDIT: I still think that our settles should be asking for more at the moment. So X10 West of Styx, and X11 around the gold hopefully. Get stuff that's contested first, backfill later.

MOAR EDIT: Don't we mostly need IW to make Calendar useful? IW also has other uses, plus deterrence value, so I would put it ahead. And with IW we have (a bit) less pressure to get HBR.

[Coeurva]

That quadruple food resource site looks just so appealing to develop, though. You're sure you don't want that one before the gold island? Actually, I think we can double-settle both sites as X11/12 after Currency. I'm more concerned that the gold might be contested by GermanJoey, though, we're still stretched thin in the area and the supply lines from Whitehall are very long, 4-move galley notwithstanding.

Maths -> Myst is probably better than the other way round, to nail those monuments with a single chop. But Myst is probably necessary at this point, most cities won't have two forests to chop for the border pop.

Maths before Currency probably doesn't pay back, on a guesstimate -- while we could assume ~30-40bpt during the 2t of 100% Currency research from the 20% bonus, we'd also be losing 20bpt whatever wealth builds will effectively bring us, for every turn that Currency is delayed by. More pertinently, as you've suggested, we need Currency quite direly to improve our still-shaky diplomatic position. Given our "harrassment" of Savant previously and the fact that we're still his avenue of conquest (not necessarily expansion... yet... thanks to the ample islands), I'm growing ever more concerned that he'll see us as a threat or snack.

Currency is also cool for demographic information, if indirectly. I want to know how much gpt Savant, JR4 and Yuri can produce.

Calendar goes hand-in-hand with Iron Working, increasing our workforce a bit more (mostly Rhapsody's task, aside from generally fueling our tech) and settling the National Epic spot. The interesting part, aside from managing the workers themselves, is how we can slot HBR into this.

EDIT: Of course, IW first for the gems. Well, and the iron, if applicable. IW also adds a metric ton of soldiers, which is its main deterrent value, as swords don't really defend that well for their cost.

late edit: Nonsense about the "metric ton", it's 10,000 soldiers (which would only weigh a ton if you assume that each soldier weighs 100 grams). Still a lot for a tech (max is 12,000), but I had been thinking ~16,000.

[Coeurva]

(This post grew terribly long, somehow. It's not "required reading", feel free to skip it.)

Taking a step backward from staring closely into the abyss. This isn't really a strategic overview; it's mostly assorted (unsorted) thoughts about how to play Civ.

At this point, I'm mainly concerned about two things in this game: our lagging crop yield (which I'll set aside for this post) and Dark Savant. I don't mean 'our military', because I think we can keep army parity in a direct race against him, as long as Yuri doesn't pile on. They both have reasons to attack us, and we cannot deter them solely with our army. Crippled as Yuri may be, we cannot outproduce both him and Savant in units, much less defend at two points which are 8t of melee unit movement apart; our empire is stretched more vertically than horizontally, especially after the Russian war -- which I still think should not have gone nearly as well for us (in fact, I'd consider it a mistake in hindsight, because we simply couldn't have predicted that our enemy wouldn't know the purpose of spearmen).

----

On the marriage of true minds?

When Mardoc mentioned earlier that we (or I) might be underestimating our performance against Krill's, or the impact of our strike against Russia, I wasn't concerned so much about a hypothetical Krill take-over of Russia purely due to Krill's "mechanical" skills -- of course, these are strong as well; but even Krill would be materially limited at what he could conjure up from five size4 cities without granaries and T90 Writing.

My fear was rather that Krill, like almost nobody else on RB, understands and considers his opponents' play from Turn Zero onward; indeed, he seems to delight in considering their options. (mackoti also qualifies, though -- the tipping point of awe for me was when he somehow anticipated suttree's early galley attack in PB13 or whichever one that was -- with Yuri on the team, btw),

Krill (if focused) isn't really only managing his own empire; he's thinking even more about his opponents' empires. Since we've already had Lord of the Rings analogies in this thread, let me add another one: Krill, from how I see him, is Denethor -- immensely capable at government, a "positional player", defensive-minded, preferring tall cities, and far-seeing to the point of obsession -- utterly respectful of others' strength when he perceives it as such, which is part of what makes him special. The only way to defeat him, barring the usual luck with neighbours or whatever random chance is inherent to the game set-up (which is not very interesting in itself, as it's not under your own influence), is to turn this against him, to project greater strength than you have, in a select few areas that he overestimates (like Sauron did), and then deal a blow that confirms this appearance to him (this part is important). But you really have to show such strength that he doesn't consider an attack before your plans are ready for execution; otherwise, he will risk his game on rendering your threat ineffective (therein lies the determined frenzy of his break-outs, and his reputation for "aggression").

