New Version - 2.5 (July 14, 2022)

Status
Not open for further replies.
Are AI bombers broken this patch? I've gifted them city states and Greece whom I am at war is not using them, ever. Also thank you for this patch, its the stablest I've ever seen for VP!
 
Are AI bombers broken this patch? I've gifted them city states and Greece whom I am at war is not using them, ever. Also thank you for this patch, its the stablest I've ever seen for VP!
Yes, fixed for next release.
 
Here is a minor issue with the new block system.
Allied units does not prevent blocking as you see here with the CS units inside french territory that doesnt unblock my 2 unit blockade.
1660241365024.png
 
I have a small question. It seems I cant disband my units. Is it a bug or a feature?
 
I have a small question. It seems I cant disband my units. Is it a bug or a feature?
You can't disband units when they are near enemy units. Otherwise, you have a problem with your install.
 
You can't disband units when they are near enemy units. Otherwise, you have a problem with your install.
Im using the auto installer... And I disband them inside my territories with no enemy unit nearby. Also anyone having the AI tell you to move troops away even when all of your troops are inside your border?
 
Im using the auto installer... And I disband them inside my territories with no enemy unit nearby. Also anyone having the AI tell you to move troops away even when all of your troops are inside your border?
I've had no issues with disbanding troops.
As for the AI being pissy about borders, if your borders are next to the AI's, and you have units at the edge inside your borders (like any competent general with itchy conquest ideas would :D), then yeah, the AI will whine about it. Vassals to ironically, even with OB active. Seriously, your a vassal, my troops tromping through your fig plantations is necessary, get over it lol.
 
I've had no issues with disbanding troops.
As for the AI being pissy about borders, if your borders are next to the AI's, and you have units at the edge inside your borders (like any competent general with itchy conquest ideas would :D), then yeah, the AI will whine about it. Vassals to ironically, even with OB active. Seriously, your a vassal, my troops tromping through your fig plantations is necessary, get over it lol.
I was not even thats close to them 😥. I always make sure to stay 4 tiles away from their city.
 
I was not even thats close to them 😥. I always make sure to stay 4 tiles away from their city.
I watched it and it seemed unreasonable to me. There was a recent bug report on this issue.
 
I didn't really think I would care, note or learn much from the tips. But that they would be a handy thing anyway. Some of them might be a bit generic but so far they have at least not been wrong or misleading. Even with many thousands of hours played I still picked up at least one tip -- I didn't know you could assign all your votes in congress by just right clicking. Don't know why I never knew.

That said some of them have a bit of a weird phrasing. Not claiming they are wrong or that I'm correct but they seem odd when I read them. Not sure if this has to do with limited characters or something such.

Not quite sure where they are stored either in the mod so I could look them all over so, I didn't really invest much time in looking either. So these are mostly from what I have seen on screen.

Example;
"Tip: never let a settler or great person unguarded"

let => leave? or 'never let a settler or great person travel unguarded/protected.




Spy - gold siphon. It was fixed some time ago to once again steal gold. It siphons gpt for a number of turns and then give you a pay out. The payout is a lot less compared to what it was previously. So much so that it is not really worth running it. The gold you get is just so minuscule that it doesn't matter. While in the previous incarnation where gold was just created out of nothing, which I guess was kind of bad to, you had an incentive to save gold for the future or big upgrades etc. That is now gone, or so to speak the risk is to great.

I'm not entirely sure how the system works but from what I can tell it siphons for X turns and then it deducts the sum from you gold pile. The workaround is to just never have any gold stockpiled. Nothing to pay out. I don't know if the AI (or you) still get any gold if that happens. But at least you the player doesn't go into the negative if you don't have any gold to steal. So you can at the end of a siphon basically negate an entire spy cycle for the AI (or another player etc).

There is also the order in which things happen. If you just follow along you can see how your gold goes to zero if you are low and then it fills up again when it's your turn as you get your GPT, income from various events (births, borders etc). So it deducts first and your income comes afterwards.

In the very late stages of the game you might see amounts similar to what it was before, thousands of golds. Which quite possibly even more makes it less and less interesting to have any gold in your stockpile as it can then swing wildly and remote several turns worth of income.

After all most people, I assume, tend to always have a tall and wide capital so you'll make many hundreds or thousands of gold per turn there. If it's during a golden age or we love the X day it's increased even more so if you siphon during a golden age but have to pay out when it has ended it could be a very large swing in difference.
 
I should bring up that city-states are very inconsistent with building workers over a long time, potentially the whole game, and it's a recurring issue that comes and goes with new versions. I know that in some cases, CS workers can be stolen by barbarians and AI, and they could just be slow to replace them. This is not the issue, because the CS in question have always been completely peaceful with AI and maintained sizable militaries, yet none of them have even begun to build any terrain improvements. When I first noticed this In a game from last year, I saw this behavior with two CS. One was an original CS and did not build a worker for the entire game, the other was an AI city that flipped CS, so they already had improvements, some remaining pillaged. The new CS did eventually build a worker after a very long time and even used the worker in a newly captured city. I wrote it off as a rare bug, but I've continued to see it pop up, leading to my current game on this version, where four out of nine CS on my continent have never built a worker by the medieval era. Has anyone else experienced this?
 
