Should the AI be, uh, *slightly* less likely to appoint sons as knights considering how likely they are to die?

[Kumicho]

I mean, this was my brother/vassal. 4 sons, all dead (probably in the same war), and only one of them had a kid of any kind (a daughter). I'm pretty sure that the AI doesn't take into account them dying and only looks at their prowess. Should this be changed to make the AI at least contemplate the possibility of them dying in combat? Maybe Forbid the primary heir from being a knight? Or possibly max out at ~50 or 75% of your sons as knights based on their prowess?

Or possibly prohibiting the primary heir from being a knight unless he had a Martial education?

 

[durbal]

I just use a lowered knight death chance mod and that resolves a lot of issues on its own. Realistically, the AI should care a lot especially when it only has one male heir or so, but even as a player I often don't mind sending my sons into battle when I use a lowered knight death mod. Knight deaths are just too high anyway and a few battles can wipe out your entire court.

I use this mod for what it's worth: https://steamcommunity.com/sharedfiles/filedetails/?id=2219643991

 

[Miroku20X6]

Knights are nobility. Worth much more as hostages than casualties.

I think the game needs to turn most (not all) knight deaths into captured knights instead. Then make it so that rulers can actually be knights, and that rulers/heirs get negative prestige for not being knights (via “forbid”, obviously if they’re not knight due to low prestige, that’s fine).

If knight deaths aren’t balanced enough that player-knights would make the game borderline unplayable due to your character dying every 5 years, then the solution is to fix knight deaths, not prohibit player-knights.

 

[Dayvit78]

No!
Death rates of knights overall should be reduced.
We need sons to be knights to gain experience, prestige, piety, etc. In fact, we need more things to do to our sons to get them ready to rule in the future. Right now, giving land is the only thing to do. There needs to be event chains for "introducing to realm" like in CK2, along with adult events for making friends, etc.

 

[Torredebelem]

The problem is Knights die or get mangled all too often while in combat. This is the root of the problem.

 

[gedsaro]

Really? I haven't noticed knights dying all *that* often, yes it happens regularly that they get wounded, some die etc, but its not like its even 1 every battle or anything.
are you guys having bigger battles than me maybe? Longer ones?
Is there an increased chance of death when you lose or get stack wiped?

Also pretty sure better knight's do, well better, I have noticed that unless I want or don't care about them, it IS bad to make anyone under about 10 prowess a knight.
Those of my knights who are in the 17-20+ range usually die of old age, or sometimes to a battle wound, but only once they are 50+ years old.

--Edit-- though i suppose if chance of death is based on ability, yes the AI could kill much of their court as I doubt they focus too heavily on making sure they have top teir knights, and probably just throw whoever can hold a sword into battle

 

[Havard]

Is there an increased chance of death when you lose or get stack wiped?

Knights seem pretty sturdy to me. I have some deaths, but not at an alarming rate. As to the matter of stackwipes:

Out of 46561 enemies, only 23 survived. 20 of the 25 knights survived.

Also:

Out of 35168 enemies, only 6 survived. 6 of the 7 knights survived.

 

[Peter Ebbesen]

Really? I haven't noticed knights dying all *that* often, yes it happens regularly that they get wounded, some die etc, but its not like its even 1 every battle or anything.
are you guys having bigger battles than me maybe? Longer ones?
Is there an increased chance of death when you lose or get stack wiped?

Knights dying in combat isn't something that should happen often to competent players except in the early game where they have few troops and quite possibly knights with low prowess values.

Knights are in many ways just like any other unit - they have a damage and a toughness value. Their base damage is 100 per prowess and their base toughness is 10 per prowess.

This is then adjusted by the knight effectiveness value.

Which is why the Gallant tree is so important for early warfare as a minor power, since you have few troops and your knights make up a major part of both your offensive and defensive strength. For a prowess 10 knight, that 75% extra effectiveness equates to another 750 damage output and 75 toughness. Later on buildings become most important for boosting knight effectiveness (military academies for feudal), but even in early warfare as tribal, you should be building war camps for more knights and the knight efficiency boost.

Think MAAs boosted by dozens of buildings are tough? They are. Extremely tough. But even mediocre knights gain toughness on the same scale without any buildings as default, and that's before taking lifestyle or other ways of boosting the knight effectiveness into account. High prowess knights with high effectiveness? They are almost unbelievably tough.

Knights can die during battle, but are extremely unlikely to unless they have low prowess. In ordinary fighting they are as noted above extremely tough.

