[BNW] First attempt to win on deity - Capital loss 64 turns

[mbbcam]

Playing Deity it is really important to learn how the Civs act. Know your enemy. I have stolen a worker from one and had them offer a declaration of friendship a few turns later. On the other hand I've done it to Spain and they have refused to make peace for the rest of the game.

If you want to learn, you will have to try things. Stealing workers is a good way to weaken the AI.

 

[YomoKimyata]

Following this as I'm making my first bona fide run at Deity as well. How much of deity success is 1. surviving the first 100 turns; 2. having non-aggressive neighbors; 3. other?

edit: versus immortal

 

[beetle]

How much of deity success is 1. surviving the first 100 turns; 2. having non-aggressive neighbors; 3. other?

I would say that 1 is 50%. 2 is 50%, as just not having Shaka nearby is half of that. 3 Other is maybe 10%. Does not add to 100% because math is hard.

 

[beetle]

How much of deity success is 1. surviving the first 100 turns; 2. having non-aggressive neighbors; 3. other?

I am changing my answer because I think surviving the first 100 turns is entirely dependent upon having not-so-aggressive neighbors!

edit: versus immortal

Sorry, can’t really say, as I have not faced immortal maps where losing was likely.

 

[raider980]

Following this as I'm making my first bona fide run at Deity as well. How much of deity success is 1. surviving the first 100 turns; 2. having non-aggressive neighbors; 3. other?

edit: versus immortal

I am not good at Deity but I have attempted quite a few Deity games, so my comments are from that perspective.

I agree with Beetle - survival and non-aggressive neighbors go hand in hand. The thing is the AI is far more aggressive on Deity. I think (better players can confirm) that the reason they tend to attack more on Deity is because you appear weaker because of how far behind you start.

I usually play immortal and the only two I really need to worry about are Shaka and Alex; sometimes Dido. On Deity there are more leaders who will attack you early.

Also, on Immortal, I can usually deal with them even if they do attack because I am closer in tech to them in the early game and don't need quite as large an army to deal with them. On Deity, they are ahead in tech plus I don't have enough money to maintain a large enough army to defend.

For #3 other, I think the main thing is precision and attention to detail. That includes scouting, build order, tech path, as well as knowing when do to x, y and z; I am sure there are more.

 

[beetle]

The other thing of course, even with the same settings, not all start are equal. For example, the game’s algorithm under values dessert and over values plains. Consequently, if your start has lots of dessert, it will probably be strong. If your start has lots of plains, it is prolly relatively weak. The algorithm also heavily discounts resources in the third rings, so always consider moving a hex or two.

 

[Nizef]

what...? stealing workers from an AI seems really dangerous...especially on this level where I am never going to have the military that the ai has...

May I suggest that you make it part of your play in almost every game to steal at least one worker from an AI and one from a nearby (hostile) city state. This is important also from a production perspective, as it free up the building queue for other important stuff and thus get you off to a faster start. Stealing from an AI has of course to be done in the very early game, preferably before turns 30-35 (standard speed). The AI will usually only have 3-5 military units at that time and as long as your warrior has a clear path out of the city bombardment range, it should cope with city bombardment + one (or even two!) hit from a unit. Prioritize cities that are founded on flat land. Hang around the AI city for a couple of turns before you make your stealing effort. This is important for knowing where the AI military units are. Stealing also has an other objective than just getting a worker for free: it slows down the AI. Don´t fear that AI will attack your capital so early. They rarely do and even if they would, you can beat their small number of units with just city bombardment if necessary. 95% of the time the AI will make peace with you after only 5-6 turns. AIs like Dido or Izzy might get crazy and not offer peace at all, but even this you can cope with: build a few units yourself and kill off any incoming enemy units.

Of course there are situations when stealing is not possible: if the AI has the +30% city bombardment pantheon and the city is founded on a hill, you have to cancel your hit-and-run effort. Even spearmen might die!

 

[Blatc]

With worker stealing, remember that if the AI is willing to make peace, that means you can steal another worker and immediately make peace. For the same warmonger penalty, it's better to get two or more workers - perhaps some caravans - so be sure to set yourself up for another steal and check to see if they're willing to make peace before moving. Consider what will be high priority tiles, which you can learn from observing as well as common sense (tiles with horse resources; mining luxury as they start the game with that tech).

Look for great situations, such as:
*hill tiles with mining resource in 2nd ring with a river between the hill and the city hindering melee attack
*high priority 3rd ring tiles
*for city on flat land, 2nd ring tile on hill or with a hill in the way, so that a ranged unit in the city cannot attack you. Less common, but same idea with a mountain between the city and 2nd ring target tile.

You can be more aggressive with CS, since you can make peace immediately; but, you can similarly try to set up repeat steal scenarios to maximize your one 'free' CS DOW.

 

[beetle]

Replying to this question for the third time, I guess I think it is interesting!

Following this as I'm making my first bona fide run at Deity as well. How much of deity success is 1. surviving the first 100 turns; 2. having non-aggressive neighbors; 3. other?
edit: versus immortal

One big variable I find is if there is a far-away run-away civ. If that AI is going win by SV or CV by T300, there is not much one can do — except to be faster than that.