Information about developments at snowballing problem

[babelfisch]

We do not give lots of free soldiers to AI, only about 10 free soldiers when they first spawn. It has nothing related to our problems. This is nothing big.

I am not fan of AI cheats and try to remove as much as possible but we have to give passive XP to NPC parties because player party and NPC parties so different in lots of terms. Player make more battles compared to an average lord party (average lord party life (time between spawned - prisoned) is less shorter than player party) and if we do not give passive XP as AI cheat 80% of troops at Npc parties be tier-1 or tier-2. This damages gameplay so badly, we have tons of troop variety but we cannot show them to player. With passive XP cheat this ratio (tier-1, tier-2 troop ratio) reduces to 50%. I want to reduce passive Xp more but I fear to damage gameplay especially at combat side.

You suggested increasing fief income but lets examine a test from 1.5.8 results :

As you see total kingdom budget is 8000K at 1084 (game start) but it rises 30000K at 1104 (20 years later) so this shows we already have a money inflation at world and as you see kingdoms with high number of fortification (Khuzait, Vlandia, Aserai, S.Empire in sample test) have high money inflation. Increasing fief incomes will make rich kingdoms even richer and create new problems and make money inflation worse. Reducing loot income can be one solution but these lords mostly do not earn money from looter loots already. Loot is income for everybody while fief income is income for mostly powerfull kingdoms. Already 66% of total income is fief income currently remaining is loot income. Increasing fief income do not solve our problem, Battanians have already 2 town 2 castles in total for last 10 years of test they already have limited number of fiefs. Making fief income 1.5x will give them daily 2000 more maybe while we will be giving 10000 more to each other kingdom. This will help stronger factions with bigger territory more and total kingdom budget of 30000K will be 50000K at 20th year.

What about diminishing returns for lots of fiefs?

Simplified, currently income is pretty linear, 5 fiefs give you 5 times more money than 1 fief. If you add diminishing returns you would buff smaller kingdoms.

This might sound artificial at first, but it's actually realistic. More territory means more costs for administration.

 

[scarface52]

@mexxico your settlement value and prosperity and related change was included in the latest hotfix or am I mistaken?

 

[mexxico]

@mexxico your settlement value and prosperity and related change was included in the latest hotfix or am I mistaken?

Yes, valuing kingdom's settlements for war / peace calculations changed with latest hotfix, a minor change we discussed here (making 1000 + prosperity * 0.33 instead of directly using prosperity). However these have minor effects on snowballing. Did you see code changes?

Your 1.5.8 example looks pretty good in terms of snowballing. Vlandia is huge but I am ok with only one kingdom getting huge in every campaign after 20 years if it is not always the same (Khuzaits). Interesting to see the SE’s come back and how Khuzaits were pushed back after getting 32 fortifications.

If you have time in the following days, would be great if you could share what changes we can expect for 1.5.8 related to snowballing. Thanks!!!

Actually this can be a bit luck having balanced run. I do not think snowballing score will be so different from 1.5.6 or 1.5.7 at 1.5.8. Because there is no huge changes.

 

[scarface52]

Yes, valuing kingdom's settlements for war / peace calculations changed with latest hotfix, a minor change we discussed here (making 1000 + prosperity * 0.33 instead of directly using prosperity). However these have minor effects on snowballing. Did you see code changes?

Yes, I check diff after patches out of curiosity. Asked because it wasn't mentioned in the patch notes with bunch of other stuff.

 

[Marbles20]

I'm still curious to know what the tribute income for each kingdom is over that 20 year period. I know +/-200 income per day per clan doesn't seem like a lot but it still seems like it would create a growing disparity between the strong kingdoms recieving tribute versus the weak ones giving tribute.

 

[Blood Gryphon]

@mexxico so what didn't make it in hotfix? Guessing mercenary changes is 1.5.8?

Nice catch detective @scarface52

 

[Abyssos]

So im hopping dev read this couple issues im having: i spend alot time to gather a good army with only tier6 troops for me and companions, then to see them dumped in garrisons,so far its ok cause if they are mine i can ge them back dispite is damm annoying keep doing this, problem is im trying now playing a diplomatic way and getting vassals, and everytime i get a castle/town i noticed that everytime i conquer one i cant keep them, have to give them to someone else(and when they been sigged we dont get the "alarm" sign).the other thing is everytime we visit vassals town/castle our companions dump a number of troops on it..wich make them impossble to get back..please change this its impossible to keep a good army this way,until there i go marry svana again and going back to "beheading all" mood until this is fixed ...regards

 

[mexxico]

@mexxico so what didn't make it in hotfix? Guessing mercenary changes is 1.5.8?