They have their own events for getting maimed or killed in battle and barring knight-vs-knight interactions such as berserker, it is very difficult to get rid of a high prowess knight. The higher the prowess the lower the risk, scaling linearly; At 40+ prowess a character is nearly immune to getting maimed, at 30+ nearly immune to getting killed. There's still a risk, but it is very, very, low. Furthermore, there's outright immunity to death so long as your side outnumbers the other 5:1, so you are not going to see your knights dying in battles where you are mopping up over small enemy stacks.

EDIT: @Wethospu mentions in a post below that he has tested with a high casualty conversion factor without seeing any knight deaths except by event, which strongly suggests that the knight combat events are the only way for knights to die in combat. If anybody has a counterexample, let us know.

Of course, if you have many knights and fight often you will see them die every now and then, and if you have long close fought battles where you end up with few troops you may see a bloodbath amongst your knights during a last stand, and perhaps you fight some climatic battle against an enemy that also has lots of high prowess knights with high knight efficiency and experience mutual knight annihilation, but these cases ought not to occur regularly.

Knights can also die or be captured if an army loses a battle; this is a separate knight-only check.

If your knights are dying frequently, then they either have low prowess, are in long battles and hence have more chances to die, or you are cursed. Possibly all three. But outside the early game where your options are restricted, strategic miscalculation remains the most likely root cause.

There are mods to make combat less dangerous, but frankly I don't see the need given the plentiful tools available to making your knights shine.

EDIT: Tidied up and clarified.

-----

I realize this does not address the OPs question about whether the AI should use sons - and especially heirs - less as knights due to the risks involved. That is a hard question, since for roleplaying reasons it definitely should use heirs as knights, but since the AI is almost per definition incompetent, it would probably work better if the AI was more careful in that respect and didn't allow low prowess sons to be knights except when in extremis.

Players, however, ought not to have major problems keeping most of their knights alive if they invest in having good knights, and if they don't players should protect those sons who deserve it for the master plan and expose those who don't to the dangers of battle.

 

[Karlington]

I mean, this was my brother/vassal. 4 sons, all dead (probably in the same war), and only one of them had a kid of any kind (a daughter). I'm pretty sure that the AI doesn't take into account them dying and only looks at their prowess. Should this be changed to make the AI at least contemplate the possibility of them dying in combat? Maybe Forbid the primary heir from being a knight? Or possibly max out at ~50 or 75% of your sons as knights based on their prowess?

Or possibly prohibiting the primary heir from being a knight unless he had a Martial education?

View attachment 642120

Really? I haven't noticed knights dying all *that* often, yes it happens regularly that they get wounded, some die etc, but its not like its even 1 every battle or anything.
are you guys having bigger battles than me maybe? Longer ones?
Is there an increased chance of death when you lose or get stack wiped?

Also pretty sure better knight's do, well better, I have noticed that unless I want or don't care about them, it IS bad to make anyone under about 10 prowess a knight.
Those of my knights who are in the 17-20+ range usually die of old age, or sometimes to a battle wound, but only once they are 50+ years old.

--Edit-- though i suppose if chance of death is based on ability, yes the AI could kill much of their court as I doubt they focus too heavily on making sure they have top teir knights, and probably just throw whoever can hold a sword into battle

Knights seem pretty sturdy to me. I have some deaths, but not at an alarming rate. As to the matter of stackwipes:

View attachment 642764

Out of 46561 enemies, only 23 survived. 20 of the 25 knights survived.

Also:

View attachment 642774

Out of 35168 enemies, only 6 survived. 6 of the 7 knights survived.

It depends a lot on who you are fighting. OP @Kumicho's screenshot is from the British Isles, where one often fights the Norsemen.

Knights from a faith with the Warmonger tenet can get the Berserker trait while fighting in battle, which basically turns them into the CK3 equivalent of WMDs. I remember playing as a Norse king of the Danelaw. The king of Brittany decided to declare holy war on me, and he lost all his sons within months because my berserkers kept ripping their heads off.

That being said, even in the first screenshot from @Havard we see a casualty rate of 20% among the knights from a single battle. That's pretty high once we consider that a) many wars have multiple battles, especially early on before the player develops stackwiping capabilities, b) that the AI usually never develops good MAAs, so AI vs AI wars tend to have more battles, and c) the AI is war-crazy and keeps going to war over and over and over again.

This is one of the main reasons significant parts of Europe are ruled by women - all the male heirs died prematurely.

 

[Havard]

That being said, even in the first screenshot from @Havard we see a casualty rate of 20% among the knights from a single battle. That's pretty high once we consider that a) many wars have multiple battles, especially early on before the player develops stackwiping capabilities, b) that the AI usually never develops good MAAs, so AI vs AI wars tend to have more battles, and c) the AI is war-crazy and keeps going to war over and over and over again.

