deity aw 3

[Dirk1302]

Until turn 95. I played again from my 1600 BC save. I had played earlier until 300 AD but i was tired than and didn't concentrate that much which is necessary to play this format well. Besides i had a much better idea now as to how to continue which i wanted to try out. Normally i wouldn't reload a game but i feel it's allowed here as the format is difficult enough anyway. I didn't play it further than this so comments are welcome.

Spoiler :

Instead of going for monarchy as i did in my first try i decide to try the calendar route this time. This will give me useful IW/Sailing/Math instead of med/PH which i don't really need right now. I intend to build loads of workers/settlers so i won't have a > 15 capital that soon anyway. It worked out really well, i managed to improve banana and even dye.

Right now i have 6 workers, 4 are going to improve spices in the next few turns. Horse city looks a bit flat right now but with some farms up it can work horse and 3 grass mines, it's going to spit out spies first units later. I didn't do much to improve Djenne, i will 2 pop whip this city from here on at size 4 or 5 so any further improvements there are useless.

Whip mechanism will be

2 pop spy
over flow 20 H to skirm

So i settled all 3 GG's there.

 

[Habitus]

@Dirk

Spoiler :

If the AI takes the 2 Pop Barb City would it start skipping your front blocking city? If so might need to take and raze it to stop them. The other barb city might be useful depending on what it picks up, its even on a hill Depends on if you can spare any troops at moment for either

 

[Dirk1302]

^

Spoiler :

I don't think it matters much if the ai take that city. They would have to make an enormous detour to get to Awdaghost, a city that is also on a hill. Prime target is Kumbi Saleh but i have to take an eye out for Djenne sometimes so i keep troops on that road atm. Especially the south side of that Djenne isn't good for cottaging.

If i don't get a reli myself i may get one by capturing that barb city from ais later , build a missionary and give it back. Would give me an espionage advantage as well.

As for the rice city, i intend to take that one as it has iron (yes i knew that from my first play through and was too lazy to send that warrior there to establish the fact again. Np if Ai get it first though, it'll probably be lightly defended and again i might get a reli.

 

[Dirk1302]

Until 1 AD

Spoiler :

Continued improving the empire. Got to currency, researched prereqs for monarchy and Col then started accumulating money. I'm cautious with whipping except for Djenne and Kumbeh, both cities whip tons of spies, Djenne alternating with Skirms. 25 BC and 1 AD monarchy feudalism and Col are stolen. I'm actually tech leader now. I have more or less a lock on philo now, can get it in some 9 turns from here. It's probably worth it, i'll get a religion, can switch to pacifism to get a new gspy quickly, probably gambling by putting on scientists as well. If i'll get a GS it can make an academy for even better research, i don't want to hang around in paci for ages as it'll be very expensive on troop expenses. Also OR at a time mints, courthouses, and grocers and not yet built markets come around is huge. I want to be in that civic asap.

I have 9 cities now, not bad for 1 AD as Snaaty already mentioned. I'm hoping to get that rice hill barb city at least before 500 AD, Drill III/Cover lb's must have a + against these archers. Then there is that other barb city that hasn't been taken, i can't defend it right now so i'll leave it alone for now. clam site is still risky to settle right now but it should be possible around 500 AD with walls and some 5 lb's able to defend. It will probably draw some troops but i don't think it'll draw those 10+stacks with loads of cats. Actually i'm thinking of building a fort in the jungle before it, that fort'll be attacked probably instead of the city itself and with 50% jungle it has more defense than the clam city itself.

More than in previous AW games i shuffled units in and out of the blockcity. As soon as Drill IV/CGIII is reached i tend to take units out of block and replace with fresh Drill 4 units. This way i don't have one unit with > 100 XP but >8 defenders with Drill IV/CGIII. I keep those Super defenders on roads between cities so they can defend once an AI stack goes for an inner city.

No idea when i'll get to offense. Techwise CS and machinery are high on the stealing list, i have just about enough spy points left. I'll self research guilds->banking probably from that point. Techwise i'm so strong atm that i may just try to get to the really good things like cavs or at least cuirassiers, then backward charles is easy prey. Medieval war is so tedious with trebs against castles.

