Deity Semi Isolation

[Lain]

Challenging map with a fun neighbor, thought it's worth sharing. What's the best approach here ?

Spoiler :

 

[Anysense]

The start looks pretty good. What can make it challenging? The land is just too good; there must be something totally crazy about the map to make it difficult. For example, some real fun neighbour like Monty or Shaka and no metals.

 

[Pedro78]

Played a few turns, and can already tell that the best approach is most likely..

Spoiler :

HAs -- with that much production, even mighty GK will be crushed. Also better to tech Archery early in case he declares first.

I don't like the "gift a city and pray for the best" approach -- it does usually work but leads to very long and tough games later on.

 

[Pedro78]

Spoiler t50 :

Looks like he'll have ~10 cities when I attack. What a fun neighbour!

Edit - after all let's not be so optimistic

Spoiler :

At least I'm not a landtarget, not sure if that matters in semi-iso though.

 

[Anysense]

Funny how we settled the same spots but in a different order (swap Medina and Damascus).
Thought I could do with a little break from time-consuming space games and played it until t80.

Spoiler :

Reached archery on t69 and started whipping HA. Built 3 barracks and a few chariots while teching. On t70 GK is already at my doorstep with swords and axes. Not sure whether he was actually targeting me, there is a barbarian city in the north.

Spoiler T70 :

I declared war and killed the lone axeman immeately, wiped out his stack on the next turn with no losses and took cease fire two turns later. That was lucky, winning 6 fights in a row at 70-80% odds.

T77 getting ready to DoW again. Mustered 8 HA and 5 chariots

Spoiler T77 :

T79 I take Beshbalik only to lose it on the next turn. I retake it loosing most of my stack and there is no way to keep it. GK won't take peace. This is the end, I think.

Hard to beat someone like GK when he's got metals and you don't, especially with a subpar leader. That would require a very thoughtful approach.
Perhaps the key is to get to horse archers ASAP. Settle gold first and skip Pottery; or tech Pottery early and do some cottaging; or settle only 3 cities; don't bother with early warfare and ~8 cities. I doubt that the last option is any good, GK will only become even bigger problem later. May be, Oracle->Construction?

Edit: As an afterthought crippling him with chariots early on might be very efficient.

 

[Pedro78]

I played a bit further, will write-up tomorrow.

@Anysense

Spoiler :

• Did you research AH before BW? Would explain why you settled sheep first, but kind of a weird move when you have all this food+forests & gold spot.
• On most maps I'd have researched HBR before writing like you did, but here we have so many forests that maths is a tremendous help. I'd rather attack a bit later with a bigger stack of 5XP HAs -- Agg melee units are no joke. Another advantage of getting Writing first is that it allowed to scout him out.
• You're playing without BUG?

 

[Anysense]

Spoiler :

Why weird? Food first, as usual. Not sure what to do with early chops. Quickly expand to even more than 4 cities? That would cripple research and t69 HBR is already extremely late. Normally, one must attack with ~10 horse archers around t65 to stand a chance. Nah, I'd rather chop into horse archers.
Perhaps, better approach is to steal workers with chariots and try to keep him away from metals. No idea how to pull this off though, he's got horses too.
I forgot about BUG, been playing only with BUFFY lately.

 

[Pedro78]

Spoiler :

Well there's plenty of food already (can share one corn), and a few chops speed up expansion a bit. Getting the gold earlier allows to reach everything else quicker. I only ended up not chopping anything until writing because GK stole the western cows spot very early.

I remained at three cities until DoW -- didn't scout the corn spot early on because I didn't want to lose a warrior to the barbs. Anyway, I don't think I'd have settled a flatland city right to GK when planning to attack him early -- would leave me very vulnerable to an even earlier DoW.

Right now plan is to declare a bit earlier to let him suicide most of his stack on fortified CG2 Archers in Mecca. Then take his flatland corn city (lightly defended), wait for the rest of his stack to move in the open and destroy it. Then he'll defintely take a cease fire.

Normally, one must attack with ~10 horse archers around t65 to stand a chance.

