A Better AI.

[Blake]

First of all, thank you for this lengthy information. It is always nice to see a bit of how the game really works.
...

Hammers get multiplied by things like forges, food does not. But I guess that is taken into account when building workers and settlers.
I personally value commerce a bit higher than 1/3 of a hammer, more like 1/2 of a hammer.

...

These values must have taken quite some thinking and experimentation.

No chance , they are the defaults the game come with (or close to) and I haven't changed them, in order to maintain familiarity. Not that the thinking and experimentation hasn't been done...

The valuation function I use tends to be:
f = 1.0, h = 0.60, c = 0.4

(comparison) Emp All Value:
f = 1.0, h = 0.52, c = 0.25

(comparison) Emp none Value:
f = ???, h = 1.0, c = 0.33


My value rates 3c = 2h, or that a Scientist and Engineer specialist are roughly equal. it does depend on context, a "growth enhancing build" (worker, settler, workboat, wonder etc) tends to increase the value of hammers (relative to commerce) by somewhere around +100% (ie 3c = 1h). So the 3c = 2h weighting is for when cities are basically idling.

If I ever get around to writing a new governor it will use the new valuation with more context sensitive weighting

For me there is no connection at all between the slider and the settings that I want in one specific city. The one is a global thing, the other a local. If I locally have a bigger multiplyer on gold/science (from library/market), then I would rather have a merchant/scientist as it would lead to a bigger total output. If I emphasize commerce, I would like it to value science as much as gold. There's really no difference between the two as one can exchange one for the other by changing the science/tax slider.

I'm in complete agreement. I've just argued this point long enough that I was too tired to be bothered arguing it yet again... although I do still think that 100% settings should act as tie breakers... especially for cases of 100% science w/ positive income and 100% gold w/ negative income...

The governor doesn't consider specialists in its optimization code for the output of the city until you click the emphasize great person button. The emphasize great person button wouldn't increase the value of specialists in this optimization algorithm. It would just count them as the amount of resources that they provide (scientist = 3 science). Without the button pressed, specialists have a value of 0.

....

The logic behind this is that you can now choose in each city whether you want specialist and thus great persons or not. 'Emphasize great person' should in this case be renamed to 'allow specialists'.

I'll consider doing this because I'm not sure if anyone actually uses "Emphasize Great People" since forcing specialists tends to get the desired results.

But it is a radical departure from the familiar behavior of the emphasize buttons so I wouldn't say it's likely...

Also unfortunately for you I personally think great people pollution is a relatively minor problem in the grand scheme of things - a functional largely hands off governor is more important.

Ideally I'd like an "Avoid Specialists" button since cities where you have to take care with pool purity tend to greatly be the exception rather than the norm.

 

[Quagga]

Oddities with 1-8 build:

As mentioned by others, Scouts get the blue circle, often pointing out into the dark.

I noticed a curious lack of cultural defenses in AI cities. I captured a size 10 city from Alex using Trebs and I didn't even need to bombard. That seems impossible.

At one point I had a worker showing an incredible number of turns to build something. I think it probably meant -1.

I had an AI offer me a trade, their gold for my tech. When I said let's negotiate, he wouldn't give me as much gold. I know this used to happen in CivIII, but I don't remember ever getting a worse deal doing this in CivIV.

I had Ragnar declare a long distance war on me. It was pathetic though.

Alex declared war on me, but then waited about four turns to attack. Maybe he was bribed.

 

[uberfish]

Lol, thanks for listening Blake. I'll take another shot at deity this weekend with the latest version, I went back to unmodded warlords after the awful war we had with Shaka in acid09.

For those unfamiliar with deity, there are already enough random game losses caused by AI war declarations at the wrong time. With the mass units built by the new AI the players needed to be cut a bit of a break on the military front. By the way, it's not so easy to just camp in your own territory and kill AI attack stacks on deity to reduce WW, they can keep up a constant stream of units to pin you down and use their home territory advantage well to counterattack you when you've taken a city or two while they tech up. This results in heavy WW.

It's not a case of wanting to win more easily - Just that 'more AI production' + improved combat AI already makes wars considerably harder and reduces the need for the frustrating game mechanics that were put in to defend the original AI against being overrun too easily. I can guarantee it's still going to be considerably more challenging than 1.61 given the same difficulty level.