Obviously i'm in a good position for lib as well but the problem there is that it doesn't gain here. Lib is useless, the freebie compensates but i'll also have to research paper which is useless as well. I like Edu but not as much as guilds/banking.PP is a great tech as well with all those willages of course but i think that can be stolen. Very much uncertainties, we'll see. At least i'm not in a hopeless position i feel.

 

[Snaaty]

@ dirk:

your save looks very, very strong

...

i also just tried to play up to 1 ad, but my laptop decided to end my set a few turns short (didnt play at home + didnt take my accu with me and went out of battery)

i will post my save + some comments once i find a bit of time so you can compare a bit between the 2 different approaches

1. gw - greatspy (dirk)
2. oracle - col - spy - greatspy (snaaty)

from a short look at dirks save and knowing my save i already can say dirks save seems a lot stronger, at least reseachwise (didnt have time to look more into detail).

 

[LuCiver]

@Dirk

Seems like you have mostly ignored culture and the benefit of accumulated cultural defenses. Is it simply not worth the hammers for a library in Goa to speed it up?

 

[Dirk1302]

^I don 't understand you fully. Gao doesn't need cultural defense. But it needs a library and it will get one once it has grown to 7. It'll be whipped down to 4 but can still work fish/2 fps/incense.

That incense tile surely slows whipping/regrowing down. But it's 7 commerce on 1 tile. When playing it really feels like it slows me down but in the meantime it's so much commerce that i should always work it i feel.

@snaaty, it's not the only comparison i feel. Going calendar first is the winner i feel, that's why i replayed, it's so much stronger than monarchy first (on this map its at least).

 

[LuCiver]

Sorry for the confusion, commenting from my phone. My point was that frontline city kumba salah only has 33 culture and 20% culture defense. Is this because your skirms already have high enough odds in combat, other reason, or just cultural neglect?

 

[Dirk1302]

It's not that easy to get Kumbeh to better cultural defense, needs a second border pop which needs a library. And it's useless, all that cultural defense will be pulted away over the next millenium. Before pults i defended it relatively lightly to give several inexperienced units a chance to improve. It'll need serious defense in the near future. Have to check this turn but there's no big stack right now iirc.

 

[fjordan]

I see tou guys use the Gspy to infiltrate. Are you planning to keep doing that? A settled Gspy is the specialist in the game 12 + 3 before modifiers. Maybe I am thinking a bit too long term but some settled spies plus SY plus maybe 20% slider could free your gold for rush-buying because you can shut off research. Or maybe turtle up and beeline communism, or is that too much of a detour and is it just going for military and going on the offense?

 

[Habitus]

As far as i know you need to steal a large number of techs to keep to a level that can defend against the AIs, settling doesn't get you enough in time to steal things such as Feudalism which are needed to live.

 

[Dirk1302]

The next gspy will be used to build Scotland yard in the capital giving me the opportunity to switch between research backed up by libraries and Espionage which get a 100% modifier in capital.

For the first gspy infiltration is probably best as it gets you a lot of important techs in a short time.
You'll typically get Feudalism/CS/MC + 3 techs like construction by infiltration.

 

[fjordan]

Oh sure, I would use the first one for infiltration too. I had the impression that in the AW2 game Snaaty just kept infiltrating and was wondering what was the consensus on this. But he seemed to aim for a quick domination and then settling is a waste. Anyway, this AW/Spy thing is a whole new ballgame but parts of it are surely useful for normal games so I am curious about the details.

 

[Dirk1302]

Have to check what Snaaty did in AW2. Infiltrating is a matter of diminishing returns quickly. By the time you'll get your second gspy it's ~500 AD, now you can typically steal 2 techs in the magnitude of Nat/Edu. But by building SY you can generate 3000 spy points (the equivalent of infiltration) in ~ 7 turns at 100%, costs gold of course. Since it's hard to breed gspies after the second i think i'd go SY unless i had a direct knockout in store. I wouldn't settle quickly. It's 12 spy points, it takes 250 turns to get to 3000 points from there. Of course this gets better with modifiers too and doesn't hamper research as SY does but it's not good enough i feel.

