Literally crushing hard AI in 3.0.7. Let me tell you how.

[AdvFlightWins]

So, I have in recent history compiled 3.0.7 for playing. For a long time I used 2.5 because that came within my package manager selection. Because of this, I am used to playing on what is considered 'classic' ruleset. In 2.5 it was default. Enjoying it.

So, I am crushing the AI on hard.

I am surprised too. I normally played on normal enemy (or easy if I had taken a big break and need to get familiar again) and just kinda chill and evolve my land till the Advanced Flight comes into play and I crush em. Almost always the AI would out tech me, at least one or two of the enemies. I just shore up and keep advancing until I bomb them out of existence. What is different now? Why am I absolutely crushing Hard AI today?

One thing.

One thing is all that matters to advancing tech tree fast. Not how well you develop your cities. Not how many special tiles you have available to those cities. Not how many trade routes you have established. Not how many allies you make. None of it. Only one thing gets you through the tree faster than AI.

Code: Select all

You have more cities than the AI does.

Hopefully a lot more. Shoot for 2-3x as many as they have, at least. Go for more. That is it. Simple right? Now, how did I figure this out? Because I was getting bored with the normal play so stared mixing it up. I did a world with 50 civs, normal. 25 civs, normal. I was looking for an optimal number of enemies to allow a good and challenging early game but still have longevity and survivability. I wanted to be forced to take over enemy cities via ground war and not just hold out until advanced flight comes into play. This is how I was able to come to this conclusion. My current hard AI game is 15 civs. Many don't have more than 2 starter cities still. The top civs have 6-10. I was able to secure a peninsula to develop all on my own by locking up a land bridge with a phalanx until I could drop a city in the path. So I was still highest starter city civ, even though a few others were closer.

When I was only playing 4-5 enemy AI, the world was too empty. I would go grab enough land to get 8-10 starter cities. Then I would develop those. I did this because that was all I really wanted to manage. I did not want to micromanage 20 cities, too much effort. But when you have 50 enemies all packed into the default world size, you have to move fast to ensure you are even able to drop your second settler onto a tile. If you get a third settler in, you are the biggest empire on the globe. No AI has 3 cities in a 50 enemy world, unless they got onto an island all by themselves. So I crushed the 50 AI world starting with only 3 starter cities. No AI had more than 2. I was able to out tech-tree all their 2 city civs with my 3. Once I took over 1 or two of their civs with pike-men and catapults, truly game over for AI.

Basically, the 4-5 enemy world is too wide open. I think many casual players don't want to micromanage the number of cities it would take to actually fill this open space. So they don't. Because of this, the AI does. AI don't care how many cities they manage. AI does not get stressed if there is too much going on. AI just algorithms through it all.

If you want a challenging early game, but have a chance to out tech tree the AI, increase the number of enemies. Find your borders fast, because AI will. Or if only 4-5 enemies, go big or go home. You'll have to expand way longer than you think you will need to. Fill up every tile with production, because the AI will try to.

End game is still the same regardless. Advanced Flight Wins.

[Mahler]

Very interesting. How many cities do you have usually? What government do you have, Monarchy or Republic (Democracy)?

[Jacew]

Crushing Freeciv AI is pretty easy, at least in 2.5 - 2.6 and the style of play I prefer (big maps, slow research, lots of nations).

How I play usually:

- Explore island/continent I'm on right away, and either assimilate any AI nations on it, or set up defenses on a chokepoint with mountains/hills/forest so I have ample room to expand through early game.

-Build at least two defensive troops on every city + some on good defensive points. Also build few attackers spread out and ready to be deployed, if there's suspectible threat around.

-Research for trireme (if room to expand is limited), or/then Monarchy as fast as possible.

-Never cease to expand, there's always settlers going out if there's a room to go.

-After Monarchy, get Knights, and use them with catapults (or Dragoons/cannons) to invade couple of cities quickly and get the war on enemy territory, if there's war.

-------

AI really suck with boats. They never ferry big invasion troops across water, so once you get the land routes secure, you're pretty safe. Not even Pirates and Barbarians bother you, if you keep ample defenders in cities.

AI does expand their settlers over water, but they don't put high preference for researching Map Making, so that slows them down a lot, when room to expand is limited. Similarly, they don't seem to care how much sticking with Despotism hurts their growth, so they don't put high value on researching Monarcy or Republic. Or target research for specific units, like Pikemen, if they're getting crushed by Knights.

Another issue is that AI always sticks with what they consider "best" defender based on cost/value. When it would be tactically better to have a garrisons of varied defenders (like archer, legion, pikemen or chariot, phalanx, phalanx etc..) to counter different strategies.

I'd like to hear these issues that make AI fumble are still relevant in 3.x? I'd hope it would be bit smarter, or there would be improvements coming?