[SPOILERS] swance bitten, twice shy

[ljubljana]

billion-yard overview of GT



so um.... due to our obnoxious forward-settling.... the next prize after lalibela is THEIR CAPITAL. good for us, as its a production monster and may outstrip Abi in heroic epic candidacy, if it's not held back by too many cottages and too much water. it's not much closer to us than it is to nauf though, so we will have to hurry, and bring along a shitton of 2-movers to race for it with

that said, it's WAY more expensive but i may prefer diverting to feudalism over HBR. i am not sure i'm convinced that there are THAT many situations in which (1-promotion, 50-hammer) horsies or (1-promotion, 60-hammer) elephants will be more hammer-efficient than an equivalent number of (2-promotion, 35-hammer) impis and (1-promotion, 30 hammer) chariots, and clearly we don't have time for stables (would it be THAT overpowered if AGG gave these at half-price too?). maaaaybe attacking into mixed axe/spear stacks from 2 tiles away is that situation, but as far as i can tell 5 chariots seem about as (in)efficient into spears as 3 horse archers...

but longbows though do a lot for us, i think. you have no doubt noticed that i'm a believer in the offensive value of defense (almost certainly too much so ) - namely, attacking GT is only a good idea if we do NOT open a retaliatory attack on our flanks, so longbows in ginger and greenline's face would effectively serve our offensive needs by freeing our offensive units to actually, um, offend. moreover, aksum and gondar are both on hills and represent the border cities in our preferred split with nauf. but i think one of the biggest dangers in this kind of thing is that if we outrace nauf to them, nauf's stack will keep on rolling and roll right over us before the cities come our of resistance, whip walls, etc. well.....not if there are PRO longbows on hills there they won't! first feudalism and then engineering seeeem to open up a pretty defensive era in the game in terms of warfare and it'd be nice to open that up AS we are trying to establish a contested border in former ethiopia and not after

i would like lalibela to be our first target to cut off adulis, but tactically it looks like a real nightmare



as much as i love 2-movers and think we will need them to race for aksum and gondar, they are no faster attacking lalibela than 1-movers are and if we face whipped walls and a pile of axe/spears, we probably need cats and our own axes to stand a chance of getting through. moreover, the approach to lalibela from the north forks adulis almost by accident, so it might make sense to go there first just to take away one of their better sources of emergency whips... then again getting bogged down at lalibela is looking like a serious risk so perhaps ANY delay in attacking there is unconscionable

also an outside chance we could send troops at matara



but that city is weak, hard to defend, and isn't a 2-for-1 like lalibela would be so i don't think so. not in the initial wave anyways...

re troop-producing capacity, i'll have to think about it, and also think about what mix we want of cats, axes, impi, and chariots... but here are some overview shots with commentary about what i expect can be contributed in 10t



atamifuji has just a+3 food surplus and no way to increase it, so it is probably exempt from whipping unless we REALLY think it's necessary. perhaps it can be squeezed to add one unit.... but we need that library up before the great person pops so we can run specialists here during the GA, which is ALSO going to slow down its growth. i'll go ahead and swap to a slow-built unit for now though.

the rest can whip though. abi will likely get its pigs back before the war, tomokaze can borrow hoshoryu's wheat and 4-2 whip, wakatakakage will get its fish next turn and give hoshoryu the cow back, hokutofuji has just +3 food but is food-capped until we have a lighthouse or civil service. so maybe hokutofuji double-whips once and the others twice each, plus some chops at tomokaze and wakatakakage, for ballpark 10-15 troops of some description? plus natural production for another 5ish?



capital i would prefer not to whip off the size 10 food box but it will continually make troops the slow way. mita likewise after it finishes the courthouse. kirishima REALLY can't whip if we want to remain economically relevant but it can slow-build one impi. takayasu is even worse with unworked river cottages and no hammers of any description lol. but i think WMH and kotozakura can double-whip once or twice, and koto has good production as well. WMH i think doesn't want to slow down moai by dumping ALL its hammers into troops but it has been whipping them and overflowing into moai so it will probably continue to do so.