 

[JammerUno]

^unless you have a city with 400+ commerce, that 7 turns isn't right.

 

[Dirk1302]

SY doubles spy points in city where it's built. Have a burocratic capital and some other good commerce sites why shouldn't it be possible by 500 AD? Of course i assume you run at 100% spy points percentage, that's why i mentioned it costs some gold.

 

[Snaaty]

i only got 1 gspy in aw 2.

next gps were:

gscientist (academy)
gpriest (shrine)
gmerchent (ga)

-> except the ga, theses gp´s were planned (i wanted a ga, so i didnt care about the gp. but a gspy i still would have infiltrated into ceasar stealing pp + rep parts, bringing us closer to rifles). you dont need more then 1 gs to run a decent spy-economy. from 1000 ad on, i was generating something like 100 spoints/turn, when raising the slider to 100% it was close to 600 spoints.

costs are around 50% of beakers, so each 5 turns you get 500 spoints anyways, to steal techs up to 1000 beakers, if you want a more expensive tech, raising the slider 1-2 turn to 100% does the trick. before my whipping orgy, 3 turns a 0% did finance 1 turn at 100% so that was no big deal and i still could selfresearch at a decent rate

...

so far, i never settled any gspy or built the sy-thingy... ...my thoughts went along this line (they basically always do when civving: short term>>>>>>>long term)

 

[fjordan]

SY doubles spy points in city where it's built. Have a burocratic capital and some other good commerce sites why shouldn't it be possible by 500 AD? Of course i assume you run at 100% spy points percentage, that's why i mentioned it costs some gold.

Well as the other commerce sites don't benefit from SY you can leave these out of the equasion. And without SY you can also turn up the slider so you must only count the extra points. In that case you would need tot produce 6000 spy points in 7 turns in 1 city and that looks a little steep to me. I think on the long term, settle first and then SY is better, even a jail is better with a settled spy than with SY. But if you want a quick boost right after you got the spy SY may be better. Allthough I come nowhere near your quality of play I am not convinced here.

 

[Dirk1302]

^Good point about only counting the extra points. I'll give the matter some extra thought. I still think settling takes too long to give a sizeable reward. So maybe infiltrate again as Snaaty suggests.

I take 1250 AD as a cut off point as other priorities such as acquiring land from Ais should take over afterward.

Settling takes a lot of time to compare with infiltration... Let's say i'll get gspy 500 AD and i'll get constitution around 1300 AD. No modifiers before 1250 AD nets 50 * 12 = 600 points (+ some beakers), nowhere near the 3000 points you'll acquire from infiltration. It's better after jails (+50%) but even after agencies (another 50%) it takes time to get to those 3000 points.

Let's do the SY calculation right this time. It's not unreasonable to assume that i will be running at 50% slider (a bit conservative here, might be 60%) approx from 500 AD until 1250 AD. Capital will have ~200 commerce throughout. 100 commerce is going to espionage and SY nets me 100 esp/turn on top of that. 50 turns until 1250 AD is 5000 points, better than the 3000 infiltration esp. But i can't do any self research then as this assumes generating esp every turn. Every turn not invested in esp detracts from those 5000 points. If i infiltrate again i have 3000 esp immed and can now choose to ignore esp and self research enhanced by libraries and possibly universities.

So i think Snaaty is right, if i'll get a gspy again i should infiltrate again but i'll check this again if that happens.

I think in my game gspy first , then a GS would be best. 2 gspies should be ok as well, bad research in that case but 4 major techs for free.

There is one other point to consider. Infiltration fixes points on one target. SY gives flexibility, if i were to attack Charlie with curis SY would be better than all those fixed points on Cathy. Still with esp slider at 100% i'll get enough points in 2 turns to generate the revolt and not every city needs to be revolted.

 

[Dirk1302]

I wonder if i should go philo in my current save. A possible strong tactic would be: research philo take barbarian Bantu, inject it with the reli from philo and give it back to some AI, not Charlie as he's backward. This gives a sizeable discount on espionage against the AI that takes it iirc? Holy city and small distance. Techs may be almost free this way.