Nice catch detective @scarface52

I did not understand first question. Yes mercenary changes is at 1.5.8

By the way something can be annoying for player and can be also related to snowballing (long term side effect) is sometimes parties around enemy cannot catch it even they should :

Here normally weak and fast parties should catch 733 men Khuzait army. Black numbers are speed of parties. However 82 men Empire party come close to Khuzait army and run away back because it thinks 1208 men army cannot join battle (can be out of radius at first moment but it will catch up). This behavior saves Khuzait army from that bad situation. You can see how Khuzait army is fast (3.4 vs 1.9 (forest and bigger army disadvantages are effective also)) compared to Empire one because of horses it carry and better cavalry ratio it has. But even this speed advantage here weak & fast parties should catch Khuzait party and it should not escape from this situation. I will examine it and think about improvements. Something valueable can be found from here to improve gameplay & kingdom balance.

 

[SGT_Night]

I did not understand first question. Yes mercenary changes is at 1.5.8

he just asked if any changes you talked about here were not in this new hotfix that came out

 

[mexxico]

he just asked if any changes you talked about here were not in this new hotfix that came out

I can add several lines of small changes at hotfix at most if there is nothing gamebreaking. Value calculation was a small code change so I added it only.

 

[Ragemaster9999]

I did not understand first question. Yes mercenary changes is at 1.5.8

By the way something can be annoying for player and can be also related to snowballing (long term side effect) is sometimes parties around enemy cannot catch it even they should :

Here normally weak and fast parties should catch 733 men Khuzait army. Black numbers are speed of parties. However 82 men Empire party come close to Khuzait army and run away back because it thinks 1208 men army cannot join battle (can be out of radius at first moment but it will catch up). This behavior saves Khuzait army from that bad situation. You can see how Khuzait army is fast (3.4 vs 1.9 (forest and bigger army disadvantages are effective also)) compared to Empire one because of horses it carry and better cavalry ratio it has. But even this speed advantage here weak & fast parties should catch Khuzait party and it should not escape from this situation. I will examine it and think about improvements. Something valueable can be found from here to improve gameplay & kingdom balance.

DUDE I have this happen ALL the time in my playhrough, it is so annoying when there are two armies playing ping pong with smaller "tackle" units going back and forth between chasing and retreating. Last night I had a 1000 army in the middle, and a 600 man army on the left and a 600 man army on the right, the 1000 was inbetween the other two armies and they refused to engage because of the reinforcement radius. I really suggest increasing the range at which a party will consider another party "in range" for engagement purposes. The AI is overly squeamish sometime! Even if it starts the fight outnumbered, it should factor in the chance of nearby parties dog piling into the fight.

 

[SGT_Night]

This is very annoying indeed, also there is the fact that smaller army's keeps them selfs pretty close to bigger enemy army, almost following behind, they only ran away when an enemy starts following then, this happens quite alot, and the weaker army always get caught, would be great if they mantain more distance between an bigger foe

 

[Dabos37]

The prosperity change for not be the only factor to take into account concerning fiefs when declaring new wars, is working pretty fine as far I have seen. Now I can see kingdoms declaring war on Khuzaits which rarely happened before. Thanks for that.

I did not understand first question. Yes mercenary changes is at 1.5.8

By the way something can be annoying for player and can be also related to snowballing (long term side effect) is sometimes parties around enemy cannot catch it even they should :

Here normally weak and fast parties should catch 733 men Khuzait army. Black numbers are speed of parties. However 82 men Empire party come close to Khuzait army and run away back because it thinks 1208 men army cannot join battle (can be out of radius at first moment but it will catch up). This behavior saves Khuzait army from that bad situation. You can see how Khuzait army is fast (3.4 vs 1.9 (forest and bigger army disadvantages are effective also)) compared to Empire one because of horses it carry and better cavalry ratio it has. But even this speed advantage here weak & fast parties should catch Khuzait party and it should not escape from this situation. I will examine it and think about improvements. Something valueable can be found from here to improve gameplay & kingdom balance.

Thanks! Everything you can find to harm Khuzaits will be pretty welcome!

 

[SGT_Night]

DUDE I have this happen ALL the time in my playhrough, it is so annoying when there are two armies playing ping pong with smaller "tackle" units going back and forth between chasing and retreating. Last night I had a 1000 army in the middle, and a 600 man army on the left and a 600 man army on the right, the 1000 was inbetween the other two armies and they refused to engage because of the reinforcement radius. I really suggest increasing the range at which a party will consider another party "in range" for engagement purposes. The AI is overly squeamish sometime! Even if it starts the fight outnumbered, it should factor in the chance of nearby parties dog piling into the fight.