ichiyamamoto at +2 food surplus can't whip at all, but it has extra chops so it can contribute SOMEthing. daeisho i was stupid and overflowed into the settler ONE turn before that became a silly idea, so i might be committed to finishing it....... or we cut our losses and swap if we want 60 more hammers of units and to not create a point of vulnerability right now vs ginger. kinbozan doesn't even have an ikhanda yet and has bad growth problems so it probably can't contribute. tobizaru is mr 3rd-best-commerce-city-with-no-library-or-market so maybe we'd better whip those things instead of troops lol. but ura is pretty strong and should be ok adding an impi or two, though i also think that at least its resource tiles are too strong to whip off of

so that's all very ballpark but i think we can have around 30 troops in theater 10 turns from now without making unrecoverable economic sacrifices. should be able to get our choice of military tech from HBR/feudalism/machinery in time for it to matter as well. so seems not crazy to send nauf a 10-turn commitment. if anything i'm a little worried it won't be fast enough to get all the territory we want:



surely the spiking milpower implies there are more troops in the fog, but i don't really expect that stack to do more than bounce off nauf's cats and axes.... ideally we'd attack either after GT's stack gets annihilated or when it is wandering around in nauf's land and therefore unable to pivot.... before and they can divert troops, after and we lose the race. when do we expect THAT to happen? i think it might be in as little as 5 turns if GT actually charges in head-on.... or perhaps they won't do that but will instead see that their odds are bad and linger unproductively on the border, in which case maybe we have the flexibility to wait until nauf potentially starts taking cities.

tbh, while they obviously aren't going to wheel around and hit us now, i am still a little worried this whole war could turn out to be a false alarm, which is why i'm a little hesitant to do stuff like whip troops at tobizaru or swap off a half-finished settler at daieisho.... now that GT can be more or less certain superdeath will not come to help out, i really don't know what they can expect to accomplish in a 1v1 with nauf.... maybe this whole wardec is a bluff to get them to whip a bunch of pop during their golden age? still, there's no better time to PREPARE an attack than when the target's troops are diverted to their opposite border, so i do think we'd better make troops and send aid signals to nauf so they aren't tempted to peace out, but i also don't want to commit so much that we'd set ourselves back if this whole thing just kind of fizzles...

edit: ps, filed under the "RB civilization is one big iterated game" theory, it's not NOT helpful that i am so prone to overposting normally, so THIS flurry shouldn't be setting off alarm bells now

[ljubljana]

Turn 124 - Zululand

decided we will send nauf a show of support if we log in and nothing significant has changed. and



...how significant is THAT? pretty damn well-timed against our horde of impi, continuing to mass which now looks silly. i don't think it fundamentally changes whether this is a good time to attack, but i guess it DOES change which units we should be massing.... against axe/impi and in favor of just piling on a shitton of catapults i guess, intending to promote some of them city raider. and maybe it favors a trip to HBR instead of feudalism? well, let's see.... 10 chariots = 6 horsies, let's see how they do against an equal number of crossbows. 10 c1 chariots vs 2 CG1 xbows in lalibela (40% culture on a hill)



wow! these kinds of numbers win only MOST of the time?? ok, do the 6 horsies do any better



...yes actually. i'm surprised, but perhaps this is the effect of immunity to first strike combined with the likelihood of weak chariots taking very high damage FROM such first strikes...

or for those 300 hammers we can have 5 (nerfed, which i'll represent badly by taking away the combat 1) elephants



for a similar result. although all of these of course will run into big trouble with spears in the city, and also leave us with no credible stack defenders against ATTACKing crossbows.... which really only longbows seem to be able to provide and those of course are 1-movers....

well, quibbles about what unit type to use aside, i don't think this changes what our decision should be, which, i think, is



sent. i believe next turn or the turn after is the turn to try to renew peace w ginger, while they hopefully don't connect the dots to the GT war and while the resulting peace is long enough to actually buy us the time we need

sigh. really going back and forth on hbr vs feudalism next. without hbr we now don't have a 2-mover that can crack GT's best defender, which poses a serious problem in a race for aksum and gondar. and the best thing our more-distant cities can whip that will reach the front in time is chariots/impis which now look much worse than they did a minute ago.... but horchers are worthless as deterrence on our other flanks or for defending aksum and gondar if we actually GET that far. i'd love to hit hbr AND feudalism before the war but then our workers will have to endure quite a few turns with literally nothing worthwhile to do but road... although, does that matter if we now expect to be whipping too much to grow onto hypothetical plains farm chains anyways?? like, is the marginal utility of a plains farm if we expect to find it to not be a good enough tile to not-whip during wartime supposed to be something OTHER than zero?

balls. i have to ponder. sticking max science on HBR for now but we'll see how i feel, and if more awake me is possessed of any further insights to lend

[aetryn]

Some thoughts:

It's fine to get Longbows after the war starts. We can bring them up to defend new conquests and it should be easy enough to get them in place on other fronts. What you need to start the war is your offensive units, whatever those are going to be, and enough other stuff to protect your offensive units as they move up and heal. Longbows aren't great for mobile defense since they really need to be standing in a city or on a hill.

