[SPOILERS] swance bitten, twice shy

[Chevalier Mal Fet]

The key thing to remember about multiplayer wars is that the enemy is almost always going to defend better than you expect him to (but also almost always won't attack as strongly as your worst case doom-predictions). Humans are extremely inventive and when properly motivated will find lots of ways to surprise you. So come to a reasonable estimate on the number of troops you'll need (based on things like how many troops HE can build, assume roughly proportional losses, ground to be covered, etc) - and then add 50% more to that.

Personally I prefer to continuously build troops all through a war until the outcome is no longer in doubt.

[Tarkeel]

Personally I prefer to continuously build troops all through a war until the outcome is no longer in doubt.

I second this. The most common error is to stop whipping out forces because you fear they will not be needed, and then not have then when they're needed.

[ljubljana]

yup, that's the plan. with 3 exceptions (wakamotoharu overflowing from a horse whip into the almost-finished moai statues, our third-best commerce city with the double gold building a LONG-delayed market, atamifuji which hard-requires finishing the last ~30h of its library or it can't run specialists during the GA) every city is on military builds and will stay that way for the forseeable future. and i'm not 100% confident the exceptions are correct either... i also get the sense that in civ4 the point where "the outcome is no longer in doubt" comes pretty late as there will always be the danger of our other neighbors whipping out their own stacks to pile on, which is why many of the troops i expect to build during the war will be not horsies but (CG2 and later CG3) longbows to secure our flanks (and cats which we need for both purposes)

only question, i think, is at what point during the war do we want to stop whipping specifically and switch over to hand-built units. in theory we could whip every city down to size 2 to get even more units, but that seems like it would crash our economy so hard that we then couldn't win even with a successful war... so presumably there exists SOME point where more whipping becomes incorrect even during wartime. the big question, i think, is whether that point is near enough to where we are now that we can afford to flip to serfdom at some point during the GA. it may not be necessary to do so right at the beginning of the GA since we have whipped off a good percentage of our farms, but being out of it long-term would be pretty devastating economically... not sure what the right call is and it probably depends on how the war is going in 15 turns or so

edit: also there is a question of if, during wartime, a riverside plains hamlet/village is a sufficiently bad tile that we should whip off them. my current perception is that, even during war, we should not whip off resource tiles (including 1f ones like horses, but excluding stuff like 0/5 desert hill copper) or grass river cottages, and definitely should whip off plains river farms at least until serfdom... but not sure about plains river cottages and grass non-river cottages.... the answer might also depend on how many commerce multipliers are in the building (this is why hoshoryu is starving right now, as i gave it all the plains cottages in the area). what do you guys think?

one wonders if by this logic we should be delaying the GA until we have higher pop and are better-positioned to take advantage of it..... i do not think so though for one very simple reason - vassalage is worth a free promotion on literally every single unit we are building AND for most of our offensive units it is the high-value second promotion level that opens up cover/shock/sentry/etc as well as guerilla 2 on the units that can take it. but that does mean that other than the vassalage swap we are at some risk of getting very poor value out of our GA, especially if we continue to whip off tons of pop even during it....

edit: re the mess at atamifuji, i really want this city to produce the 300 point great person so the 50 banked points from chichen itza (which will be more like 100 by the end of the GA) will get used for something, and because the prophet points in the pool lower the odds of getting the same GP as from kirishima and thus being unable to launch the second GA. but that could be wrong and we could simplify things a bit if there is some other city (likely the capital) that has market and library in place already and can afford to run 4 specialists during the GA. not 100% sure it would be correct but i'll look into this

edit: btw, i did not expect this to materialize THIS early but AGG is now up to saving us a full 17 gpt on unit maintenance costs..... it looks like by the time our full stack is assembled it might be already rivaling FIN and ORG in economic impact

[williams482]

I'm detecting a bit of irony RE worries about workers having nothing to do, and working unimproved hill tiles for micro purposes. I know mines are suboptmal tiles to be working constantly, but it sure seems like having more of them around to provide extra production when a couple more hammers really matter, or to give Waka a way to progress on Moai that won't leave it unable to work Moai tiles for lack of happiness, might be well worth it.

Also: no harm on putting an Impi in a scary spot in view of that worker, but I'm dubious on the value of declaring early in order to capture it.

[Tarkeel]

Mines are mainly a bad tile to work when you're in the low-pop whipping stage; once cities grow and cap out on happy/health, they're actually pretty decent tiles. Still not as good as riverside cottages, but then what is?

