Wodan
Deity
I see... not happening to me, in any event. Anyway, I'm sure they'll appreciate the screen shot to help track it down.
Wodan
A Better AI.
- Thread starter Blake
- Start date
Wodan
Wodan
Deity
Iustus/Blake,
Something still not quite right with the latest build AI settlement placement. Ragnar was on an island with just enough room for one more city. He places that city so that it has 2 overlap with the capitol and thereby misses out on a coastal tile with clams.
The overlap tiles are the only hills available, so perhaps it's valuing the hammers over the clams.
About the only value there I see is the tactic of two cities sharing the tiles, swapping back and forth at need. Is your AI clever enough to do that? Even so, I'm not sure it's worth missing the clams and the two extra workable tiles.
Wodan
Something still not quite right with the latest build AI settlement placement. Ragnar was on an island with just enough room for one more city. He places that city so that it has 2 overlap with the capitol and thereby misses out on a coastal tile with clams.
The overlap tiles are the only hills available, so perhaps it's valuing the hammers over the clams.
About the only value there I see is the tactic of two cities sharing the tiles, swapping back and forth at need. Is your AI clever enough to do that? Even so, I'm not sure it's worth missing the clams and the two extra workable tiles.
Wodan
I did not play you AI, but my impression form 2.08 pach and notes and discussion here is that some of AI "Improvement" come from the price of AI character and influence of some Game elements, like religion and trade.
That narrowed game, making it more poor, less diverce in posible strategies.
Game gearer more and mor eto just strigth tech feast and warmondering. Diplomacy and AI character component shrink.
I would suggest to rethink direction of you mod and concentrate on improving AI city and army managment and military strategy and logistic.
Stop reducing richness of the game, stop making all characters absolutly a like.
That narrowed game, making it more poor, less diverce in posible strategies.
Game gearer more and mor eto just strigth tech feast and warmondering. Diplomacy and AI character component shrink.
I would suggest to rethink direction of you mod and concentrate on improving AI city and army managment and military strategy and logistic.
Stop reducing richness of the game, stop making all characters absolutly a like.
analyst
Chieftain
- Joined
- Oct 23, 2005
- Messages
- 43
Mutineer, if you had bothered to read even 10 of the 823 posts in this thread, you would have seen that an overwhelming portion of the mod is devoted to military strategy. The 'dagger' approach has made fighting wars exponentially more challenging.
The AI personalities remain; Monty continues to warmonger, and Mansa will trade like crazy. The difference is that they do what they do better.
The last 5 pages or so have been a lot of discussion over AI city placement and city management, and Blakes mod has bettered both.
Read the thread (at least partially) before you post (please...)
The AI personalities remain; Monty continues to warmonger, and Mansa will trade like crazy. The difference is that they do what they do better.
The last 5 pages or so have been a lot of discussion over AI city placement and city management, and Blakes mod has bettered both.
Read the thread (at least partially) before you post (please...)
yes, but effect of this was reduced, as one think which were effecting diplomacy, technology trade, now have mach less effect on diplomatic situation.
This cause that masta no longer create friends try his trade and Toky does not create enamies true his refusin of trade. This tendancy continue.
Now, creater start to modify starting positions with out undestanding what part of start position generator does.
In part it give civ starting position a character. If you create a few starting position in Vanilla you will find that different civ tend to have different starting terran and position on the map.
This cause that masta no longer create friends try his trade and Toky does not create enamies true his refusin of trade. This tendancy continue.
Now, creater start to modify starting positions with out undestanding what part of start position generator does.
In part it give civ starting position a character. If you create a few starting position in Vanilla you will find that different civ tend to have different starting terran and position on the map.
kettyo
Seeker of Reality
Diplomacy is set in different files so it's totally unchanged.
Trading techs will give you the trade attitude boosts the very same way as always had.
There are no civ-specific starting locations and never been.
Just look at the civilizationinfos and leaderheadsinfos files.
No values affecting starting at all.
Actually Blake's mod makes the game more diverse because now AI civs pursue all the 3 kinds of victory type, military, technology and culture victory.
Previously AI only pursued technology victory.
