Not playing the map

[coanda]

The right answer to someone looking for general advice is often "it depends" sorts of advice. However, I'm a little curious about what stable strategies are (strategies that can work on a wide variety of maps, with a wide variety of leaders, civs, etc. - not strategies which rely on Stables, like horse archer rushes). So I'm going to play a "not playing the map" game, to see how well I can make this go.

Here's how it'll work:

I'll call out a strategy in advance. I'd be delighted to get advice on this from other players before I start. Here's the settings:
RandomScriptMap mapscript (will randomly pick one of Fractal, Terra, Hemispheres, Big-and-Small, Medium-and-Small, Archipelago, Pangaea).
Random size (will automatically add the default number of AIs for the chosen size).
Random Climate, Random Sealevel. Unrestricted leaders, random personalities. Huts and events are on (of course).
Everything else is pretty much default settings - normal speed, Ancient era start, etc. Random leader of random civ. I'm leaning towards Emperor difficulty, but could be persuaded to go to Immortal or Monarch if people think that would be more interesting.

Here's some pretty pictures of some of the possible starts this could throw out:

Spoiler :

Duel, isolated start...


Looks like some sort of Huge Archipelago tiny-islands...


Offhand, I'm guessing small Continents?


What looks like a 2-continent Fractal Standard map perhaps

As you can probably tell, the starting position could vary quite a bit...

I'll play ~50-turn turnsets (I might vary it a bit if there are good stopping points near there) - one following the plan, which save I'll keep; another turnset which is just for comparison, of what I'd do if I wasn't blindly following a preset plan. And a good time will be had by all laughing at the epic failure (or, possibly, epic success) of the game. I hope by picking Emperor I give a reasonable chance at each result, and at the very least that the game will not be quickly decided either way. Anyone else will be free to roll their own start with the same settings and try my plan or one of their own on a different map.

If for some reason the strategy literally cannot be followed, I'll do my best to get back on plan if possible, or just move on to the next step if it can't be salvaged. For example, if I planned on axe-rushing a neighbor, then found I had no copper and no neighbors, I might settle an extra city to get copper, tech Sailing to get Galleys, then axe rush a nearby island neighbor. But if I was fully isolated, I'd build some axes then just disband them and move on. I'll try to plan to avoid that sort of map-dependency.

So here's my offhand thoughts about the plan; I'll probably put together a full plan in more detail after other people have a chance to weigh in (if anyone else cares to).
1. It might be isolated, so all early rushes are off the table; similarly, I cannot rely on tech trades to get me caught up in tech. If there's an important tech, I'd better research it myself. I'm of two minds about a Medieval war... on the one hand, if I get boxed in to just a couple cities early, it might be the only hope. On the other hand, if I'm truly isolated, it's just a waste of hammers and beakers. I'm leaning towards just calling for an early REX to perhaps 5 cities, planning to squeeze in extra cities somewhere lousy if necessary for National Wonders. Then pick up a couple other techs to stabilize my economy, and resume expanding after that if there's still space available.
2. I could start on a small island with only space for one or two cities, so Sailing will be important early (the necessary early tech path could be a mess... AGR -> AH -> Fishing -> Mining -> Sailing -> BW -> Hunting -> Pottery -> Writing -> Mysticism -> Wheel? I feel like I have to cut some corners there, but I can't see a single tech in there that might not be necessary depending on the terrain).
3. Similarly, I could get a seafood-only food start, so I'd better work a Fishing Boat or two into my opening build. I'm not going to try to call every build all game long, but I am going to give a fair amount of detail at least for the opening. Worker-first seems like the obvious start; perhaps worker - warrior - warrior - worker - work boat - work boat - settler? If I'm not coastal, I'll end up just skipping those work boat builds anyways.
4. I'm going back and forth on victory condition. Cultural or Domination strike me as the simplest two to aim for; which of the two is more generally stable I'm undecided on. Whichever I pick will obviously have a big impact on my post-opening tech path and expansion pattern.

 

[xenex]

This sounds fun, in a "explorative" kind of way.

While rushing a very close neighbor is something often advised.I'm thinking it's better to skip that altogether in this exercise,because it's way too easy for that strategy to not be possible,isolated start, no copper/iron/horse close or no neighbor close-by.

It seems to me the best stable (as you put it) strategy, is (at least at start) the most peaceful one.Plop down settler, aim for going for 6 cities as quick as possible, salvage/stabilize your economy and go for culture.I know it doesn't sound like the most fun, but I think its the most stable

Although going for culture might run into some problems because of the religions, if you end up on a isolated map it could take a long time before you get religions to use to boost culture, going for religion, is also not a stable strategy and it doesn't work for all the leaders, especially the first 2 religions.

(something that came up while writing the above, can religions even travel to you while your isolated and haven't met anyone yet or have a explored access to somewhere with religion?)

 

[vranasm]

not sure what the aim is here honestly.

My aim going into every game is this strategy:

1) domination victory
2) using cuirassiers->cavalry for mopping up
3) getting good number of cities in BC era as whip base (can be combination of rex, BC era warfare)

So the "play the map" mostly refers to first 130 turns of normal speed game. From the T 140(150) it usually is the same every game ;-).