[ljubljana]

yeah, you're right and that hill should definitely have been mined, and i'm sending a few guys (well, one guys) up to WMH for the same purpose. that said i realized having 50 turns of whip anger at WMH is not as dire as i thought due to the imminent arrival of hereditary rule (and the need for troops in the area anyways for zone defense) so i may just whip a few more troops to finish moai

yeahhhh generally i've been resistant to mining for that reason @tarkeel.... by the time these cities DO cap out, we'll have windmill access and that will be a better tile due to the low-food variant we seem to be playing  so i couldn't think of a timeframe in which we'd ever really want to work mines, except for micro reasons. but if the workers have literally nothing better to do, then of course those micro reasons are sufficient justification

re the impi/worker sitch, yeah, i decided against it because indeed, i wasn't going to declare early to capture it. and i think there WAS a little harm in moving up a threatening guy if i definitely wasn't going to attack early.... anything that might make GT invest mental clock cycles into whether they have enough troops on the border with zulu seems to risk prompting early whipping and redeployment which is the main thing we want them not to do right now

Turn 131 - Zululand

we can finally actually see ginger's research and it is....



WOW that was fast! nice :D it's in caesarea, one city south of the capital, which is closer to drey than to ours to make taking the whole group of religious wonders in a split unlikely. still, this manifestly increases the rewards we could get from attacking them and i'm starting to think seriously about the situations in which we might be able to do so. assuming we DO get the full benefit of capturing sankore/AP/minaret if we also capture the christian holy city and convert to the religion - is that accurate?

right research, yes, they are researching construction ok cool. honestly i'm pleasantly surprised they still didn't have it



surely THIS is enough of a power spike to make GT do something? right....?



i guess i missed this in the event log but looks like nutty bars was re-recaptured. gav's back in slavery now too after a long period in SPI serfdom, so hopefully my worries about a collapse from them were a little premature

btw, how do the medic promotions work? if i already have medic I on a unit, adding medic II does NOT further increase the rate of healing for units in the stack right, only for those in adjacent tiles? also, if i have a medic I unit in a stack and a medic II unit one tile away from the stack, then the stack gets the effects of both promotions right? i see that gav made a super medic with their first GG and i'm wondering how much i should prioritize doing the same



here is a good view of the dueling nauf/GT stacks.... to face this, GT has 18 attackers (the majority axes) and 9 cats. so.... enough to kill the nauf stack if it invades, i think, but not enough to invade in its face. which means both of these stacks have just sat here uselessly for endless turns on end, each trapped in place by the other and with both sides apparently feeling compelled to pour crossbows into them to avoid falling behind



accidental micro as daieisho overflows exactly 50 hammers for horsies on consecutive turns



going to preemptively ask greenline to extend our fish/fish, which elapses next turn, as we have this worker on the border farming in a dangerous locale. and ofc i REALLY don't want them to start making eyes at our border cities



i did whip atamifuji, as you can see, and probably will whip the capital as well. ichiyamamoto is trickier... however one thing that is easier about it is quantification. it will take 7 turns to regrow, so a whip gets us 30 hammers right now, at the cost of 7 turns working a silver tile = 21 hammers + 42 gold. so we'd be paying 42 gold for 9 hammers and the tempo advantage of getting the horsie to the front 3 turns faster (and subsequent units but those will take a looong time to be produced). is that worth it? seems dubious to me, i think i'll hold off for now 



our delay on CS has allowed us to pick up not just the feudalism prereq bonus but 3 KTBs as well :D feeling very good about that move seeing as we're not even working the riverside plains farms now for the most part. maybe even good enough to delay CS even further in favor of machinery, which provides semi-useful worker tasks in windmills while also giving a unit we might make. definitely enough to delay for iron working, which we will need to make most of the upcoming strong units (not sure why crossbows require iron specifically but so it appears to be)



stack is looking a little less motley and a little more real now, yes?

another 9 troops are coming out this turn, though some are defenders on the borders to replace impis that i preemptively transferred. abi came off a floodplains village (!!!) for the stone, because that's what needs to happen to produce exactly one longbow on exactly turn 133 to cover our workers going for the combat road. hoshoryu is inadvertently ALSO overflowing exactly 50 hammers into another unit next turn, which could be a second longbow, but perhaps we should prefer a horse archer or cat... but anyways, you guys can expect to see a LOT of longbow builds next turn, as this next round will come out during vasslage, and i think g2 longbows are comparable to horse archers on attack while being a much longer-lasting investment and of course HUGEly superior as stack-coverers and defenders in captured hill cities



idk if this is worth a highlight, but for the first time all game we are working every grass tile in the eastern river valley  kirishima needs to run exactly one don't-starve turn to avoid losing time on the golden age, and this seemed like the turn to do it as otherwise we would have an unworked grass cottage. well...ok belatedly the actual time to do it was a while ago when we didn't have to work the lake but whateverrrr

[xist10]

Medic I and Medic II should not stack.
Medic (I or II) and Medic III do stack and Woodsman II stacks too.