 

[Iustus]

BTW I can't see huts graphics

You cannot see goody huts? At all? I do not have this problem. Are you sure it is not the issue of the hut drawing for a second when it was there, but some AI beat you to it before you revealed that plot?

also I can see the blue circles with scouts in the shroud ??

With the new, vastly improved explorer code, I went ahead and added blue circles to the suggested spot to explore, this is where an automated explorer would go (randomness aside). It only shows the blue circles for scouts, explorers and sea based units.

It very often will be on a unrevealed plot, because what is the point of scouting plots that you can already see?

It is pretty trivial to remove these blue circles, I added them for testing, and then decided to leave them in.

By the way, if you want to test out the new circumnavigation code, generate any water based map, use worldbuilder to give yourself two caravels on the same plot, and automate explore on both of them. I am very happy with how well it worked out. Note, if you generate these sea units in a normal game, it may take more than 2 units, each unit has a 50/50 chance of deciding to go west or east (and north or south for maps that wrap that way) but once it decides, it will always be that direction for that unit, so it is quite possible to build two 'eastgoing' caravels if you are building other units as well (if they are created without any units in between you are virtually guaranteed to have one east and one west).

As you can see in this example, the two caravels started at the red dot, and they moved just about as well as you could expect, to meet where the blue arrow is pointing.

If you want the circumnavigation bonus, you will have to get to optics more quickly, no counting on the AI to piddle around and not try for it.

-Iustus

 

[Newcomer24]

With the new, vastly improved explorer code, I went ahead and added blue circles to the suggested spot to explore, this is where an automated explorer would go (randomness aside). It only shows the blue circles for scouts, explorers and sea based units.

It very often will be on a unrevealed plot, because what is the point of scouting plots that you can already see?

It is pretty trivial to remove these blue circles, I added them for testing, and then decided to leave them in.

-Iustus

I understand why you added the circles.. they can be very helpful for testing... I do not like to see them while playing though (as I have no need for them) and would like them to be removed from the public build.

 

[Iustus]

I understand why you added the circles.. they can be very helpful for testing... I do not like to see them while playing though (as I have no need for them) and would like them to be removed from the public build.

They seemed to be in the spirit of other suggestions. If you turn off suggestions, they will not show, that said, they will be removed in the next build. It seems no one liked the explore suggestions.

Explorer automated behavior will still be improved, you just will not get any explore blue circle suggestions.

-Iustus

 

[Doomed_UK]

yes, I concur - I like the idea of the new explorer code, but I would prefer it without the blue circles.
(The blue circles for settlers annoy me enough!)

I hope the level of declarations of war in the 8/1 build is the same as the 1/1 build - I have been having great fun with the 1/1 build - it even has made me drop back to noble after losing 3 time in a row (twice to military and once to culture)

 

[sveint]

Sweet insanity!

I think the screenshot says it all... this is the 8 Jan build/Monarch/new handicaps. Great work, I got massacred.

 

[klokwerk]

51 units, not bad

Did Monty manage to take your other cities ??

I think it's a good thing to see such attacks, my dream is to see an AI win by domination or conquest.

 

[sveint]

That screenshot is taken after I tried to fight him off - I lost about 20 units. And he has well over 100 units left in the attack stacks.

But even more incredible is Ragnar, his power is 50% higher than Montezuma!

As for the game, I have it saved at the point of the screenshot, but it's over. I have no reserves except for a single unit in each city. Too much teching (I'm the tech leader for a change, usually not on Monarch), not enough units - all my inital neighbours were low in power.

 

[Zombie69]

- Wrote new cottage-valuing logic for governor (town > village > hamlet > cottage).

This is completely wrong, it should be the other way around. Personally, if i can choose between working a 3 commerce riverside cottage (financial) and a 3 commerce hamlet, i'll choose the cottage, especially if it will take fewer turns for it to grow to the next level. This is one of the reasons why i hate the new governor and think the original one that came with the game is miles ahead.