So rethink about this a little bit please.
Though i accept if someone likes the stock AI better than the Blake/Iustus one. So probably it's better to make the 'BetterAI' optional and not integrate into the official patch but it's up to Firaxis i think.
Jettio, you simple do not know or even read pach notes. Try trade yourself and compate with original game.
Simular trade in original game tend to bring about 2 times bigger + for fair trade then in 2.08.
Starting positon is effected by civ, just make experiment I asked.
For example,
Issa, Saladin tend to start with gold near by.
Hatty allmost never has gold near starting position.
I do not know what effect it, I know that there is nothing in XML files.
It could be hardcoded or linked to starting tech or some other parameters.
Simular trade in original game tend to bring about 2 times bigger + for fair trade then in 2.08.
Starting positon is effected by civ, just make experiment I asked.
For example,
Issa, Saladin tend to start with gold near by.
Hatty allmost never has gold near starting position.
I do not know what effect it, I know that there is nothing in XML files.
It could be hardcoded or linked to starting tech or some other parameters.
tofof
Chieftain
Installed the mod and am loving it. A few issues so far though -
Emphasize great person was choosing in general to stagnate the city's growth - even in cities that were at like, 14/21 health 12/23 happiness. In one extreme example the city wanted to run 6 priests rather than 5 + a single worker on any of the grassland tiles. Running 'food + gp' emphasis ended up doing 0 specalists at all. It seemed like there should be a middle ground. (I have a screenshot of this on a different computer if needed).
I always play teamed with my financee, so we're quite used to how the AI treats us independently for some things and together for others. Religion, close-border-tension, help/refused help, traded with enemies are all individual. However, with the ai mod installed, I (she and I both, actually) was getting a 'close borders spark tension' for any civ SHE had close borders with, despite the fact I was on a different continent and nowhere near. See screenshot (I'm china, spain is angry about my borders, they're an ocean apart before astronomy. On the minimap you can see the very-very close borders that Spain and India, my teammate, share.)
Next, they love to chop during wars now, which is great. However, there seemed to be a bug in worker safety issues. See screenshot. Two turns prior to the shot, she moved the workers to C and started chopping. On the prior turn, I moved my jungle-movement axeman to point A, where I also had an elephant - he's now at B, having killed THEIR elephant just prior to this screenshot. At any rate, Catherine could absolutely see the forest-axeman, had the opportunity to move them to safety / put a defender on them / put her elephant between me and them but didn't. Result: three dead workers.
Finally, through the whole war she seemed unaware of the possiblity of exploiting the land bridge to strike behind my lines. (That's a barbarian on it in the screenshot). Instead, she continually threw individual units (4-5 across maybe 12 turns) from St. Petersburg to harass my forces taking Rostov, Yakutsk, and Novgorod. If she'd sent even just three of them together on that landbridge, she could have hit lightly defended stuff and might have actually forced a peace treaty, as I was spread thin even on the front lines.
Oh, one last afterthought - the AI is always so willing to trade 7-8 gold per turn for even the crappiest resources (e.g. fish for a mostly-inland civ with no harbors). With skillful negotiations I'm often able to bring in as much as 10/turn from 6-8 civs in a 15 civ game, which is huge for research purposes - I can run at 100% and make a profit, even while supporting many cities or a large army. This was the only way I even had an economically viable empire back when I used to play at monarch or higher - but it's as abusable now as it was before.
Maybe they shouldn't be so willing to part with cash? I'll even wait to see their gold/turn go up a few points, cancel an old outrageous deal with them, and renegotiate it to take all their cash again. Then they'll turn down research (it seems), rinse repeat. For strategic resources or desperately-needed resources it makes sense to be willing to pay through the nose, but this is one where I always seem to be able to abuse their coding.
Emphasize great person was choosing in general to stagnate the city's growth - even in cities that were at like, 14/21 health 12/23 happiness. In one extreme example the city wanted to run 6 priests rather than 5 + a single worker on any of the grassland tiles. Running 'food + gp' emphasis ended up doing 0 specalists at all. It seemed like there should be a middle ground. (I have a screenshot of this on a different computer if needed).
