THE STAR

[CheeseOnToast]

Cheese, was it a standard RM Arabia, or a custom Arabia script?

It was Dry Arabia. it's defined correctly in the RMS as a Arabia style map.

[Armelon]

Welcome back mister Ruler! gogogog

[THE RULER]

I've uploaded a test version plays only a skirm(arch) flush any civ
atm needs direct access to enemies erg Arabia like
civ characteristics atm not supported
still not able to treat mounted opponents (in work)

a huge rework of the villager training
new style of balancing (units,research, villies)
now dynamically handles target selection (2v2)
can calculate the loss of units as a percentage (deceased ratio) during a raid and triggers retreat
can also calculate a transient between loss and production of units and triggers retreat if quotient is going downwards
still tries to place a tower direction to closest enemy (this is of deciding importance), while TCs will be placed in direction ally (team together)
(this may fail if ES can't find a placement with uncalculated result )
calculation of a floating average (trend) to set gathering percentage as usual for my scripts (similar to simple share-charts)

the attack itself is shurely unique..

I still need to pimp up the darkage, the scouting is not satisfactory at all
the tower with the defense position is still static, will use dynamic positioning next time

[Esty]

Hey

I tested your Ai vs mine twice, in 1v1, with byzantine as his favorite civ, on arabia map.

First game :
Your Ai feudaled arround 11:30 and made immediately one defensive tower in front of your 2 archeries and made your full army arround your garisonned tower. After having standed defensive the whole feudal age, you uped castle minute 25 without being hurt thank to your defense and started being agressive. You just rapped my ai with your massive attack and you uped elite-skirmishers fastly. I liked the way your economy was developped, 2 more tc were built fastly. You eventually uped imperial minute 40 when my ai uped only minute 50 after having suffered a lot. Your imperial bugged and your ai kept spamming something like "seems my ai needs update".; then, it kept patrolling without attack anymore, near my town. My Ai was dead for a long time btw lol. I was really impressed overal by this first game !

second game :
This time your ai uped feudal minute 12 and didn't make a tower. Your archeries were build with only one villager assigned for this. It took time to have army enough to counter my flushers. Your ai said he was gonna make a 2nd camp on stone after having made one on gold, but made it on gold again It enventually tried to make a tower in front of your archeries but i had already massed like 30 ranged units there. Your ai built 4 camps (3 on gold and finaly one on stone...) I reached castle min 25 and won.

Conclusion : This version of your ai can be really impressive with a good feudal defense at this step of the development of ur ai. Some bugs need to be fixed at least but this is the beggining of your work. Your scout dont care about boars at all and lure deers very quickly, then go explore enemy. In both games your 2 boars stayed alive the whole game ! Maybe you made a AI able to face a barbarian boars-stealer ?

Your AI is really promissing ! i wait for the next update

[ER]

Just one small thing I noticed when browsing through the code. There is a potential problem with one defconst:

(defconst medium 1)

"medium" is a reserved word in the game for checking map-size. By overwriting it you can potentially get strange/unexpected results when checking map-size in a defrule, for eg.

[THE RULER]

txs very much for testing
sry that this version spams lots of debug messages
forgot about that "enemy cleaning up" loop , forgot to fix
often ES can't distinguish between gold and stone - result is as you described ( too low camp max dist and stone closer than gold)
my script autom enlarges camp max dist; therefore it places some "test" camps to get info about the camps to gold dropdistance.
indeed that's not so much effective, but uses quite small code and runs for any map. this code was already quite old and before UP
esp the wood camp placements urgently needs a better progressing ,that, for I am shure, eats a lots of rules.
it needs to find the enemy before placing to tower correctly - lost war on drunken scout, yep, that's life !

ah yeees, no attack rule for late game yet, needs some (more) though attempts

txs ER, should not interfer, but better to redefine this constant

if you face one of the high class Ais you indeed may not miss a sheep or boar resulting in your early doom.

[ER]

There is a way to drop camps correctly at gold or stone. Working from an admittedly faulty memory it was something like:

To drop at gold use actions -

(set-strategic-number sn-gold-dropsite-distance "small-distance")
(set-strategic-number sn-stone-dropsite-distance "large-distance")

To drop at stone use actions -

(set-strategic-number sn-gold-dropsite-distance "large-distance")
(set-strategic-number sn-stone-dropsite-distance "small-distance")

Either way, the camp-max-distance must be large enough to reach that specific resource for it to work properly.

[THE RULER]

a small update available

next thoughts: will favour units if camps are ahead, enemy close , and ev town discovered resp. enemy not found
another proceeding after 35 min

less suicidal now

[John the Late]

Ooh a new AI! Very nice, thanks for uploading!

[marathon]

WB Ruler!

Tested your AI, it is interesting already. I want to study your dynamic gathering.
Remember to lure 1st boar either when you still have sheep, or after you have like 5 farms at least.
Something was wrong with a Mayan player: made no army, limited camp distance etc.
Since you store coordinates of allies, if pocket in a 4-strong team it is more critical to find closest flank rather than to find closest ally.

Happy scripting!