-Aims to get paladin asap and launch 50 paladins at the enemy. Usually does this between 40-43 min if left untouched or is not hurt by any attacks towards it.
-best times to get 50 paladins so far are:
* Persians: 40.12
* Huns: 40.19
* Franks: 40.39
-Will Team Krush an enemy in team games - is very strong in teams.
-Civ: Best with:
*Franks - Full Paladins, slight eco bonus's for good boom, best paladins.
*Huns - Full Paladins, good boom, stable training bonus for faster paladins.
*Persians - Full Paladins, best boom, paladins get archer attack bonus.
*Teutons - Full Paladins, cheap farms for good boom, BBT's for awesome defence late game.
*Spanish - Full Paladins, few bonus's that help boom, BBT's for awesome defence late game.
*can play with any other civ that gets paladin.
-Map type - Land maps (Black forest, Arabia, Mongolia, Yucatan, Gold rush etc.)
-Hard difficulty only
THIS AI IS FOR USERPATCH ONLY!!
To download, go to: http://www.aiscripters.com/forums - you may need to register
I have more ai's over at aiscripters as well.
Check back for updates often if you like this AI
|Author||Comments & Reviews ( All | Comments Only | Reviews Only )|
This AI is stronger in closed and wallable maps (as any AI prepared for Fast Imperial strategies), but it works well in any land map because if it detects a possible early attack it will prepare knights faster.
And as Cakemaphoeneige says, it has got a great team cooperation, making it very dangerous in 2v2 or more games.
I take my hat off to this incredible team game AI
Too bad it's not compatible with 1.0c
Thanks for the nice comments guys.
If enough people want Pantheon 1.0c compatible i might be able to make it happen.
- It is ideal for 4vs4 games. It only shows its power in such big games, where it is impossible to kill enough Pantheons before they overwhelm your teammates.
- If you lose all your AI allies while 3-4 Pantheons are still alive, you can no longer win. At least I can't, unless I have already giant stockpiles and Elephants.
- If your AI allies do not manage to stop its first coordinated attack, at 26-29', before it kills an AI ally, then the 2nd AI ally will fall even faster.
- Its attack is predictable: it attacks the last player number in your team, so if you are the only human you will be the last to be attacked... unless if your town lies on the road to the target town.
- If 2 humans try to cooperate with 1-2 AIs for cannon fodder, it is very possible to kill it with a sling-shot strategy.
(at least the version I tested)
- Some annoying creepy taunting when 2 Pantheons are weak and under attack. Needs fix.
- Forgets caravan, critical when you have 50+ carts per player.
- Some suicidal miners and builders, especially mining camp builders.
- It is of remarkable strength only in 4vs4.
- Ignores bodkin arrow for its TCs, bombard towers and castles.
- Ignores the opportunity to sling a very weakened but still alive ally. Often a player is left with 2 (undiscovered by the enemy) town centers and 0 villagers. A small sling for a start could revive this player very fast.
- Ignores the opportunity to strategically relocate when under severe pressure. Maybe in a future version? Strategically placed castles, TCs, markets would make it brrr.
- I somehow dislike how it deletes villagers for carts later. This plan and execution needs optimisation.
- Its known issue of attacking the last player number makes it too predictable for some people's taste.
Try it on maps with lots of gold. Maps like Relics and Resources, with lots of safe gold.
Thanks for the review marathon.
Some of those issues, particularly the targeting issue, have been fixed for a while now, i just havent updated it yet, so thats my fault.
Some of those issues i wasnt really aware of too, so ill look into it when i get the chance. I havent been doing much scripting lately.
Ill try and get stuck into some scripting soon and update pantheon.
Can't play in Scandinavia Map that lack of berries.
However in other land map, this AI is a strong AI.
Only few non-cheating AI can stand against Pantheon at the first attack.