China

[Txurce]

I played China on Emperor level with 7.2.6. Given the Paper Maker and the CKN's mid-term arrival, I decided to pursue a conquest strategy built around strong science. I found myself on a continent with four other civs, and had no problem with any of them. I focused on Liberty and then Honor, while researching to Machinery and Physics for a Medieval war machine. In the process I built four cities: the capital and the next one for the hammers, the last to for the available luxuries.

I was ready to attack France around 250 AD or so. I immediately made a huge mistake. I thought I could cancel a DoF whenever I wanted, but when I found I couldn't, I declared war anyway. France fell soon enough, but Spain denounced me immediately, and soon after I found myself at war with Spain, Russia and Siam as well. Fortunately my army was pretty devastating, and I ground my way through all four of them, finally taking all of the capitals on my continent by turn 260. But I could have accomplished this much faster, and with more gold from peace treaties and trading, if I hadn't screwed up that DoF and been at war with the entire continent for pretty much the duration. The focus on science early on definitely paid off, as I was never behind in military tech, and ahead almost always - sometimes by two levels.

By then I had infantry and artillery, but had long since quit building units in order to get the late-game Wonders and raise my culture, as well as build the air force and nukes I knew I'd need to invade the other continent, where Germany had just conquered Songhai and England. My policies now also included Scholasticism and most of Autocracy and Commerce.

I arrived in time to see that Germany had nuked the Songhai capital. One key observation I quickly made is that the AI doesn't seem to nuke cities with Military Bases. I built one instantly in every city I took, because as usual I had more gold than I knew what to do with: 30K at game's end, despite buying GDR's and buildings on the new continent at every opportunity. I also nuked the AI cities with uranium deposits, then set about my path toward far-off London and Berlin.

Germany attacked with artillery and then rocket artillery. I could ignore these as a rule, and almost exclusively used nukes, Stealth bombers and a rare melee unit to take the necessary cities. Despite my still getting the hang of the most efficient way to do this, I won by turn 311. (I also had 4.5 trees completed, and over half the world's cities.)

 

[Thalassicus]

I'm glad policy rate wasn't a problem for you in that game; 4.5 trees is a reasonable number and about what I normally get.

Sneaks added DoF-cancellation back when DoFs never expired. When Firaxis made them expire, I removed the cancellation option. It makes the decision of who to befriend more strategically challenging.

It's good to hear Germany actually conquered two civs in your game. Most of my games end around Infantry and it's very rare to see any civs destroyed by then.

 

[Txurce]

I'm glad policy rate wasn't a problem for you in that game; 4.5 trees is a reasonable number and about what I normally get.

Sneaks added DoF-cancellation back when DoFs never expired. When Firaxis made them expire, I removed the cancellation option. It makes the decision of who to befriend more strategically challenging.

It's good to hear Germany actually conquered two civs in your game. Most of my games end around Infantry and it's very rare to see any civs destroyed by then.

Yes, the policy rate seemed about right.

Now that I know DoFs's don't expire (!), I'm okay with the change you made, although I'd prefer to have the option. Removing the option does add a choice, but in non-war games, indiscriminate Friendships lead to late-game denunciations and diplomatic red marks. In the early war game, the science benefit is low; in the late game, no one's your Friend anymore, anyway.

I've also noticed the occasional AI civ being conquered early on, too. I think it's quite possible for the AI to win a conquest game on a Pangaea map (which I have yet to play).

 

[Ahriman]

When Firaxis made them expire, I removed the cancellation option. It makes the decision of who to befriend more strategically challenging

I support this. Otherwise it's a no-brainer to sign everything.

indiscriminate Friendships lead to late-game denunciations and diplomatic red marks

I see this as a pro, not a con. Late game needs some war to stir it up. I find TBC to be too peaceful overall, unless I act as a warmonger. AIs seem to invade and conquer each other less than in vanilla.

 

[Txurce]

I support this. Otherwise it's a no-brainer to sign everything.


I see this as a pro, not a con. Late game needs some war to stir it up. I find TBC to be too peaceful overall, unless I act as a warmonger. AIs seem to invade and conquer each other less than in vanilla.

I see it as a pro as well, but mentioned it as one reason why signing DoF's isn't a no-brainer.

The TBC AI may be calmer than the vanilla one, but my games are all pretty different. In my last one everyone attacked me, and Germany conquered his continent. In the prior one everyone around me was fighting almost from the start. It's when I play a peaceful strat that I find it pretty easy to avoid war. While I would probably prefer the AI to be a little more aggressive, I much prefer the relative consistency and "rationality."

 

[Thalassicus]

Late game needs some war to stir it up. I find TBC to be too peaceful overall, unless I act as a warmonger. AIs seem to invade and conquer each other less than in vanilla.

It's not for a lack of trying, that's for sure.

The AIs build huge armies, but often just sit there not attacking anyone. I set their war flavors high... in the end the amount we can do with data edits is limited. The reason they don't conquer one another as much are the extra promotions they receive. It makes combat more challenging, which is good in human vs ai battles but does tend to stalemate ai vs ai. It's difficult for two AIs to really have a lopsided victory since they are essentially the same person with the same tactics and strategy.

 

[Txurce]

It's not for a lack of trying, that's for sure.

The AIs build huge armies, but often just sit there not attacking anyone. I set their war flavors high... in the end the amount we can do with data edits is limited. The reason they don't conquer one another as much are the extra promotions they receive. It makes combat more challenging, which is good in human vs ai battles but does tend to stalemate ai vs ai. It's difficult for two AIs to really have a lopsided victory since they are essentially the same person with the same tactics and strategy.

Yes - that's why the big breakthroughs tend to come when an AI researches a game-changing tech like Flight or the bomb. But there are also some civs that seem to do worse than others. For example, the Mongols are the absolute pits in my games. The Aztecs may be next, with India and Songhai and Egypt perhaps in the next tier. The Polynesians and Danes have yet to have a strong finish, although in less tries.

 

[Thalassicus]

There's some leaders who can never do well as an AI. Mongolia relies on mobility and avoiding counterattacks, and the AI can't predict threats well enough to do this. Danes require amphibious attacks and pillaging, neither of which the AI does. Polynesia depends on knowing a chain of Moai built in a row are more powerful than the individual improvements.

This is one reason why Washington typically does well as an AI. His bonuses fit well within the AIs limitations: faster growth, higher sight range, lower purchase costs.

 

[Txurce]

Mongolia can never do well as an AI because it relies on mobility and avoiding direct confrontations.

And he's aggressive to boot. A deadly combination.