In fact, you're more familiar with how Krill reacts to a strong threat, having been involved when you and Gavagai succeeded against him in a tour-de-force in PB27.

Maybe it's just the Angora cat in his avatar that suggests all this to me.

Also, I'd actually think that the closest player to Krill in this roster -- from how they both play Civ -- is German Joey, which makes it all the more interesting that they're going to be neighbours; looking forward to that. But enough of that for right now.

Coming back to Yuri/Savant -- they might not be Krill, but nonetheless, if they truly understand our current position, and try to get into our minds -- and they have more knowledge on the relevant parts of our empire than we have on theirs -- they can find ways to coordinate successfully. Even if Yuri languishes in effective resignation, my main concern is that Savant/ipecac might be much better at reading our play than I am at reading theirs. At the moment, I certainly think so. Bluntly, I have no idea what either might be doing besides the vaguest "they want to expand and improve their GNP" (less so for Yuri, but much more so for Savant), while I think they have every idea of what we've been doing, and are planning to do, outside of minutiae such as whether our next city will have a 510 or 502 first-ring food tile. That our pivotal techs mostly telegraph themselves (Curr: wealth trades; Calendar: banana/incense hookup; IW: iron/gems hookup; CoL: Confu; Music: GArt birth) doesn't help us, except in the case of HBR where it's half of the point, but that's also the tech that doesn't improve our economy ("only" helps to preserve it).

For instance, Savant has settled for stone and connected it, a resource which (extrapolating from our own stone/marble sites) seems to be found in rather marginal places (and only on desert hills, where it's worse than the whip), compared to what else he could settle on those islands. (It also means they have researched Masonry, it wasn't just for the 2h nougat centre). But what wonder could they be pursuing, if any -- the Pyramids? Possibly, with their low GNP, and high food/MFG -- but where are their libraries, all in the north? Or, perhaps on a smaller scale, what purpose is there to the galley inside Harmondale? Where is Savant's expected barracks whip inside Fountain Head? That city has now hit size3, but still only contains a stele for culture; why a stele, if AGG barracks would have cost the same?

I want to know much more about their plans, or at least speculate -- but I don't know how we could approach this, other than by meeting Gavagai, and I don't have enough experience with Civ to speculate about such planning; I can hardly do it for our own civ. Suggestions about this, and their possible plans, will always remain very welcome. (You've been making those already.)

----

On concerns of war

The most effective deterrent against a two-pronged Yuri-Savant strike that we can exploit, apart from their other neighbours' pressure, might be their mutual concern of getting bogged down in a concentrated defense on a single front, while the other player gains cities without much resistance. We can exploit that -- we don't need to make conquest impossible from what they can see, just painful.

The terrain at Sorpigal makes it impossible to watch the city centre as we're doing with Harmondale; what's vexing me is that I'm aware how Dark Savant will move units over there, but he might be thinking that he's preparing a surprise for us. Thing is, it doesn't matter if his perceived dagger will instead prove a sledgehammer. At some point, he'll realize this himself, if he hasn't already -- we cannot stop a dozen Ancient units at Ignis quite yet, if he wants to commit (I don't doubt he could assemble such a stack; the only check on him is Gavagai and perhaps his GNP, which warfare wouldn't help immediately) -- after which our position would crumble horribly, especially if he poured 2-movers unto the breach. Even more so if he's about to finish Horseback Riding. Unlike cottages, Alpha/Currency as an economic advantage cannot be taken from us, but I have little doubt that he'd like to have silver + whales provided to him by helpful petit Ignis.

Perhaps bizarrely, Dark Savant retreating his Sentry chariot from its observational post at Harmondale, which could be taken as a benevolent gesture, worries me more, as he appears to think that it's better-used elsewhere... I'd consider it more openly friendly to stay visibly alert of us: it would send the message that he doesn't have a large military to meet us on this front; much less to take our cities. The new galley is another source of concern; I'm hoping that it's meant to ferry a settler, but be assured that Whitehall and Styx will keep building archers from here (Styx needs granary/library as well, though).