The new city healing seems to be a bit over the top. Or that it heals a bit much, or recovers to much per turn. I kind of like it but at the same time it's probably more detrimental to the AI then it is the player. As it is now a city can basically shrug off almost any amount of damage if it's not blocked by some mean. Just having units in range to semi-block isn't really enough. There needs to be actual full blocking.

The old tactic used to be to get range-3 and then bombard things back to the stoneage from safety and in the end rush in and take the city. That just doesn't work anymore unless you have a lot of range-3 bombarding going on. One or two piece of siege unit is not enough.

As the player I can work around it but the AI seems to be really struggling with it. I have so far only seen the AI take cities that are small and weak and had not had the time to proper defence up. They manged to take some city-states, mostly due to that they can outlast them and kill all the units and then swarm them. Beyond that it doesn't seem to be happening. The AI is not committing to the attack or gets distracted and pulls out to early so the city starts to heal and then bounces back more or less instantly.

To put some minor numbers on it. This is just prior to the industrial era, or I have just entered it. Bombarding Rotterdam a city with 36 combat strength, 815 hp, with two range-3 field guns. I took a few hits before taking notes.

T1: 785 hp, before damage, artillery 1 and 2 does 78 and 75 dmg. So leaving Rotterdam with 631 hp after. It's a bit odd, there is some rounding shenanigans here cause that would be 154 dmg but I only did 153 dmg.

T2: The AI engages Defense and gets 50 combat strength. It healed 145 dmg so it just entirely shrugged off the previous artillery (or field gun) barrage. Or it took 8 points of damage to be correct. That is not much. At this pace it would take a 100 turns give or take to bombard it down with two siege units. Auch!

T3: I guess I better take some shots with my musketmen to then. Rotterdam 776dmg, two field guns, three musketmen later. 612 hp. Heals 159 dmg. Back to 771 hitpoints, so this turn it took 5 points of damage then. This isn't going to work out.

T4: Lets move some melee units a bit closer; range2. Bombard. 616 hp left. We are now standing one tile out, so city-empty-myunit. By the next turn it healed back up to 659 hp. So at least now it took 43 points of damage. So it should be done in less then 15 turns at least.

The problem with moving a lot closer is that you are really standing in the deathzone if you don't have all the promotions, some medic units etc. In some regard the spy mission to put a city in blockade just became a lot more useful, and probably a lot more important. It's probably the best and most secure way of taking a city now. Cause just bombarding it won't do the trick anymore. Unless you can shuffle a lot of units in and out (probably to their death) the old and previously normal way just isn't going to cut it anymore in any kind of normal time frame.
 
The new city healing seems to be a bit over the top. Or that it heals a bit much, or recovers to much per turn. I kind of like it but at the same time it's probably more detrimental to the AI then it is the player. As it is now a city can basically shrug off almost any amount of damage if it's not blocked by some mean. Just having units in range to semi-block isn't really enough. There needs to be actual full blocking.

The old tactic used to be to get range-3 and then bombard things back to the stoneage from safety and in the end rush in and take the city. That just doesn't work anymore unless you have a lot of range-3 bombarding going on. One or two piece of siege unit is not enough.

As the player I can work around it but the AI seems to be really struggling with it. I have so far only seen the AI take cities that are small and weak and had not had the time to proper defence up. They manged to take some city-states, mostly due to that they can outlast them and kill all the units and then swarm them. Beyond that it doesn't seem to be happening. The AI is not committing to the attack or gets distracted and pulls out to early so the city starts to heal and then bounces back more or less instantly.

To put some minor numbers on it. This is just prior to the industrial era, or I have just entered it. Bombarding Rotterdam a city with 36 combat strength, 815 hp, with two range-3 field guns. I took a few hits before taking notes.

T1: 785 hp, before damage, artillery 1 and 2 does 78 and 75 dmg. So leaving Rotterdam with 631 hp after. It's a bit odd, there is some rounding shenanigans here cause that would be 154 dmg but I only did 153 dmg.

T2: The AI engages Defense and gets 50 combat strength. It healed 145 dmg so it just entirely shrugged off the previous artillery (or field gun) barrage. Or it took 8 points of damage to be correct. That is not much. At this pace it would take a 100 turns give or take to bombard it down with two siege units. Auch!

T3: I guess I better take some shots with my musketmen to then. Rotterdam 776dmg, two field guns, three musketmen later. 612 hp. Heals 159 dmg. Back to 771 hitpoints, so this turn it took 5 points of damage then. This isn't going to work out.