Another example (since Roland asked me for specifics) is that the emphasize buttons now emphasize too much for my taste. Just because i said to emphasize production, doesn't mean i'd rather have 1 hammer than 1 food (but it does mean that i'd rather have 2 hammers than 1 food, whereas normally i wouldn't). Just an example.

 

[sveint]

I think it's a good thing to see such attacks, my dream is to see an AI win by domination or conquest.

For the record, Napoleon won a Domination victory in one of my games on an early build. I was stuck on an isolated island and he vassalized all of the other AIs.

 

[marioflag]

This is an issue i have noticed since warlords was released and it hasn't been yet "fixed".
When AI's research polyteism they prioritize the construction of Temple of Arthemis.What is more useful between Temple of Arthemis and Parthenon can be subjective and vary from people to people, but building always first Temple of Arthemis and only after Parthenon isn't good.
Philosophical leaders should prioritize Parthenon instead of Temple of Arthemis considering tht their trait stacks with effect of Parthenon.Creative leaders also should get a slight preference for Parthenon.Financial leaders instead should prioritize Temple of Arthemis.I don't use worldbuilder but you can easily check this seeing that Temple of Arthemis with The Oracle and the Great Wall is built really soon, while Parthenon is built a lot later.So you can easily try to grab both temples building first T. of Arthemis and after the Parthenon, while it's nearly impossible if you try to build Parthenon first and then T. of Arthemis.
In the last 10 games i have tried to build 5 times first Arthemis T. and i was able to build both temples everytime, while the other 5 times i built first Parthenon and then Arthemis t. and i was able just one time to build both (i had a really good production capital).
I hope that something can be done about it.

 

[marioflag]

I forgot to add that in my last game with 1/1 build i noticed a strange behaviour.Hannibal asked me for the right of passage which i agreed and the turn after he declared war on me. While on the declaration of war troops inside my territory were sent out of my territory there wasn't any declaration in which he refused to continue with right of passage agreement.I was a bit weak militarily anyway he desembarked just an archer and a swordsman on my coast.Is this a bug?

 

[Blake]

This is completely wrong, it should be the other way around. Personally, if i can choose between working a 3 commerce riverside cottage (financial) and a 3 commerce hamlet, i'll choose the cottage, especially if it will take fewer turns for it to grow to the next level. This is one of the reasons why i hate the new governor and think the original one that came with the game is miles ahead.

Actually the governor does prefer a river cottage over a non-river hamlet - it looks at the yield and how long it will take to upgrade, both have identical yields but the cottage upgrades faster so is considered the better tile.

Another example (since Roland asked me for specifics) is that the emphasize buttons now emphasize too much for my taste. Just because i said to emphasize production, doesn't mean i'd rather have 1 hammer than 1 food (but it does mean that i'd rather have 2 hammers than 1 food, whereas normally i wouldn't). Just an example.

Now? I'm in agreement but I haven't change the way emphasis buttons work since I want to preserve familiarity.

The fact is the emp buttons value food at 180 and non-emp production is 15 meaning 1f = 12h. I agree that's stupid.
Please explain the "Now"? Has the behaivour changed or are you assuming what I was talking about was how I've changed it to work and not how it's always worked?

 

[Roland Johansen]

No chance , they are the defaults the game come with (or close to) and I haven't changed them, in order to maintain familiarity. Not that the thinking and experimentation hasn't been done...

The valuation function I use tends to be:
f = 1.0, h = 0.60, c = 0.4

(comparison) Emp All Value:
f = 1.0, h = 0.52, c = 0.25

(comparison) Emp none Value:
f = ???, h = 1.0, c = 0.33


My value rates 3c = 2h, or that a Scientist and Engineer specialist are roughly equal. it does depend on context, a "growth enhancing build" (worker, settler, workboat, wonder etc) tends to increase the value of hammers (relative to commerce) by somewhere around +100% (ie 3c = 1h). So the 3c = 2h weighting is for when cities are basically idling.

If I ever get around to writing a new governor it will use the new valuation with more context sensitive weighting

That would be great. My personal valuation is pretty close to yours, maybe almost identical, so I would love this new governor that still resides only in your mind. I hope it's not too difficult to add some time in the future.

I'm in complete agreement. I've just argued this point long enough that I was too tired to be bothered arguing it yet again... although I do still think that 100% settings should act as tie breakers... especially for cases of 100% science w/ positive income and 100% gold w/ negative income...