I disagree with this for several reasons. First, very few maps allow for a T65 DoW with 10HAs. Second, attacking this early means you'll have very few hammers from chops (probably only 2 or 3 workers), which means lots and lots of whipping, which is incompatible with the low happy cap, and will make recovery extremely hard after war. Another thing is that Maths opens up Currency, which will be required after war. And then having 4-5 workers & Maths chops combined with whips allow for very good timing to build an army efficiently. There are just too many good things coming from teching Writing & Maths to ignore imo.

Edit - in most HA rushes I attack around T85, but depends on the map ofc.

 

[Anysense]

Spoiler :

T85 HA rush Usually there is at least 40% culture in every city worth conquering at this point and lots of metal units, even macemen if you are unlucky. A couple of times my elepult even met pikes around t90. But if you are successful with HA rushing that late it is truly remarkable. I would not even think it's possible.
Yaeh, t65 HA rush is a bit difficult and can not always be done. Here I managed t71 attack after settling on marble, not an obvious move here of course. See how much earlier attack date means.

Spoiler Pics :

Easy as a pie.

 

[Pedro78]

The AI never has maces/pikes before at least T100 in any "normal" game. T90 longbows happen if you're really unlucky. I think you're too used to play HoF starts with handpicked AIs, hence why the dates you're talking about are so early. I also tend to play below average starts, so maybe an "average" HA rush date would be T80.

Spoiler :

See, you're talking about a T65 attack "to stand a chance" but are "only" attacking on T71 even though you have BFC gold. I'm not criticizing your play (T71 rush is very respectable on this map), just saying that your timeline assessment is wrong (for 99% of random maps)

And of course a much earlier attack date makes a big difference, but with such a good start it's more advantageous (imo) to attack later with a better set-up, because the game is not over after the rush. I will ofc not have such an easy time taking out GK in my game as you had in your T71 attempt, but my position around T100 will probably be stronger.

 

[Lain]

Interesting, I chose another approach

Spoiler :

Will write-up at some other point, but the gist of it...

- Gift city
- Stay at 4 (maybe 5) cities
- Stick 10 archers and walls into border city (protective for the win)
(- I could imagine going Horse Archers purely for defense, to flank out catapults later?)
- Build Great Library + NE
- Hope for the best

Basically if you survive until Astro and trade for Engineering etc, it doesn't matter if he has 12 cities and I have only 6 or so.

Is taking him out really better? Big investment and we are left with a big continent, but more than 6 cities in isolation are not possible anyway, so most land will be unused/razed until Astro. But Astro will probably come rather late.

 

[Pedro78]

@Lain somewhere I saw you mention a civ4 combat odds calculator. Can't remember where tough. Could you post the link here pls?

 

[Anysense]

The AI never has maces/pikes before at least T100 in any "normal" game. T90 longbows happen if you're really unlucky. I think you're too used to play HoF starts with handpicked AIs, hence why the dates you're talking about are so early. I also tend to play below average starts, so maybe an "average" HA rush date would be T80.

Probably, this. IIRC you are a bit unlucky to face t90 longbows in a normal game... In the last HoF game I actually saw t74 longbows from Darius; and I really hoped to take his GLH. Such things of course may alter a little the understanding of what is early. I don't remember how HA rush works in a normal game, ceased using it long ago. Not much stronger than axe rush but comes 15-20 turns later and has a very strong and cheap counter. You may of course have no choice when you have horses but no metal.

Spoiler [USER=212405]@Lain[/USER] :

Is taking him out really better? Big investment and we are left with a big continent, but more than 6 cities in isolation are not possible anyway, so most land will be unused/razed until Astro. But Astro will probably come rather late.

Oh, I don't know. There is quite a bit of riverside, so it may actually work if conquest gold will be enough to survive till the cottages mature and economy becomes stable. Kind of thing difficult to figure out before you actually play it. A lot depends on luck too. For instance, in my second attempt GK built Mids for me

 

[Lain]

Yes, combat calculator here: http://civ.zulan.net/vodka/

 

[Pedro78]

Awesome. Thanks.

more than 6 cities in isolation are not possible anyway

https://forums.civfanatics.com/thre...orm-fractal-nh-ne.517825/page-2#post-12985108


Edit - ofc this was with FIN leader & GLH, but here we have good land aswell. The point is there is no hard rule.

 

[Pedro78]

The first 2 screens should say everything about the benefits of Writing & OB.

Spoiler :

He has only one iron (scouted the whole island)




And his units are ready to come & suicide on my hilled, 40% culture, behind-a-river, CG2 Archers.