Of course some map situations don't allow cuirs/cavs break out (here I refer isolation basically, in all other situations you surely can use cuirs/cavs to own your continent with getting to space as free card to win)

Btw played the "RandomScriptMap" script like 1 month ago, got semiisolated with Sury getting 13 cities (him 20) and well balanced other continent (all other AI's all stagnated with ~10 cities) ending with culture (was testing some BC era CS strategy with Willem and got confu, tao and of course Juda from Sury)

Why I talk about this is...you're sure that it really chooses? because that kind of map didn't looked like any of the maps you mentioned, no familiar resource patterns, land distribution etc. all looked like unique map script.

 

[Um the Muse]

This looks like it's more for the laughs than for strategy, because you really want some sort of insight into your map and everybody can at least tell whether they need workboats or not. That said, this looks like fun

I noticed that the majority of your maps have much coastline and very little jungle. If you're doing this completely blind, though, you'll have to plan for the chance that you've got a wall of jungle five feet from your borders.

Also, with huts on, I would recommend hunting as a higher priority. There's always the chance that you end up with Deer as your food anyway. I also think that AH needs to come much later. Mysticism is never necessary; it can be really important in limited circumstances, but most of the time can be safely put off indefinitely. You can't have Pottery without the Wheel, sorry. The wheel's importance depends on the difficulty level, imho. BTW, why not set the difficulty to random, too?

Actually, since AH is such a map dependent tech, you probably want to put it off for after you're getting a steady stream of commerce and can afford it. If you do it that way, then Hunting also loses a lot of its appeal. I would say it's either something you should get really early (like right after Agriculture) or not at all.

 

[vicawoo]

Isolation kills off all non-peaceful strategies. I suppose you could always try to force a cultural victory.

Otherwise, I find a good catapult rush pretty reliable, as it deals well with high unit counts. Or steel plays.

 

[CaF]

I've been thinking about the stableness of things a bit as well. Finding things that are always correct and "never depend" is pretty difficult. There might be some, but they don't come to mind.

I'd say a culture win or an early rush into a culture win always works on Immortal and below. Now, that of course depends on the start. Whenever I increased the difficulty level I went for a couple of culture wins first.

If I understand the topic correctly what we are trying to do is going to be pretty idiotic-robotic. Are people here able to program a bot or something? Or "hack" an AI and let it play our strategy? That would save us a lot of time playing blindly.

 

[coanda]

If I understand the topic correctly what we are trying to do is going to be pretty idiotic-robotic. Are people here able to program a bot or something? Or "hack" an AI and let it play our strategy? That would save us a lot of time playing blindly.

Short answer, not without a lot of work.

Long answer... with the way the AI is set up right now, changing it to let you hard-code a specific plan would mean almost ripping out the entire guts of the game and re-writing it. There isn't so much one "AI" file you could edit as a half-dozen AI files (e.g., player AI, team AI, unit AI), plus the occasional borderline-AI function just tucked away in other files that are part of the main game engine. Many of those AI functions aren't calling each other, but instead are called as-needed by the game engine. Further, all of those files were designed specifically to be mod-friendly, meaning they don't have nice spots hard-coding stuff like "build horse archer" - because some mods might not have horse archers. Instead they might have a heuristic which gives extra value to 2-movement units, extra value to offensive units, and has a nice hefty random chance weighed in, which will sometimes lead to them naturally building horse archers.

 

[civvver]

This is interesting, and it might just work on emperor. If you say I'm definitely researching bronze working, chopping pyramids and running representation, peacefully expanding and teching towards liberalism for a cultural victory, I can see this being possible on any map type, with any leader. And you will usually be able to accomplish this on emperor. Only super hard part will be if you have no stone, non industrious leader, no production and no forests. Having any one of those will probably make it possible.

Despite people always saying play the map I find myself doing mostly the same things every game anyway. I think where this gets drastic is you go from fractal and archepileago maps to continents and pangea. Being isolated and on islands makes a huge difference.

 

[Mantic0re]

There are different levels of "Play the map".

Worker Micro and city placement are near the lowest level and a given VC at the top. Start with Automated workers and shoot for a previously chosen random VC (or score based VC) to eliminate those map-based decisions.


A list of stable plays should include:

Worker first
Slavery when available
Whip away unhappiness
Get non-warrior military unit for barb defense asap
Cottage 1-2 green cities
Hereditary Rule when available
Research what AI doesn't research and trade techs


I'm still learning Monarch so some of these may not apply well to emperor or above. Higher strategy decisions (DoW targets, current era unit builds, city builds and timing, etc,) are very likely to require some map-based considerations.

 

[gorf37]

This is a fun idea. It depends how much playing the map you're willing to allow. For instance, how do you decide where your 2nd, 3rd, 4th cities will be placed? That can have a huge effect. Still, Emperor is the highest level where I could see the concept of a pre-set strategy working out.

Some elements of my 'brainless' strategy:

1) Worker techs first. (ie. no early beelines) AG,Mining,BW,AH,TW,Pot (up to you if order has to be pre-determined)

2) Spawnbust with warriors to decrease barb pressure

3) Cottage the capital for future bureaucracy. Yes, with some capitals this is a really weak play...

4) 2nd or 3rd city must be hammer specialized to pump military. The other must whip a library ASAP and run 2 scientists for a GS who will build the capital's academy.

5) After worker techs go Writing, Math, Currency. Build wealth while teching CoL, Bureau.

6)Plan on having an AI research alpha while you're working on Currency/CoL/Bureau. Trade them for Alpha and backfill as much as possible.

Typically this is all heading towards: 2nd GS bulbs Philo, 3rd GS partial bulbs Education. Lib>>MT, build cuirassiers>>win.