Denmark

[BiteInTheMark]

Here’s the difference between me and you: you see something, but you don’t know why it is. You observe, but you cannot interpret. You are like the man who witnesses advanced science and calls it divine: you see the event, but you lack the language to explain it. If you stopped at observation, that would be fine. But you persist in your blind explanations of the scientific, and in the process reveal your ignorance.

I am always thankful for observation and reporting. I love to answer questions and be challenged on topics or ideas. What irritates me is when people without access or insight into the code decide to use a bug or a quirk to explain away the thousands of man-hours spent making an AI that thinks and acts as close to a human player as we are capable of achieving on a single threaded, 32bit program.

G

If an architect builds a house for me, which is completly instable, and I already see cracks in the walls, stones falling from the ceiling, I dont need to know why this building is instable, I dont need the ability to calculate static of a house. Its enough to see a house which is already falling apart to know, the done work was a failure. If you call me ignorant because I do not know the exact cause of a problem, I call you arrogant because you suggest to me that I am unable to properly observe and draw logical conclusions. Not knowing the exact causes of a problem does not absolve the problem from any error. But that's exactly what you're trying to claim.

Denmark AI does, for whatever reason, not play like a human would play. Iam not talking about deity pro player ways to play, with survival 3 scouts sneak attacks or lightning warfare. I talk about the elementary understanding of the worth of pillaging. Every calculation of the AI may be correct, but if the base values are chosen wrong, the result is wrong, even if the calculation was done properly.
As an example, the AI is much more hostile towards others, if they have atleast one of their UU. The value for hostile aproach to others is multiplied by 10. Why 10? Why not 5? Or 20? The hostile approach towards others is multiplied by 10 independently if this civ is only able to have 1 of their UU ever or already have 20. Even multiplied vs nations which are much stronger, or those with a defence pact. Theres no technical bug, but the result will be irrational war declaration cause a human has set this value to 10.

Would you consider everything is fine, if a 2 city backwater civ declares every time war it can against a 20 city leading empire? Would you? You wouldnt need any insight to the code to know, something is going wrong.

 

[Gazebo]

If an architect builds a house for me, which is completly instable, and I already see cracks in the walls, stones falling from the ceiling, I dont need to know why this building is instable, I dont need the ability to calculate static of a house. Its enough to see a house which is already falling apart to know, the done work was a failure. If you call me ignorant because I do not know the exact cause of a problem, I call you arrogant because you suggest to me that I am unable to properly observe and draw logical conclusions. Not knowing the exact causes of a problem does not absolve the problem from any error. But that's exactly what you're trying to claim.

Denmark AI does, for whatever reason, not play like a human would play. Iam not talking about deity pro player ways to play, with survival 3 scouts sneak attacks or lightning warfare. I talk about the elementary understanding of the worth of pillaging. Every calculation of the AI may be correct, but if the base values are chosen wrong, the result is wrong, even if the calculation was done properly.
As an example, the AI is much more hostile towards others, if they have atleast one of their UU. The value for hostile aproach to others is multiplied by 10. Why 10? Why not 5? Or 20? The hostile approach towards others is multiplied by 10 independently if this civ is only able to have 1 of their UU ever or already have 20. Even multiplied vs nations which are much stronger, or those with a defence pact. Theres no technical bug, but the result will be irrational war declaration cause a human has set this value to 10.

Would you consider everything is fine, if a 2 city backwater civ declares every time war it can against a 20 city leading empire? Would you? You wouldnt need any insight to the code to know, something is going wrong.

The funniest part is that you literally have no idea what you’re talking about w/r/t how the approach code works. You’re just parroting what other people wrote about code they didn't write.

It’s not arrogance to be skilled at something and get angry when someone without said skill makes up nonsense to puff up their own ego. That’s called expertise.

I spend far too much time humoring your bad faith behavior.

G

 

[BiteInTheMark]

The funniest part is that you literally have no idea what you’re talking about w/r/t how the approach code works. You’re just parroting what other people wrote about code they didn't write.

It’s not arrogance to be skilled at something and get angry when someone without said skill makes up nonsense to puff up their own ego. That’s called expertise.

I spend far too much time humoring your bad faith behavior.

G

I..... dont... care... about.... the.... code....
It isnt necessary to know the code. I dunno why you are only talking about the damn code.
But I care about your irrational claim, Denmark AI is working proberly as intended. Everyone can see, they are not playing like a human with such an UA would. I have no clue why you claim such thing. Their simply repeating the standard warmonger procedere. I didnt see any difference in playstyle to any other warmonger civ, played by AI. They come to conquer, always. I never saw any sneak or fast attack to plunder as much as possible and then retreat. Never. Not in 2 years.

