Report Questionable Behavior

[kettyo]

For some reason AI especially Japan i think is neglecting a place to settle which i consider would be quite a good resource outpost.
1 sheep for food so not a totally crippled city and also 2 furs and 1 whale worth an outpost city in my calculations without doubt even more for someone so close as Japan.

 

[kettyo]

AI (Korea) is training a settler in an 1 pop 'city' Cheju while there's nowhere to settle.

 

[kettyo]

Coastal cities are almost undefended (1 defender or 2) which is not right IMO.
It's ok the AI is not that great at invasions right now but that's just too much gamble. And I can attack from the sea hard enough and can also amphibious ! so they'd better care for their people more. Not to mention i'm sure AI's invasion/amphibious skills will also be improved in the close future.

 

[kettyo]

I don't see unit overspam at noble* at all so if we gonna turn them down it has to be tested very throughly to not leave civs defenseless.
At least a fast-response defensive stack or 2 would be necessary to defend the otherwise quite open cities.

 

[kettyo]

I think barbarian intruders should be attacked by the defenders with more courage.
Maybe inner cities are considered under low-threat so much they won't train anti-pillagers. This should be dealt with somehow. (like fast response antipillagers)
Also this could maybe to avoid luring out defenders. In such a case if victory is sure and there are no enemies nearby defender could move out for a kill nonetheless.
Maybe barbarian cities hold out longer than they should. AI's might do some kind of limited military buildup to take barb cities considered worthy.
Because of the new handicaps (bigger threat) AI's should consider escorting settlers with more units not just 1.

All this with betterai handicaps 1.00

 

[Blake]

For some reason AI especially Japan i think is neglecting a place to settle which i consider would be quite a good resource outpost.
1 sheep for food so not a totally crippled city and also 2 furs and 1 whale worth an outpost city in my calculations without doubt even more for someone so close as Japan.

Do you have this save?

I'm looking into coastal defenses, especially post-astronomy.

 

[kettyo]

Do you have this save?

I'm looking into coastal defenses, especially post-astronomy.

All my recent observations are from this game.
I now have got astronomy. Probably others don't have it.
This might be a factor i didn't think of.

 

[Roland Johansen]

Coastal cities are almost undefended (1 defender or 2) which is not right IMO.
It's ok the AI is not that great at invasions right now but that's just too much gamble. And I can attack from the sea hard enough and can also amphibious ! so they'd better care for their people more. Not to mention i'm sure AI's invasion/amphibious skills will also be improved in the close future.

There seems to be a difference between the number of defenders on the left below the spy and the number of defenders shown vaguely visible next to the spy on the bottom of the screen. Is the information from 2 different spies?

 

[kettyo]

There seems to be a difference between the number of defenders on the left below the spy and the number of defenders shown vaguely visible next to the spy on the bottom of the screen. Is the information from 2 different spies?

Yes i think the icons show the stack beside the spy that's moving (i've placed spy to every AI cities to observe 'unit overspam' which i didn't find)

And the text list at left-bottom shows the city on the screen because the pointer was hovering over these cities when the shots were taken.

Sorry for a bit of confusion.

Always look at the text list.

 

[shadyforce]

(from bug reports thread)


Julius Caesar managed to get himself in quite a financial hole in this one. He is at 100% to gold, and still negative, and this is not the first turn he has been so. I am happy to see that too many units is not the cause for his financial hole.

It seems it lost Rome to Wang Kon, but took all his other cities and then vassalized him (Rome is Wang Kon's only city).

Being at war with Washington isnt helping JC, but it isnt hurting him that much, it looks like he just took Washington's capital, that gold influx probably helped some.

The real problem is he is paying high maint cost because his capital is now in a poor place, and he has not built any courthouses or even his UB market.

This may be a case where he has been using wars to finance himself, but that is causing him to fail to build enough buildings, keeping the war effort going.

-Iustus

Well, this was a weird game. What happened was that Wang Kon launched an attack against Ceasar, took Rome, and eventually Vassalised him. (Wang Kon and Ceasar were not adjacent civilisations, Wang Kon was near me, Ceasar was on the far side of the pangea.) A little while after that, Wang Kon then abruptly attacked me (and thus his vassal Ceasar had to declare too). After fending off the initial wave, I launched a viscous counter attack (as you do) against Wang Kon. Ceasar didn't get involved in any of the fighting because as I said, he was far away on the pangea.

My counter attack captured and razed all of Wang Kon's main cities, leaving him with only the captured Rome. This freed Ceasar as a vassal as Wang Kon's population had plummeted. Rome was meanwhile under trememdous cultural pressure from Ceasar's surrounding cities. And not too long after I ended the war, Wang Kon now became Ceasar's vassal as the slave became the master!

Ceasar also gained I think Shaka as his vassal a while later and ended up back on top in score, and quickly entered a war with Washington shortly after all this as well.

 

[Ishon]

Testing the newest build.

The German scout ignored a barb village, which enables me to take it in the next turn. I enclose two savegames + a screenshot.