A couples of screenies -





Sadly he didn't suicide all his metal units -- spear and axes didn't attack. But I don't think he has much left. Could also send all of Karakorum's HAs as scouts to his other cities before peace, only saw a couple of axemen. He almost retook his border city, but I had a few HAs coming in anyway so it wouldn't have been a big deal.

Basically I teched Pottery-Hunt-Archery-HBR after Writing. I ran max specs in Mecca (2 priests + 2 scientists), plan was to either bulb maths or settle a GP. Popped a prophet (would've prefered a GS, but I still prefer a Prophet now than a GS 6-7 turns later). I originally planned to attack with more units & bait his stack, but when I discovered that his only iron was in his flatland capital that could be attacked on T1 I didn't wait any longer.

Concerning the rest of the game, conquest gold will get me through Currency, Masonry, Aesth, Litt. Karakorum has enough forests to quickly chop GLib & NE. After that techs will probably be Calendar --> Astro, Monarchy not needed with +3 happy (4 with forges). Might go for CoL though. Tech will be slow for a while after war, but I'm sure that Astro wouldn't come any later than T150.

I don't think I'll play this through (mostly lazy), but that was definitely a fun start, a good reminder that double corn+gold does not necessarily mean an easy game. Thanks for the map

 

[Lain]

Interesting, I always thought that attacking in semi-iso is a big no-no and I only did it if boxed in (limited war to reach 5-6 cities at least). Hard to weigh conquest gold against tech detours, unit cost etc. I remember another map that forced a construction attack with Axes and I failed hard. Got attacked before I was ready and 3 vs 10 cities is a bit annoying with one-movers. Captured quite a bit, but went broke and abandoned it.

I need some meta for orientation. When to pursue military solutions in semi-iso?

 

[gaash2]

not sure it works on diety but on imm I've been wiping the other guy out completely, whipping courthouses/wealth everywhere, and then catching up in tech very quickly with many cities.

 

[Pedro78]

I don't think seeking meta is a good thing -- that's the conclusion we came to after months of seeking meta for isolation.. Can look at some patterns to get a better understanding of certain situations though.

The main thing to consider about attacking/not attacking, whether or not in semi-iso is the cost/return ratio of the war. There is only one cost that increases dramatically in semi iso, and it's the cost of war recovery. When you're boxed in with few cities, war is almost always profitable, but as you stated, there's no point in a total war if you simply go broke at the end. So the biggest question about war in semi-iso is probably "what will I be able to leverage in order to get my economy back in shape?". Here it was good/great land + marble + lots of forests. In other games the AI can build the Mids or the GLH, or just a bunch of wonders that give GPP. Or it might tech Alpha in which case you can take peace for some techs. A lot of stuff can happen, but there's few cases where I'd just sit back and tech when boxed in. Oh and don't forget about Axe-rushes when you have copper -- these are really powerful.

I'd be interested to play the map you've just stated if you still have the save.

 

[Lain]

Don't know, I do have some basic meta I follow for isolation games and it works quite well 90% of the time.

4 Cities asap, Monarchy, 3x GS (academy, 2x Astro), possibly city #5 at Monarchy, in exceptional cases #6. Attack with Cannons/Rifles, use Globe Theatre (this pushed me to the next level, always build this now), deal annihilation, preferably to target without own Rifles. Works well. Most maps have some nice space for cottaging at least. No more minimalistic approaches.

I ignore stuff like Aesthetics and CoL for the most part. It hurt my brain on that Mao map, trying to figure out what's best. Turned out when I followed that "meta" and just ignored marble I got way better results. And ultimately, if someone tried to learn Deity Isolation from scratch, something like "just play the map" is not so helpful. I need quick guidelines for my playstyle, especially when recording, don't want to do 15 hour series anymore .

Adjustments can still be made like skipping Monarchy with lots of happies, but I go into the game with a clear framework. Maybe it's time for a Semi Isolation Workshop! Because right now, if I attempted a 100-win-streak, getting mauled in a nasty semi isolation seems most likely to me. My biggest nightmare, next to JC without metal. Got an idea for that one except "gift and pray"?

Sure, here is the map. Dunno how trivial it is, but some documentation of build-up and economy balance would be very valuable information for me.