If you master, divine might of the mod, have the full insight of all behavior of the AI, please enlighten us, why the AI didnt pillaged the little ...... fishing boat in @infidel88 example, even it would give plenty of very worthy culture to them and has no danger to get killed or even damaged?

 

[Gazebo]

I..... dont... care... about.... the.... code....
It isnt necessary to know the code. I dunno why you are only talking about the damn code.
But I care about your irrational claim, Denmark AI is working proberly as intended. Everyone can see, they are not playing like a human with such an UA would. I have no clue why you claim such thing. Their simply repeating the standard warmonger procedere. I didnt see any difference in playstyle to any other warmonger civ, played by AI. They come to conquer, always. I never saw any sneak or fast attack to plunder as much as possible and then retreat. Never. Not in 2 years.

If you master, divine might of the mod, have the full insight of all behavior of the AI, please enlighten us, why the AI didnt pillaged the little ...... fishing boat in @infidel88 example, even it would give plenty of very worthy culture to them and has no danger to get killed or even damaged?

And therein lies the problem. The only thing you care about is being right. On the internet! Such glory...

I already said why. But you don't seem to read.

G

 

[CrazyG]

I'm actually inclined to agree here.

Humans use rewards for pillaging a lot better than the AI does. You can say this about many things, like rewards for unit kills or city capture. However, the gap is biggest by far for pillaging.

Even from just a human perspective, there are many mechanics that are fun and interesting. Killing as many units as possible, demanding as much tribute as possible, city conquest rewards, these are present an interesting puzzle to solve, if you wish to maximize them. Maximum pillaging does not. It has a very boring solution. It gives you incentive not to take cities (the AI does not do this well). It gives you incentive not to capture enemy workers (AI does not do this well). This is why I get so bored when I play Denmark. You just need to attack a loser civ once every 15 or so turns to pillage it all again. You don't even accumulate much warmonger because you don't take many cities.

The move to yields for pillaging was a mistake in my opinion. The reasons for it seemed to be that people finally acknoledged how broken Runestones were when they rewarded killing units. Which didn't address the problem at all, it just made the AI worse.

 

[BiteInTheMark]

And therein lies the problem. The only thing you care about is being right. On the internet! Such glory...

I already said why. But you don't seem to read.

G

"The AI doesn’t know what you know or see what you see. The AI is an algorithm. It views things like tiles as a probability model. Likely it viewed the plot as too risky without proper visiblity."

Yeah, the AI is an algorithm. Its math. And not a beeing which can "understand" things like we human can do. Its not able to "understand" its UA. In the end its only a mathematical comparison of 2 numbers.
And if your unable to create a code for the AI to gain a greater view of everything and evaluate values and risks better (no blame on you, thats high level AI coding, nothing for a single privat person), you shouldnt give an AI civ such an ability.

I agree with CrazyG. To get a maximum out of the UA of Denmark, the AI would need to forget nearly everything about city conquest. Else its only a maritime mobile warmonger nation with some yields by war.
Maybe its better to completly throw away the whole pillaging thing on Denmark, cause its impossible to balance it for both, human and AI in same way.

 

[Gazebo]

And not a beeing which can "understand" things like we human can do. Its not able to "understand" its UA. In the end its only a mathematical comparison of 2 numbers.

There's the jump to a conclusion, and I disagree with it. The AI can understand it's UA without being human. It doesn't have to play 100% like a human to understand what UA it has. That's where I disagree, and I'm not going to budge on that point.

And if your unable to create a code for the AI to gain a greater view of everything and evaluate values and risks better (no blame on you, thats high level AI coding, nothing for a single privat person), you shouldnt give an AI civ such an ability.

This is such a back-handed, asinine thing to say. Absolutely unacceptable. You should be ashamed of yourself for having such a miserable attitude towards this project, and towards me, a person who has spent far too much time engaging with you and far too much time on this project. I hate that you derail every conversation and make it a personal attack on me. I hate that you constantly barb and prod and poke at me, to try and provoke a response. I hate, ultimately, that you constantly and consistently operate in bad faith to egg me on. I'm tired of it. Contribute constructively and positively like everyone else, or - and I'm being serious here - I won't be sad to see you go populate another subforum.

The move to yields for pillaging was a mistake in my opinion. The reasons for it seemed to be that people finally acknoledged how broken Runestones were when they rewarded killing units. Which didn't address the problem at all, it just made the AI worse.

Huh?

Also, I disagree with the point behind this. Humans will do everything better than an AI. So we cut it all? Piece by piece? Strip it down so that each UA is just a passive bonus? No, that's a terrible idea.