Spoiler :

http://forums.civfanatics.com/uploads/89703/AutoSave_BC-3835.CivWarlordsSave
http://forums.civfanatics.com/uploads/89703/AutoSave_BC-3820.CivWarlordsSave

 

[Ralgar]

I'm looking into coastal defenses, especially post-astronomy.

How does the AI handle colonisation post-astronomy?

Terra, standard, monarch, non-aggressive, 01/25 build:



Saladin beat me to the circum-navigation (something around 1200 AD), so in frustration I didn't concentrate on the new world. But later in 1750 the barbarian continent is still pristine like a virgin! I conquered without effort the northern half of the continent (infantry vs barbarian macemen). Till 1900 no other Civ settles in the southern part, so i can grab that too.

The game itself was good. There were many wars and powershifts among Victoria vs Napoleon vs Isabella. Saladin managed for the most part to stay out of war and could keep up (almost) in tech with me, but was backstabbed by Isabella.

No defender spam (or unit spam in general), but the wars, especially those the AI lost to me, were costly for the other Civs.

 

[jkp1187]

How does the AI handle colonisation post-astronomy?

I wonder if this has something to do with the fact that Terra's "old world" is a sort of "mini-Pangaea", meaning that the AI civs tend to go into land warfare mode to the exclusion of interest in overseas colonization? (All the more so since you were steamrollering them!)

I noticed in my recent Tectonics map game, a variety of land masses were generated, and the AI not only colonized some, but also fought wars over those colonies, too. (The civs starting locations were spread out over two or three continents, but there were also one or two empty island continents worth settling; I didn't try to grab them myself b/c I was focused on a culture strategy....)

 

[Iustus]

Testing the newest build.

The German scout ignored a barb village, which enables me to take it in the next turn. I enclose two savegames + a screenshot.

Spoiler :

http://forums.civfanatics.com/uploads/89703/AutoSave_BC-3835.CivWarlordsSave
http://forums.civfanatics.com/uploads/89703/AutoSave_BC-3820.CivWarlordsSave

Hmm. This is an interesting case. I am not sure if it makes sense to 'fix' this or not.

Basically what happened here is that the AI decided where to move before he saw the hut. I know I have done the same thing on accident.

It is pretty easy to change it so that explore units only move one plot at a time, but that will slow down the turn times more, as scouts will make twice as many decisions.

If you did not take the hut that turn, he would definitely grab it the next one.

Anyone have any thoughts? Is it alright for the AI to make some 'mistakes' or should it act perfect every time?

-Iustus

 

[marioflag]

Hmm. This is an interesting case. I am not sure if it makes sense to 'fix' this or not.

Basically what happened here is that the AI decided where to move before he saw the hut. I know I have done the same thing on accident.

It is pretty easy to change it so that explore units only move one plot at a time, but that will slow down the turn times more, as scouts will make twice as many decisions.

If you did not take the hut that turn, he would definitely grab it the next one.

Anyone have any thoughts? Is it alright for the AI to make some 'mistakes' or should it act perfect every time?

-Iustus

if it slowdown times i think things should stay as they are now.Missing a hut the first time is not an huge mistake and the AI can compensate for its error a turn later.
I think this is also common among expert humans to use the scout moving 2 movement for time.
If the slowdown is unnoticeable why not improving this aspect otherwise, this "mistake" is not worth a slowdown

 

[jkp1187]

I vote we leave it as is.... It balances out all the errant keystrokes that I make during a game.

 

[Roland Johansen]

I vote for perfect movement. I never move 2 tiles at a time through unexplored territory. Far to risky. You never know where you're going.

Does this really make turn times that much longer? It's just one of the many things the AI does in its turn. Or is the movement aspect that important for turn times. It's also only while exploring unexplored terrain with 2 move explorers where this specific error can happen.

 

[Arlborn]

If the slowdown is about 1 second or more(I think you cant tell it, can you?) then nah, not big importance over slowdown...

 

[Wodan]

I vote for perfect movement. I never move 2 tiles at a time through unexplored territory. Far to risky. You never know where you're going.

Does this really make turn times that much longer? It's just one of the many things the AI does in its turn. Or is the movement aspect that important for turn times. It's also only while exploring unexplored terrain with 2 move explorers where this specific error can happen.

I would hazard to guess that this would also happen with any two move unit when dealing with fogged terrain. Such as pillaging horse units. They move two and get whomped by your pike. Or, they move one, see the pike, and then run away.

Wodan

 

[Ishon]

I also vote for perfect movement, as I also never move 2 tiles at one time while exploring - this could be costly if you have bad luck. If an AI scout moves 1 tile and then finds out it's facing a dangerous animal/barbarian, it should pull back to some 50%/75% defense tile and not commit suicide.

If Wodan is right and this is just a natural feature of units that are capable of moving more than 1 tile/turn, I think this becomes a very serious issue. But I don't know how much this would make the game work slower. If it's 10-25% slower, I think it's acceptable. If it's more than 50%, let's think twice.