I agree with the breaking points that you added with 100% science with positive income and 100% gold with negative income, but I never get in those situations.

I'll consider doing this because I'm not sure if anyone actually uses "Emphasize Great People" since forcing specialists tends to get the desired results.

But it is a radical departure from the familiar behavior of the emphasize buttons so I wouldn't say it's likely...

Also unfortunately for you I personally think great people pollution is a relatively minor problem in the grand scheme of things - a functional largely hands off governor is more important.

Ideally I'd like an "Avoid Specialists" button since cities where you have to take care with pool purity tend to greatly be the exception rather than the norm.

I know the idea was rather radical. I wasn't even sure that you'd consider it. It doesn't actually have to do that much with great people pollution. It's just that I sometimes want to try to get the right combination of tiles used without specialists (who will generally be less productive than improved tiles) and the normal non-emphasizing governor picks the wrong tiles. Maybe an improvement on the standard governor like you suggested above would work for me.

Lol, thanks for listening Blake. I'll take another shot at deity this weekend with the latest version, I went back to unmodded warlords after the awful war we had with Shaka in acid09.

For those unfamiliar with deity, there are already enough random game losses caused by AI war declarations at the wrong time. With the mass units built by the new AI the players needed to be cut a bit of a break on the military front. By the way, it's not so easy to just camp in your own territory and kill AI attack stacks on deity to reduce WW, they can keep up a constant stream of units to pin you down and use their home territory advantage well to counterattack you when you've taken a city or two while they tech up. This results in heavy WW.

It's not a case of wanting to win more easily - Just that 'more AI production' + improved combat AI already makes wars considerably harder and reduces the need for the frustrating game mechanics that were put in to defend the original AI against being overrun too easily. I can guarantee it's still going to be considerably more challenging than 1.61 given the same difficulty level.

Just to warn you, the latest build doesn't contain a modified handicap file so it is still using the old handicaps. Other handicaps also wouldn't change your war weariness but only the AI's war weariness (at least that would be the result of the changes that were proposed by Blake). I guess that Blake is considering other handicap values for the 1.0 build and not for these intermediate builds.

 

[Wodan]

Does the AI take pity on humans?

Take a look.

I had to rally all units down south to fight off first Ragnar and then Shaka. "All hands on deck" type of situation. Lo, to my surprise, Wang shows up in the NW with 3 galleons worth of troops.

Exactly what a human would have done... wait in the wings until others declare war, and then swoop in.

But, my point is that Wang did NOT take the ungarrisoned city. WTH? Not that I'm complaining, but this is the Better AI mod, after all. He should have taken my city with butter and lemon, maybe a spritz of fresh pepper. All he needed was a napkin to clean up.

Wodan

 

[Blake]

Oh are you talking about him not unloading the troops directly into the city and capturing it from the sea? I guess the AI ignores that possibility at the moment...

 

[Uncle_Joe]

One observation on why the AI sometimes seems so 'peaceful':

The AI is VERY good at spreading its religions. In all 3 of the games I've started with 1/8, the other continent was all one unified religion, making for a very peaceful coexistance. I've also noticed that many of the AIs are reluctant to switch to Free Religion and thus break-up their little sewing circle.

I dont know what could be done (or even if its desirable for anything to be done), but I think that the AIs should consider Free Religion a little more as the game goes on.

 

[Mannu]

Oh are you talking about him not unloading the troops directly into the city and capturing it from the sea? I guess the AI ignores that possibility at the moment...

I've seen this referred to as the "Syrian Doctrine". The human player certainly has this in his bag of tricks, so IMO it should be an option for the AI as well. Basically the thinking goes that it protects your units from counter attack (except by other ships attacking your transports) on the turns when you are next to the city but not attacking (end of the first turn, or bombardment turns).

Considerations:
- Amphibous promotion of attack troops
- Naval superiority (at the target city and overall)
- defensible terrain next to target city
- counter attack units in the target city

Purposes
- establish a beachhead on another landmass
- capture and raze poorly defended far flung cities for long distance wars?

I'd love to see Ragnar capable of pulling this off...