I always play teamed with my financee, so we're quite used to how the AI treats us independently for some things and together for others. Religion, close-border-tension, help/refused help, traded with enemies are all individual. However, with the ai mod installed, I (she and I both, actually) was getting a 'close borders spark tension' for any civ SHE had close borders with, despite the fact I was on a different continent and nowhere near. See screenshot (I'm china, spain is angry about my borders, they're an ocean apart before astronomy. On the minimap you can see the very-very close borders that Spain and India, my teammate, share.)
Next, they love to chop during wars now, which is great. However, there seemed to be a bug in worker safety issues. See screenshot. Two turns prior to the shot, she moved the workers to C and started chopping. On the prior turn, I moved my jungle-movement axeman to point A, where I also had an elephant - he's now at B, having killed THEIR elephant just prior to this screenshot. At any rate, Catherine could absolutely see the forest-axeman, had the opportunity to move them to safety / put a defender on them / put her elephant between me and them but didn't. Result: three dead workers.
Finally, through the whole war she seemed unaware of the possiblity of exploiting the land bridge to strike behind my lines. (That's a barbarian on it in the screenshot). Instead, she continually threw individual units (4-5 across maybe 12 turns) from St. Petersburg to harass my forces taking Rostov, Yakutsk, and Novgorod. If she'd sent even just three of them together on that landbridge, she could have hit lightly defended stuff and might have actually forced a peace treaty, as I was spread thin even on the front lines.
Oh, one last afterthought - the AI is always so willing to trade 7-8 gold per turn for even the crappiest resources (e.g. fish for a mostly-inland civ with no harbors). With skillful negotiations I'm often able to bring in as much as 10/turn from 6-8 civs in a 15 civ game, which is huge for research purposes - I can run at 100% and make a profit, even while supporting many cities or a large army. This was the only way I even had an economically viable empire back when I used to play at monarch or higher - but it's as abusable now as it was before.
Maybe they shouldn't be so willing to part with cash? I'll even wait to see their gold/turn go up a few points, cancel an old outrageous deal with them, and renegotiate it to take all their cash again. Then they'll turn down research (it seems), rinse repeat. For strategic resources or desperately-needed resources it makes sense to be willing to pay through the nose, but this is one where I always seem to be able to abuse their coding.
Attachments
Iustus
King
No intentional changes were made to reduce the distinctiveness of each personality.
Pursuit of religions or not, should be if anything more diverse than in the default game.
The one area that was somewhat changed is that AIs are less willing to give away all of their money in trades. Now, you can argue that this is an improvement in their intelligence, in that they now realize that that money can be used to run at a higher research rate. Or you can argue, as I think you are, that this change makes it harder for the human player to tech trade, ruining the tech trading aspect of Civ4. I do not feel that it goes too far on this point, but I am certainly willing to discuss it.
If there is some other specific thing that the AIs are doing or not doing, I would love to look into it, but it is all but impossible to respond to general statements like "the AIs have less personality than in 2.08".
There are some things that are going to be less random. AIs are much more likely to be building a granary pretty early in the process, rather than building whatever random building they feel like and never getting around to building a granary. (Or some other essential building).
On the personality side, Better AI civs like their unique buildings and unique units and will favor them and the techs that enable them more than in the default game.
Better AI civs use the warmongering variables stored for each leader to make more decisions, so the personality of each leader can have an even bigger effect on decisions such as when to go to war, when to accept peace, how much to value peace, and so on.
Hope this clarifies things. I would encourage you to try out a build and give some specific comments about issues that you see. Include a save game if possible, hopefully the turn before something interesting happens, so we can look into it directly. (I suggest changing your autosave interval to 1 turn, to make this easy to do, you can increase the number of saves at the same time, so you can still go back many turns if needed).
-Iustus
Iustus
King
Yes, this is really a tricky situation. I will see if I can get it to work a bit better.
-Iustus
tofof
Chieftain
Great. Is there anything I can do to help with this? I have save files for many of the situations I described, I haven't sent them to that gmail account Blake made because I wasn't sure which if any would be useful.