This is a profoundly disappointing thread, I thought better of all of you.
G

 

[civplayer33]

This is a profoundly disappointing thread, I thought better of all of you.
G

Now, now; some of us still appreciate your efforts
Also, our future robot overlords will surely reward you for attempting to bring sentience and enlightenment to the dank, dark, unconscious hell of single threaded 32 bit programs

 

[ashendashin]

Huh?

Also, I disagree with the point behind this. Humans will do everything better than an AI. So we cut it all? Piece by piece? Strip it down so that each UA is just a passive bonus? No, that's a terrible idea.

This is a profoundly disappointing thread, I thought better of all of you.
G

On the topic of 'passive' bonuses... I think auto-pillaging after killing a unit would be really cool. Just throwing that out there. please don't hate me.

 

[Gazebo]

Now, now; some of us still appreciate your efforts
Also, our future robot overlords will surely reward you for attempting to bring sentience and enlightenment to the dank, dark, unconscious hell of single threaded 32 bit programs

Oh definitely. I'm not going anywhere. I've been around too long for all that. I just get a bit fatigued with every.single.thread. getting derailed by the same two or three forum users, and with the same topics and themes every time. It gets old, no? Or perhaps it's just me getting old...

On the topic of 'passive' bonuses... I think auto-pillaging after killing a unit would be really cool. Just throwing that out there. please don't hate me.

Your picture is appropriate here.

G

 

[Gazebo]

That's certainly not what we want. I like the idea that the AI tries to save their units. However, it does seem like the humans can use that to their advantage but that's a different topic altogether.

What's the order the AI makes its decisions for its units? Let's say we have a melee non-mounted unit pillage at the start of the turn and have 2 movements still due to Viking. Then, it withdraws to heal up and fight another day. Will another unit now notice the pillaged tile and account for a boost to its attack? I know it can be hard but it would be cool if a wounded unit can heal and withdraw while setting the stage for another unit to use that pillaged tile to continue the push. I haven't seen quite something like this in my games but maybe I just haven't played enough games.

Been looking into this, and I've found a few places (don't kill me @ilteroi) where I think I can make the AI a little more aggressive at pillaging. We'll see. It's a very delicate balance.

G

 

[Gazebo]

Just dawned on me as I was looking at the code - I've got the solution: I'm going to 'Dojo' the Runestone, make it so that the pillage bonuses from a unit apply to the city the unit is from. We can actually bump the values up to appreciable levels that way. I'll do the same for the Colosseum/Acropolis. No reason to have those out-of-control values when we can easily control them with directed yields.

G

 

[civplayer33]

Just dawned on me as I was looking at the code - I've got the solution: I'm going to 'Dojo' the Runestone, make it so that the pillage bonuses from a unit apply to the city the unit is from. We can actually bump the values up to appreciable levels that way. I'll do the same for the Colosseum/Acropolis. No reason to have those out-of-control values when we can easily control them with directed yields.

G

And then reenable pillaging of anything and everything, I presume? I like it.
Hmm is it possible to change unit names at all? Would be nice to be able to mark units' origin cities so that I know which city will get the production boost on the next pillage.

Edit: apparently it is possible, but only when the unit is about to be promoted; there is a little Edit button in the top right of the promotion window.

 

[Gazebo]

And then reenable pillaging of anything and everything, I presume? I like it.
Hmm is it possible to change unit names at all? Would be nice to be able to mark unit's origin cities so that I know which city will get the production boost on the next pillage.

Yeah, and then remove the restrictions. I'd probably keep the land/sea tables separate for API purposes, but they could be standardized.

 

[CrazyG]

Just dawned on me as I was looking at the code - I've got the solution: I'm going to 'Dojo' the Runestone, make it so that the pillage bonuses from a unit apply to the city the unit is from. We can actually bump the values up to appreciable levels that way. I'll do the same for the Colosseum/Acropolis. No reason to have those out-of-control values when we can easily control them with directed yields.

G

Beautiful solution.

 

[Gazebo]

Beautiful solution.

I'm thinking 25p/25c to start, no? Since it's only going to one city?

G

 

[Deadstarre]

make it so that the pillage bonuses from a unit apply to the city the unit is from

G

the building_yieldfrompillage table worked as intended perfectly for a very long time. then you changed how it worked, split it into land and sea tables and for resource only. now its changing again, and the yieldfromkills table with it... its fine that you have new ideas for what tables should be but please bear in mind other people are using these long-time CP tables as well, and there is no need to scrap perfectly useful and functional tables just to trade apples for oranges. the table as it currently works has its merits, and as you want it to work has its merits- theyre just different. if you want to make a different table that the yields will now go only to the city unit is from, i'd appreciate you make that a new table rather than scrap the useful long existing ones.

