Age of Kings Heaven
Downloads Home
Best Files
Review Guidelines
Main site
Code of Conduct

Advanced Search
AI Files
New Releases
New Reviews
New Comments

Submit File
Browse Categories
New Files
Recent Updates

Downloads Home » AI Files » KeyC AI beta 1.1

Download File Post Review Post Comment

KeyC AI beta 1.1

Author File Description
File Details
Version: The Conquerors 1.0c
The KeyC AI beta 1.1...!
My first AI.
It's not finished, but it's well underway :)

Random map
Arabia - scandinavia works too
200 pop
Dark age
Standard/low resources

**Updated 18. May 2011**
It researches all technologies worth researching.
Now trains a lot of knights, partly instead of LC.
Trains trebuchets.
It trades with allies.

Advice/suggestions/whatever is always welcome! But make sure to read the 'things to do' list at the top of the script :)
AuthorComments & Reviews   ( All | Comments Only | Reviews Only )
Campidoctoris Very good script to be a first AI.

Your 19-20' of castle time is not bad, and pumping scouts during castle transition is a good idea IMO. It reduces your time to be vulnerable to enemy flushes and they are upgraded automatically into light cav once in castle.

Things that I consider you should improve:

- Once you are in castle age you should change into your also bonused knights. With your gold mining bonus they are "cheaper", although they drain mines as fast like any other civ, but the main reason is because they are the most powerful castle unit. Of course you can turn again into light cav if low gold, monks or in late castle.

- Unbalanced eco with gold excess in castle time, although you could correct it better with knights.

- Lack of counter units. Your cavalry and swords are more than enough to protect your ranged units, and they are great protecting your cavalry from sou haven't apears, but you need something against cavalry. Yes, Turk spears are bad, but they are still useful, specially in feudal and early castle, and you can turn your cavalry into camels if enemy has tons of knights, etc.

-Flush detection, although this is more complicated, but even the fastest FC scripts need it if they want to become top AIs. If you detect a really fast feudal time to your enemy you could set your strategy into a feudal army to portect your eco against enemy flush, and then you can flush by your own or train only the necessary and jump faster to castle age.
File Author
Thanks! :)

I've tried to get some camels going when it detects cavalry (see g-enemy_cavalry), but it wont work for some reason.
I've always thought of camels as a replacement, the turks got from the bad spears, and therefor wiser to use camels, but maybe I'm wrong... It just works extremely well in the sassanid empire script. Those camels are a pain in the a** if there's enough of them.

I have scripted it to build more barracks - and therefor more swordsmen when the enemy builds spears against my scouts. That was the main purpose of them - to crush the spears.

Knights you say? That sounds like a good idea ;)

And a small question: how do you detect what age the enemy is in?
Campidoctoris Some scripts only counts the enemy unit that have to be countered, and if they are greater than an specific number they train the counter unit. Other scripts like mine (I think that CPS_Alexander does something similar, but I could be wrong about it) use goals or obsolete strategic numbers to determine the level of threat, the higher the threat the more counter units you train.

You can study my Threat Rules.per file, which determine the levels of threat. And the files that load my main file for Turks in the part in which Crusade select what units it will train.

For example, in the Castle Age my Turks prefer knights until transition to imperial (although scouts in the transition to castle because by then I can't train nothing better and because the ligh cav Turk bonus). But the more cavalry has my enemy the more camels I train instread of knights, except if enemy has monks, then I change to scouts again. If I have enough anticav camels or antimonks scout I return to standard knights.

Of course,every scripter can choose other number of enemy units to set the threat levels, or the desired number of counters to that threat level.

Or even choose another way. For example James I AI counts the enemy buildings of any type, and it doens't work bad.

Finally, there are scripts which doesn't train counters but are surprisingly strong. Strong_bow only trains archers, monks and siege and it's hugely difficult even with skirms and cavalry. And there are KLEW scripts really difficult to win even with man@arms, despite to sword are suposed to counter eagles.
Leif Ericson
To detect a players age, just use "players-current-age any-enemy" ;)

There are two main reasons why Turks could consider using spearmen over camels. First they don't cost gold, and if they are training knights, cavalry archers, or gunpowder units, they'll need their gold. Second, camels are trained at a stable, a building that trains two other types of units which the Turks will use heavily (light cavalry and knights). Spearmen don't conflict with the training of these units.

However, that being said, camels are probably the most cost-effective way to beat knights, so if the situation calls for a cavalry-countering army, camels are a good idea.
File Author
Thank you :)

I think I'm gonna work on a flush that will be activated if the enemy reaches feudal very early.
I'll also add spears instead of the camels (they didn't work anyway).
lws735 Very good AI,I test it with BOSS 1.6.

Post Review Post Comment

HGDL v0.8.0

Forum Username:


Create a new account
Forgot password?
Favorites: [Who?]0
Size:6.24 KB