Wodan
Deity
I ran into similar things a few times, not with GP emphasis, but with other buttons (hammer, commerce, research).
My intuitive feeling is that the emphasize food button should make the city grow always, no matter what other buttons are checked.
Wodan
Iustus
King
I think the issue is emphasize other things should not stop the city from trying to at least grow 2/turn if there is growth to be done.
-Iustus
darrelljs
Immortal
I don't have enough experience playing with the patch to know if Mutineer's concerns are valid, but just because the diplomacy code is unchanged doesn't mean that diplomacy is unchanged.
Darrell
Wodan
Deity
I agree with what you just said. But, that doesn't change that Mutineer made some very general, sweeping condemnations without much in specific to warrant and validate his position.
I thought that when I first saw it. It felt more like a rant to me than any kind of critical analysis. Actually, I thought Iustus did a pretty good job in replying to him. I think it's clear that Mutineer needs to provide more in the way of specifics at this point.
Wodan
Wodan
Deity
Agreed. Actually, that's true even if there is no emphasis whatsoever. The city should grow by default... the player should be able to force stagnation but it should be just that: forced. Anyway, IMO.
Wodan
Iustus
King
Without emphasis, it does just that. But when you emphasize something, right now it is drowning out all other concerns. It is a tricky thing to balance, if you make the emphasize do too little, then it sticks with food even when thats not what you want. If you make it do too much, then it stagnates the city.
I am probably going to just have to add an extra bonus to force that first 2/turn.
-Iustus
player1 fanatic
Fanatic
Emp. shuld keep city growth while weighting emp. thing more then other.
Only if you add prevent growth too in the mix, it should go to super emp. mode.
P.S.
That how I used it in unmodified Civ4.
Only if you add prevent growth too in the mix, it should go to super emp. mode.
P.S.
That how I used it in unmodified Civ4.
DevilJin
Warlord
Blake and Iustus.
I played with the latest build and I saw that Even when the AI is going to capture a City, and he brings tons of trebuchets(35 in my case) he doesn't attack with lot of them. and after the turn ends he keeps the trebuchets near the city so I caould easily destroy them.
I played with the latest build and I saw that Even when the AI is going to capture a City, and he brings tons of trebuchets(35 in my case) he doesn't attack with lot of them. and after the turn ends he keeps the trebuchets near the city so I caould easily destroy them.
During the last game I played I observed interesting AI behavior.
It came with a stack (catapults & swordsmen) to my city that was defended by a single axeman. I took my poor axeman anr run leaving city undefended. To my big surprise AI instead ot taking the city started taking down its defenses. It was standing with this massive army by and empty city until its defenses were down to 0...
I guess the city attack routine should be changed to something like this:
1. If city is empty - enter city with stack;
2. If city has any defence and there are catapults (trebs, etc) in the stack - take down the city defences?
3. Cycle through all the stack units (except for those dedicated to defending the attackers and those taking down city defences). Pick the one with highest success probability. If that probability is above 50% attack.
If probability is below 50% choose one of two strategies: collateral damage strategy (if there are several strong units) or take-out the strong guy strategy (if there is one or few strong units). In the first case start the attack with catapults etc. In the latter case take down the strongest guy with best attacker and then start with collateral damage attack.
It came with a stack (catapults & swordsmen) to my city that was defended by a single axeman. I took my poor axeman anr run leaving city undefended. To my big surprise AI instead ot taking the city started taking down its defenses. It was standing with this massive army by and empty city until its defenses were down to 0...
I guess the city attack routine should be changed to something like this:
1. If city is empty - enter city with stack;
2. If city has any defence and there are catapults (trebs, etc) in the stack - take down the city defences?
3. Cycle through all the stack units (except for those dedicated to defending the attackers and those taking down city defences). Pick the one with highest success probability. If that probability is above 50% attack.
If probability is below 50% choose one of two strategies: collateral damage strategy (if there are several strong units) or take-out the strong guy strategy (if there is one or few strong units). In the first case start the attack with catapults etc. In the latter case take down the strongest guy with best attacker and then start with collateral damage attack.