Also, from your screenshots, Nauf appears to have lost a city as his city count went down from 14 on T122 to 13 on T123 (GT did NOT increase in city count, so looks like a raze or auto-raze). So whatever's going on, it's not entirely bloodless. I admit I have no idea how this is going to go. GT may have burned a city through surprise and that may be it. Or they may hurl armies at each other. My impression of Nauf's "I need vengeance" meter is that it's moderately high, but obviously if he doesn't have the forces right now or the desire to build them, that vengeance may come later down the line.

[xist10]

Try vodka again, after you reduced city defense to 0% and attack with a few catapults more.
Against 1 or 2 crossbows, catapults are not that effizient, but against 4+ ?

A significant part of your mobile army should be catapults. - this is CF 1 mover doctrin

[williams482]

From the tile yields, looks like GT's iron is the grass hill 1W-1SW from his capital. Reasonably well sheltered from us, certainly not a plausible target for a lightning pillage raid.

My instinct here is to get HRB as quick as possible, double whip a few horse archers in time to join an attack stack of impis and chariots, tip your hand one turn early by roading up to and preparing to road on the hill 2W of Abi while placing your stack 1W of Abi, and then declare war and move your two mover stack 4W of Abi to fork both Liabela and Adulis.

GT will see you coming but have two total turns (and depending on turnsplits, one or two production windows) to counter, with two cities that currently have basically nothing on defense. The offensive forces we have seen from them do not inspire a great deal of confidence from me, so if they want to make any meaningful gains then fresh units like that crossbow will have to flow west towards Nauf, not east towards our border. If they are fully committed to the Nauf attack, which will be obvious to us and we pick the second half of the turn split, the worst we could be facing for our attack is whatever we see in the cities on our roading turn plus a whipped, promoted crossbow in each. By the look of it, horchers with some impis/chariots in cleanup roles can handle that.

I feel like the two-mover fork is important here for all the usual reasons (it's fast, it limits reinforcement options, etc) and also because in this case it basically guarantees that you'll get Adulis. That city is much harder to reinforce and would be very easy for you to hold if taken, even if Liabela turns out to be stuffed full of unexpected units. More likely given what we can see on defense now, you'll get the lightning strike on Liabela and cut off Adulis, allowing a slow stack including cats to take it a few turns later with fewer losses while the two movers can either hole up in Liabela and dissuade a counterattack, or (if the battles go unreasonably well or GT is already collapsing on the Nauf front) push on immediately for the capital.

Does this seem like a practical plan to those of you who have actually fought humans in this game?

How many horse archers specifically can you get out 10 turns from now?

[ljubljana]

hi friends! as mentioned in the tech thread i am Abroad right now, so my reporting proooobably won't be up to my usual standard. but i am in broad agreement with all three of you. @williams, this attacking tactic seems right to me, possibly with one exception - it looks like on the same turn we would be forking adulis and lalibela, we could simply attack adulis while also threatening lalibela. depending on the defenses in play, i may do that.... waiting for cats at both lalibela and adulis will give fewer losses but also slow down a push for the capital, which i expect to be relevant here. right now i'm vaguely envisioning sending the 1-movers at lalibela directly while TRYing to use the 2-movers to first handle adulis, then converge on lalibela once the 1-movers are in position

@xist i agree about cats. i think for 1-movers i will build exclusively cats here, with a few war elephants thrown in as stack defenders to prevent efficient attacks out with crossbows (or if i get them in time, i may use longbows for this instead). hoping to have something in the neighborhood of 10 cats, 10 horsies, and 10-15 random other units come attack time