 

[Gazebo]

the building_yieldfrompillage table worked as intended perfectly for a very long time. then you changed how it worked, split it into land and sea tables and for resource only. now its changing again, and the yieldfromkills table with it... its fine that you have new ideas for what tables should be but please bear in mind other people are using these long-time CP tables as well, and there is no need to scrap perfectly useful and functional tables just to trade apples for oranges. the table as it currently works has its merits, and as you want it to work has its merits- theyre just different. if you want to make a different table that the yields will now go only to the city unit is from, i'd appreciate you make that a new table rather than scrap the useful long existing ones.

Until VP goes 'gold,' I consider any of the expanded API elements like new tables to be subject to some changes/tweaks for the purposes of promoting balance, even outside of the VP. If someone is using these tables specifically for their 'global' nature, we can address that, but I make new tables very rarely these days for memory purposes.

G

 

[Deadstarre]

Until VP goes 'gold,' I consider any of the expanded API elements like new tables to be subject to some changes/tweaks for the purposes of promoting balance, even outside of the VP. If someone is using these tables specifically for their 'global' nature, we can address that, but I make new tables very rarely these days for memory purposes.

G

i get that, i didnt say anything when you replaced the authority unit-per-citizen table and there have been other apple to orange table changes in the not too distant past i remain silent about as well including the resource only pillaging because frankly i dont consider them worth posting about. but please lets not bring up 'gold'. If a table exists in the CP for X amount of years unchanged, and works perfectly and is being used, it would be nice to let that table be- a modder somewhere is likely already making use of it as it is.

and yes, in this case i am specifically using the yieldfromvictory / pillage tables for the global nature, as they currently function, and im perfectly happy with these apples. replacing them would be an unwelcome bucket of oranges. do with that information as you will =)

if tables must be cut hopefully it wont be something as basic as the way these building tables have been operating, and having multiple ways of handling super common actions like killing/pillaging are fine. Personally i think there is more eclectic chaff in the coreadditions file that could be cut if needed (how often am i gonna yieldfromspydefense, and comes to mind i remember you acquiesced to a huge chunk of trait additions to another modder who'd already done the work not so long ago... and frankly some of those tables looked dubious) nor do i think the land/sea pillage distinction is worth memory either, but i know its all subjective. Subjective as it is, hopefully table seniority counts for something.

 

[BiteInTheMark]

This is such a back-handed, asinine thing to say. Absolutely unacceptable. You should be ashamed of yourself for having such a miserable attitude towards this project, and towards me, a person who has spent far too much time engaging with you and far too much time on this project. I hate that you derail every conversation and make it a personal attack on me. I hate that you constantly barb and prod and poke at me, to try and provoke a response. I hate, ultimately, that you constantly and consistently operate in bad faith to egg me on. I'm tired of it. Contribute constructively and positively like everyone else, or - and I'm being serious here - I won't be sad to see you go populate another subforum.

Oh, Iam sorry for this, it was late yesterday and I didnt wanted to sound like this.
I wanted to say, that I didnt believe the AI will be ever able to reach an acceptable level of understanding of this UA. You pushed the AI to such a better understanding than any Firaxis coder have done, and Iam thankful for it. The AI may have reached already 100% of normal human skill level in building order, policies and is even better in calculating yield gain.
Their ability for diplomacy may be 85%, and the skill level of going to war maybe 75% (dont kill me about those numbers ), but the proper understanding (and conclusion with subsequent action) may have only reached 30% of the skill of a normal human (numbers are only here to show the discrepancy).
I dont think its possible to balance such a civ for human and AI, cause the yields a human can gain are not only a bit higher (10-20%) by abusing some little mechanics, the discrepancy can be extremly high (5 times? 10 times?). Balancing Denmark based on the yields a human can gain, would create a civ, which has a much too weak UA for AI hands. My comment about "AI excells at civs with lot of passiv skills" was aiming this. Its not possible to abuse Chinas UA, the AI is already good at accumulating GW to trigger and is also able to conquer cities, even shes not as good as a human. The mentioned civs didnt have UA which can be abused by human the AI cant. This makes them much easier to balance. But civs like Denmark, Venice or Aztecs are so special in their way to play optimal, making it a huge difference if a AI or a human is playing.

Also, I disagree with the point behind this. Humans will do everything better than an AI. So we cut it all? Piece by piece? Strip it down so that each UA is just a passive bonus? No, that's a terrible idea.

Of course they can do better than AI. But what counts, is the possible difference. If a human is 10-30% better than an AI, we will find a balanced spot. But if a human can get 10 times more yields/benefits out of a UA, it will be a nightmare to find a fair balance.