Yes, the smaller army should consider that he is outnumbered and should tried to escape as well

 

[Dabos37]

This is very annoying indeed, also there is the fact that smaller army's keeps them selfs pretty close to bigger enemy army, almost following behind, they only ran away when an enemy starts following then, this happens quite alot, and the weaker army always get caught, would be great if they mantain more distance between an bigger foe

Agree with this, I hate when this happens usually in Khuzaits’ favor.

By the way, increasing the range for battle reinforcements would be great, also if it would affects the player. I feel like cheating when attacking one enemy party and another one is pretty close but not in range for a bit. It is unfair for the AI because the player stops the time in battles.

 

[Blood Gryphon]

I did not understand first question. Yes mercenary changes is at 1.5.8

My bad it was basically the same question as 2nd. Thanks for the answer

Off topic but something I've been curious about, wouldn't factions be able to respond to things a lot better if they could dynamically call and release people from their armies (like the player does or at least me)? I feel like right now an army forms with a goal and after that goal they roam around doing what is within their means until they disband. Often times I wish these armies would disband and reform to deal with the issue they are attempting instead of with the remnants of their previous goal. This is especially true when armies start to go around and recruit, has to be the least efficient way to rebuild your forces.

 

[Pejot]

@mexxico if You reduce loot income, increase fief income and disable cheat income to lords You will at least give the option to weaken big empires by raiding villages. This will cut off some money sources. Also how is the cav upgrade cos calculated for NPC? Do they pay additional cause of no need for horses cheat?

One more thing that can slower bigger kingdoms would be giving additional influence cost multiplier for creating an army based on how many fiefs a kingdom has. So the bigger You grow the hard it is to create big armies. It would simulate NPC lords loosing interest in participating in wars cause of how much they already have. Smaller kingdoms could get positive multiplier for that so raising army would cost less influence for them.

 

[Nalgasucia507]

I did not understand first question. Yes mercenary changes is at 1.5.8

By the way something can be annoying for player and can be also related to snowballing (long term side effect) is sometimes parties around enemy cannot catch it even they should :

Here normally weak and fast parties should catch 733 men Khuzait army. Black numbers are speed of parties. However 82 men Empire party come close to Khuzait army and run away back because it thinks 1208 men army cannot join battle (can be out of radius at first moment but it will catch up). This behavior saves Khuzait army from that bad situation. You can see how Khuzait army is fast (3.4 vs 1.9 (forest and bigger army disadvantages are effective also)) compared to Empire one because of horses it carry and better cavalry ratio it has. But even this speed advantage here weak & fast parties should catch Khuzait party and it should not escape from this situation. I will examine it and think about improvements. Something valueable can be found from here to improve gameplay & kingdom balance.

+1. You are fighting the good fight mexxico. We are rooting for you.

 

[Dabos37]

I have tested this afternoon the result of reducing loot income and to be honest, it does not change much.

The changes I would like to see in the future for reducing snowballing:

- A much slower war pace. Seriously, it is not fun to see kingdoms at war 1679 of 1680 days in a 20 years campaign.
- Something to prevent AI going bankrupt in peace times (wages limits or whatever).
- A more defensive AI.

Concerning the AI, I am usually seeing this scenario:

1- Khuzaits create a 1500 men army, and siege a pretty well defended Empire town (400-500 men).
2- The SE or NE kingdom has an 800 men army, it evaluates that the 1500 men army is too strong, so decides to attack an useless far away castle.
3- The Khuzait's army gets reduced to 300 men or less after taking the well defended town, if the 800 Empire army would have waited close to the well defended town, they would have been able to protect it.
4- Some days later, the 800 Empire men army gets defeated by 600-700 Khuzaits army when was assaulting the useless far away castle. After losing a key town.

In the same way than the Player does, the AI should take into account that enemy armies will get tons of losses when assaulting settlements, so they could be a bit smarter and wait for a good oportunity, instead of going to attack a silly castle and getting defeated when trying it.

 

[Dabos37]

I want to share an image with all of you, especially with @mexxico

Day 1409:

6 hosted at ImgBB

Image 6 hosted in ImgBB

ibb.co

Khuzaits did overexpanded and got ganked.

First time I have seen this since the release. While snowballing is still present and Vlandia is almost gone, I think this is pretty encouraging. The hotfix change about settlements fixed value + prosperity * 0.35 has been simply amazing. We are pretty close!!