I think, irrigation spreading is not the most important part of CS.

[ljubljana]

Medic I and Medic II should not stack.
Medic (I or II) and Medic III do stack and Woodsman II stacks too.

I think, irrigation spreading is not the most important part of CS.

so there is no way at all to make medic I and II stack? even having a medic I unit and a medic II unit in adjacent tiles? if so, what is the point of ever taking medic II.... it seems like the scenario in which we'd want to have 2 stacks of units healing next to each other should be extremely rare...

agreed about CS... it's only on this oddball food-poor map that irrigation spreading seems important. but the other two benefits are maces, which require machinery anyways, and bureaucracy, which i'm not using until at least the end of the GA in favor of vassalage. so irrigation-spreading would be the only reason for us to get CS BEFORE machinery. and imo it is not enough to justify doing so, with our cities that would otherwise want to grow onto plains farms probably content to just stick to whipping units for now. and once we hit machinery the choice will be CS vs guilds or engineering, and i'm not sure we will pick CS in that situation either...

[Tarkeel]

assuming we DO get the full benefit of capturing sankore/AP/minaret if we also capture the christian holy city and convert to the religion - is that accurate?

I forget if I've said this before in this thread, but it's worth repeating: Apostolic Palace will lock onto the state religion of the builder, and give the benefit to all buildings for that religion regardless of owner. The other three monk wonders (Sankore/Minaret/Sistine) follow the state religion of the current owner. In no way shape or form do you need to control the holy city, that just gives you an extra bonus for spreading the religion around in the form of shrine-money. However, in order to get the benefit of AP and captured Sankore/Minaret/Sistine you need to swap to the AP religion.

i guess i missed this in the event log but looks like nutty bars was re-recaptured.

The re-capture would only show up in the event-log if you regained vision on it while under new ownership.

so there is no way at all to make medic I and II stack? even having a medic I unit and a medic II unit in adjacent tiles? if so, what is the point of ever taking medic II.... it seems like the scenario in which we'd want to have 2 stacks of units healing next to each other should be extremely rare...

Yes, Medic II is only for spreading the effect to neighboring tiles, which occasionaly comes in handy but often isn't. It's mainly taken as a pre-req for Medic III.

[williams482]

Short and sweet: Units will get the healing boost from whichever single unit offers the best benefit to them. If you have a Medic I unit in the stack, any other Medic I or II units are irrelevant to units on that tile. If you have a Woody III unit (+15% healing) then that will supersede the +10% from any Medic I or II units who would otherwise affect that stack.

Medic III units are awesome because they combine the +10% from Medic I and the +15% from Medic III onto a single unit, which means the stack benefits from a combined +25% healing bonus. The near-mythical Medic III + Woody III superhealer combines +10% from medic I, +15% from Medic III, and +15% from Woody III for a +45% healing bonus.

As for the actual value add here, those promotion bonuses stack with the base healing rate for the tile a unit is occupying: +5% for enemy territory, +10% for neutral territory, +15% for friendly territory (including cities currently in resistance), or +20% for a friendly city which is not in resistance. So after capturing and entering a city, a Medic III stack will heal all units by 15 + 25 = 40HP per turn, while a MedIII+WoodyIII stack will heal by 60HP per turn, the basic Medic I stack will heal by 25HP per turn, and the stack with no healing promotions at all will reclaim just 15HP per.

In hostile territory, those numbers become 15 HP/t for Medic I, 30 HP/t for Medic III, and 50 HP/t for MIII+WIII. Really, really big differences there. I'm guessing medics are less enormously powerful in multiplayer than in singleplayer, when I'd imagine stack combats are more likely to be resolved decisively in a single turn and healing up collateral damage on the interturn is less likely to happen, but they will still enormously speed up conquests when units can go from 20 HP to full by sitting in an occupied city for two turns.