@aetryn i do think longbows work for mobile defense here, at least, because the approach to lalibela is so hilly.... because ALL the relevant cities seem to be on hills, we can promote them guerilla during the advance and not expect to lose anything from a city defense perspective once that is their main task. i am increasingly thinking that if we are going to be doing heavy troop whipping over the next 10 turns, it doesn't make sense to push for CS..... it feels a little counterintuitive to be okay with our workers doing NOTHING (well, besides roading) when they could be chaining farms, but what is the point of making CS farms now if we'd just whip off them into units anyways.... at least feudalism/machinery give us SOMEthing to help the war effort and feudalism does speed up CS when we do decide to go for it. and HBR will be our first research for sure - i ended up putting 10% into it last turn with the intent of finishing it this turn, delaying the big 100% turn in case we pick up another KTB from someone else in the meantime

[ljubljana]

turn 125 - zululand

empire-wide celebrations as abi's borders expand to third-ring :DDDDD though it will still take a few turns to actually get the pig back ofc

nauf echoed 10g/10g, which i accepted (is there a way for them to know i did this?) but didn't send back 9/9 in case we don't want to commit to something that precise. whipped a few cats, reallocated a few chops from infra (not courthouses, dw, just delayable stuff like lighthouses in the new coastal cities) to more cats. we have 2 extant as of now and 7 in production. next turn we get HBR and start horsing around (sadly the KTB gambit failed)

new GT crossbow is indeed moving west, and though they build a new one in lalibela this turn i expect it to do the same. they are out of golden age mode now and so should be unlikely to outproduce us + nauf going forward

we should at least consider starting this war before we're ready... like, if in 2 turns adulis just has one spear on defense still, we COULD move up like, 7 impis and two workers after GT moves so they can't whip..... then unless adulis produces a unit between turns we can just walk up and take it. it would be a little slimy (and piss off GT into accelerating their defensive buildup and probably viewing us as a primary aggressor instead of a vulture) but could save us a lot of headache - once we start whipping for real i think GT will at least make ministacks to defend there and killing the city while we can do so basically for free could save us significant casualties. i doubt the window to do this will be very large as once they see our power spike from HBR and then our increasing trickle of troops i expect real defenses to start manifesting themselves

sent drey cow/cow, still waiting to send to ginger. i decided to use our EP edge from the courthouses to pour 10 points/turn into ginger in a bid for research vis. that, and because i hope to scare them a little with the combination of this + our power ticking upwards.... scare them into accepting the otherwise ill-advised cow-cow proposal i will be sending in the next few turns

[williams482]

Oh, of course. You can move the one-mover stack right up next to Adulis on the same turn the forking stack takes position between the two cities, or you can move it up to be in position to attack Liabela one turn later. Too bad there's no tile that allows for both, but that's one-movers for you.

I think we can probably wait for the last possible moment to decide if Adulis' defenses are going to merit having the slow mover stack deployed against it, or if we can skip right past and drive on Liabela instead while the two movers handle it. Notably, that last possible moment should be after GT whips whatever they have to whip in Adulis once they see our stack in an obvious staging position.

Although, our half of the turnsplit is already decided by the GT/Nauf war, right? who has which half right now?

As for workers, it looks like there are a lot of unlikely-to-be-worked-immediately plains river farms still unmade. Are you actually going to run out of things for workers to do, or just immediately yield-beneficial things for the workers to do? If the later, I definitely favor going HBR -> Feudalism ahead of Civil Service, because I think you're right about double promoted vassalage longbows being very useful all-around units in this environment.

[ljubljana]

wait is that how turnsplits work? right now GT has the second half of the nauf-GT split; if we attack GT do we get the first half of the turn as well? or is there the possibility of a 3-way split going nauf-GT-us? if the former, i think that pretty much closes out the early attack on adulis idea, that kind of thing REALLY need to not run into a whipped crossbow to work imo

re farms, well.... other than at mitakeumi and kirishima, most of them now have workers on them and are partially complete. the eastern workers might be able to do kirishima cottaging and not run comPLETELy out of stuff to do, but some of the western workers probably will. still, if we are going to whip a bunch we have probably hit the point where we generally are pop-linited rather than worker labor limited, so idk if it will affect our short-term growth curve if we do have to do a bunch of marginal roading now. re longbows we won't have vasslage until yhe GA (by which time i do think we can get both if we do CS first) but even 1-prono longbows seem like gold builds now, enough so that i feel ok about delaying CS for them

[xist10]

You don't have to play in the same half as Nauf, you can start a 3 way turnsplit and attack, after GT has moved. But I would refrain from doing so.

Turnsplits are one of the most common reasons for delays and the chance of delays increases (exponentially) the less time everyone has for their turn.