Savant has also never sent us a single trade offer throughout the game so far. You'd think he'd try for Open Borders if he's interested in closer relations, although perhaps he's afraid of getting scouted (he knows a bit more about our cities than we do about his) or exploited by Currency -- or just of fueling our GNP further. This reinforces my belief that his diplomatic strategy can be condensed as "befriend the Khatunate, deter Gavagai, and devour Carthage". We might have won a trick by being the first team to Open Borders with JR4/Ref, as trade routes increase in value over time, and they would not gain much from OB with Savant when our cities provide the same or better income; so they'd have less reason to sign them with him. Better yet if we're not competing against them too much while expanding, which I hope will be the case for the entire game -- until the final stages, at least.

Speaking of which -- should we sign open borders with him? I think we can work such a deal into the gold gift, actually -- a lopsided OB trade will not be taken for a sign of weakness as easily as a gold gift might, I think, but it will signal readiness to cooperate (i.e. agree to focus our expansion elsewhere) and concede lopsided treaties. He knows about our border cities already, including the exact composition of our garrison at X9; he'll probably visit Westward as well.

When I moved those chariots north while in view of his work boat, it was less for their actual defensive value (and the barb axe wasn't even visible) and much more to signal that we're mindful of Savant gathering knowledge on us, and to keep him uncertain that it will stay accurate for long.

----

On cooperation in games of conflict

The Khatunate will prove the fulcrum of our own geopolitical future. I hope they realize that, when we plant X10 and the gold city, we will also strengthen our position against German Joey and Dark Savant, whom they should probably count among their enemies (not quite sure about Savant, but Joey is an incredible front-runner right now), and thus allow us to serve as better allies to them; I'm certainly willing to owe them a few favours for this concession, because our enemies are the same (except for Yuri, whom they'll probably support because he can both distract dtay and check our own expansion, although maybe not if they need us to grow larger).

Speaking more broadly, what I've been finding is that "AI diplo" doesn't impede or water down diplomacy at all -- on the contrary, it just eliminates the wishy-washy deceitful rhetoric that some players might think constitutes "diplomacy". Instead, because the players are left pondering the situation according to their own perceptions -- which will always be skewed in favour of their own interests, such that getting an objectively fair share of the world will not suffice -- it becomes much more meaningful to actually realize and respect the other players' strategies, and genuine opportunities for mutual benefit that arise from them, and to signal this with in-game actions -- you can't stall with words here.

More concretely, if a plum pudding would be divided equally between 10 players and I can work with someone to ensure he gets 20% of the pudding instead, while I end up with 15%, I would still be increasing my chances to bite the king-making coin hidden in the dough: so in my own best interests, I should always take the opportunity, even if I increase his chances by more than my own. Of course, in-game situations are usually not that simple. For one, the advantage that your opponent gets must not be so large that it allows him to take all of your share afterwards.

At the same time, you cannot allow your direct neighbours to get lazy about their border with you, because that will give them leeway to concentrate their expansion on another front. But if you overdo this, they might pick you as a target instead and lock you into mutually-assured destruction. The threshold varies depending on both the game state and the players involved.

It's fascinating, and I want more of it rather than less (unlike plum pudding).

----

Final words: I mostly do the focused turn reports for Bacchus' sake, because the hearth of down-to-earth information is needed so that strategic ideas can be incubated, but this airy tangent-drawing is more in line with how I actually play the turns, rather than those post-hoc wrap-ups.

Which, you know, probably explains at least part of why my micro tends to suck.

The race for more food will feature in the usual turn reports, since those are naturally more oriented at turn-by-turn organization and development. Coming back to you in three, two, one...

I'd always like to read your own thoughts on these topics, they interest me more than whether we can complete the next pair of settlers in 6t or 7t (that is to say: to a fault).

I'll play in the evening today, probably ~20.00 of local time.

[Bacchus]

Firstly, I've never really bothered with a deep analysis of play styles, I think there is a great danger in leading yourself astray once you start trying to reason explicitly what other players are like, as information is woefully incomplete, and whether you introduce analogies, or data (percentage of games Mackoti built the Pyramids), or any other method of rational analysis, your analytical method is very likely to get in the way of your actual knowledge. I definitely have some ineffable sense, neural-network style, of what, say Gavagai, is likely and not likely to do in a given position, but to hone that sense I prefer to gain experience with Gavagai's actual play and feed my perception with observation, rather than build an explicit model of Gavagai in my mind. System II thinking just doesn't perform well enough in this case, or, more to the point, no-one has the thousands of hours of brainpower to invest into developing a robust, System II-amenable model of playstyles.