T4: Lets move some melee units a bit closer; range2. Bombard. 616 hp left. We are now standing one tile out, so city-empty-myunit. By the next turn it healed back up to 659 hp. So at least now it took 43 points of damage. So it should be done in less then 15 turns at least.

The problem with moving a lot closer is that you are really standing in the deathzone if you don't have all the promotions, some medic units etc. In some regard the spy mission to put a city in blockade just became a lot more useful, and probably a lot more important. It's probably the best and most secure way of taking a city now. Cause just bombarding it won't do the trick anymore. Unless you can shuffle a lot of units in and out (probably to their death) the old and previously normal way just isn't going to cut it anymore in any kind of normal time frame.
To me what you are describing is the system working exactly as intended. The whole idea was to remove the idea of a few long range artillery just shelling a city while the player sits back and sips coffee, that you needed to get "get your hands dirty" if you wanted to take a city quickly. And also 2 field guns isn't exactly a major siege force, I would at least have 3 before I consider it a "major offensive".

That said, if the AI is struggling with the concept that is of course a concern. Is it just in the range 3 era, or are you seeing that problem in the early wars as well?
 
At T3 it healed 43 damage but took 112 damage so not 15 turns but 5
So are you saying the corrected numbers are:

Rotterdamn 776
After attack (112 dmg): 664
After healing (43): 707

Total Damage: 69
Number of Turn Remaining to get city to 0: ~10.2
 
To me what you are describing is the system working exactly as intended. The whole idea was to remove the idea of a few long range artillery just shelling a city while the player sits back and sips coffee, that you needed to get "get your hands dirty" if you wanted to take a city quickly. And also 2 field guns isn't exactly a major siege force, I would at least have 3 before I consider it a "major offensive".

That said, if the AI is struggling with the concept that is of course a concern. Is it just in the range 3 era, or are you seeing that problem in the early wars as well?

I did note that I do kind of like it but I think the healing or recovery amount might be a bit much. Also that it seems more problematic for the AI then for me, I can always plan and work around it better. The key issue now is to time the spy mission with the actual attacks. I think in this particular case it was all that the terrain would allow for or what was available at the front at the time. That said I am not really seeing the AI take land locked cities at the moment. They took some very early or they have been able to snipe new cities that pop up that are basically at spawn undefended. They have the usual fairly good success with coastal cities that they can more easily rotate ships around and eventually bombard down. City-states they eventually swarm due to all the units getting eliminated. But beyond that I'm not really seeing any mid-late game wars where the AI is grabbing land and cities.

/edit/
An AI example. We are now in the middle-end part of the industrial era. Mecca is never going to fall to the Inca. They shoot, city takes minor damage and then heals to full at the start of the turn. Meanwhile the arabian camelarchers (which are now a bit old) just slaughter incan musketmen and push them back. The Incan charge next turn and back and forth it goes. The Inca is clearly not grasping the new siege mechanics; they have no siege units, they are not moving in melee into range to block, they are not doing spy stuff etc. There is some melee units behind the mountains to the south but it's I guess kept in reserve for protection of the Inca homeland.

While this is only an image it's fairly symptomatic of AI vs AI war, or even AI vs player. They just don't get the mechanics and instead just resort to killing units and the one that runs out of the units first will lose. Borders mainly only move due to citadels, not from cities actually changing hands and lands. Citadels are in that regard the only border movement I see between AI vs AI. Cities are in some regard now more or less unconquerable artillery bastions for the most part -- not saying that they won't once in a blue moon or so manage to pull one but it's very rare. In some regard it feels like the AI is just messing around.

c5vox-defence-mecca.png

(As always this is deity, marathon, etc etc so mileage may vary depending on such factors)




"T3: I guess I better take some shots with my musketmen to then. Rotterdam 776dmg, two field guns, three musketmen later. 612 hp. Heals 159 dmg. Back to 771 hitpoints, so this turn it took 5 points of damage then. This isn't going to work out."
This one might have been unclear. I took just sporadic notes so it might seem odd. Lets see if I can make it more clear.

Rotterdam started the turn with 776 hitpoints. Two field guns and three musketmen brought it down to 612 hitpoints; so 164 points of damage. During it's turn it heals back 159 points of damage bringing it back to 771 hitpoints. So over a turn cycle with the start_hitpoints - attack_damage + city_healing it took 5 points of damage (diff between 776 and 771 -- or the stating hitpoint values at the start of each turn) when you include the repairs.
 
Last edited:
So are you saying the corrected numbers are:

Rotterdamn 776
After attack (112 dmg): 664
After healing (43): 707

Total Damage: 69
Number of Turn Remaining to get city to 0: ~10.2
No. I see that T3 Rotterdam 771 then next turn after attack 616, 659 after healing. Total damage 112
Number of turns to get city to 0 about 6 because last time you don't give a turn for healing but take it with melee.
 
Status
Not open for further replies.
Top Bottom