I doubt Savant has a good idea of what we are planning, I think our mutual knowledge is about the same, albeit I'm surprised that you feel so much at a loss regarding Savant's actions. As we have, actually even earlier, Savant has discovered the abundance of islands. His priority, just as ours, and indeed anyone else's, at the moment is to grab as much of the islands as possible, without falling to an attack. This is a delicate balance, potentially one civ might decide that they rather act as takers, than builders, but given that OT4E isn't playing, I don't think anyone will. I mean, why do it? You have no idea of the schedule of other powers planting cities, they will probably be inconvenient to defend once you take them, etc. So most people are going to be pushing for settling, Savant included. Things are trickier around the borders of two collapsing civilizations, but that doesn't really concern Savant directly, if anything it pushes him more towards establishing a naval advantage, as he has no China or Russia to use for expansion. He does have circumnav, which he can actually use.

As a result, more galleys, including in Harmondale, is really not surpising. Whatever happens, fleet is going to be important, and even just for regular settlement, one galley is just not enough. Why wouldn't he build more? No barracks in Fountain Head -- well, the stele is better long-term from a cultural perspective, and you would build first it exactly if you planned to NEVER build barracks in this city, which is a perfectly legitimate decision.

You are also greatly overthinking stuff, as with CML's naming scheme. He settled on stone -- so what? It may just have been the only way to get double-food in the city, there is plenty of awkwardly positioned fish around the place, is there not? The piece of information that you happen to have is not necessarily the important one, it's just the one you got in a lottery. Not to mention that stone is just very useful for Moai, and Moai is very useful for this map.

Going to a higher level, why on earth do you think that devouring us is a preferred option for Savant? Is there any specific fact about our border, our land, our play or anything else that makes this a necessary, or an attractive move compared to others? Yeah, sure, you need to eat someone to win, and he is no doubt open to devouring us, just as we are open to devouring him, but why would he actually tunnel in on this conquest at this point the game? Where is the opportunity? To me, that would be not just an arbitrary thing to do, but a pretty stupid one at this stage. Now, does he have a strategy around trying to put us in a position, where attacking us would make sense? That is surely the case, just as it is for all neighbours in the came. You need to make a lunch of your neighbours, but normally you need to find creative ways about how to do that.

What are Savant's options here? Dogpile with Yuri? Well, Yuri is already pretty much out of commission. If you are afraid of this dogpile, what should Savant be afraid of? Of course, a dogpile of us and Gavagai, and that's a much scarier prospect. Neither of them will actually come to fruition though, they are more relevant as threats, which take some countering, but are eminently counterable on their own. No, the play here will be much more interesting, and it will involve cross-sea conflict. You eat your neighbour not at your own convenience, but when he is already facing pressure on the sea. I have no idea whether Savant thought this through, but from our side, we want to encourage him to settle and thereby make as much of his stuff vulnerable as possible, create possibilities for conflict with others.

I don't think Savant has much MP experience, so I will leave him at that, I really don't think we should be focussing on him. Everything I said applies to us -- Savant will only become a problem if we get other problems, and that's what we need to deal with. The land differential that you make such a big deal out of is not a problem, our land is not actually dry, we are maybe at a small disadvantage, but that's not what will drive this game. Our actual problem, as is everyone else's, is how to not make enemies whilst claiming enough islands not to fall behind. This problem actually has no solution, you can't avoid making enemies entirely, and we are very lucky that we have a nice predefined enemy -- GermanJoey. Our other sea neighbours are either out of commission (China, Russia), or have already been pushed strongly in a different direction of expansion (Ethiopia). Stygian Terrace was a really, really huge move. We just have two powers to worry about -- Mongolia and Greece. We befriend one, and fight the other. The purpose of our fighting is two fold -- 1) gain a bigger share of islands than normal at Greek's expense (actually, also at Russia's and China's, which is what we are competing over with Greece); 2) destroy his ability to be a meaningful problem to us. If we achieve this, Savant will never have an opening to attacks us, we will develop an economic advantage, devour Yuri, and then fight it out in the major league with Mongolia and Inca.

[Bacchus]

Oh yeah, what Savant needs to do to compete with us, just at the moment, is get happy. That supposed food shortage has driven even charismatic us into happy cap, think how much more pertinent that problem is for him.