I'll have to stress the fact that examples are not representative for the general survival rate of knights. They both show stackwipes (as that was what I responded to) where the overall casual rate is 99,96% and more or less the only survivors are the knights.

 

[Havard]

I realize this does not address the OPs question about whether the AI should use sons - and especially heirs - less as knights due to the risks involved. That is a hard question, since for roleplaying reasons it definitely should, but since the AI is almost per definition incompetent, it would probably work better if the AI was more careful in that respect and didn't allow low prowess sons to be knights except when in extremis.

Players, however, ought not to have major problems keeping most of their knights alive if they invest in having good knights, and if they don't players should protect those sons who deserve it for the master plan and expose those who don't to the dangers of battle.

I agree. There is of course a risk, and in those situations where I have only one son and heir I will forbid him to fight as a knight. If I have several sons I tend to use them if their prowess is good enough. The prestige and development they get as characters are more often than not quite worth it. Both my last campaigns have been with elective succession, and having a great warrior as heir makes it a lot easier to sway the electors in my way.

 

[klinkvon13]

As pointed out by Miroku20X6, knight were worth being captured so they were incentive not to kill them. But mangling them as a bit similar.

So either this has to be reduced or the AI should be smart enough not to endanger a full line of heirs.
With 4 kids, it can be smart to improve experience of the two olders. When already 2 died, any half-witted king would tend to protect the remaining heirs.

 

[Peter Ebbesen]

That being said, even in the first screenshot from @Havard we see a casualty rate of 20% among the knights from a single battle. That's pretty high once we consider that a) many wars have multiple battles, especially early on before the player develops stackwiping capabilities, b) that the AI usually never develops good MAAs, so AI vs AI wars tend to have more battles, and c) the AI is war-crazy and keeps going to war over and over and over again.

You consider 20% death rate amongst knights on the losing side in that battle to be HIGH??????

A more appropriate term would be that the death rate is MIRACULOUSLY LOW.

46561 enemies took the field, including 25 knights
46538 enemies died, including 5 knights
23 enemies survived, including 20 knights

Or to put it another way, the levies and man-at-arms had a casualty rate of 99.994%, while the knights only had a 20% casualty rate.

Just how many knights would you expect to survive when only 23 people survived out of 46561 people, that included 25 knights at start? 21? 22? 23? (I refuse to accept answers that 24 or 25 out of 23 survivors should be knights )

@Havard 's first screenshot is a good case of showing just how tough good knights are, able to survive nearly anything.

 

[Wethospu]

@Peter Ebbesen I think knight effectiveness only affects their damage, but it's still a good boost.

I don't think Knights can die like any other unit. Only way for them to die is through battle events or at end of battle when losing. Knights survive even when setting BASE_RATIO_CASUALTIES_CONVERSION to 1 (which means all casualties are fatal).

 

[Zhetone]

something else that annoys me about this is that they'll usually have one son survive and their land never splits as a consequence.

 

[Peter Ebbesen]

@Peter Ebbesen

I think knight effectiveness only affects their damage, but it's still a good boost.

If this is the case then the tooltip for knight effectiveness is wrong and needs to be updated, because it currently states as an example that with 200% knight efficiency a knight fights as if his prowess is doubled, and it is the prowess that determines the damage and the toughness.

Of course, should the tooltip be wrong this would hardly be the first time, but lacking evidence to the contrary I believe the default position should be that it probably does do what it says.

I don't think Knights can die like any other unit. Only way for them to die is through battle events or at end of battle when losing. Knights survive even when setting BASE_RATIO_CASUALTIES_CONVERSION to 1 (which means all casualties are fatal).

Now, that is very interesting. Thanks.

 

[Xenrek]

Honestly I wish their was more to knights than just assigning.

Knights between realms could be everything from simply better trained soldiers to the medieval equivalent of tanks in large part based on how wealthy the average landholder/nobility was in the realm. Wish things like development and wealth played into their likelihood to die just as much as their prowess does.

 

[Wethospu]

If I'm reading files correctly, after combat:

• 10% * (1- prowess / 30) chance to get captured (minimum is 1% chance).
• 5% * (1- prowess / 40) chance to get killed (minimum is 0.5% chance).
• Stackwipe multiplies chances by 3.125. So about 30% capture and 15% death for low prowess knights.
• Brave, one legged, maimed, disfigured and one eyed double the chances.
• Craven halves the chances.
• Wounds increase chance of death 1.5x, 2x and 3x depending on the wound level.

Above values are not exact because of the weighting system. For example at 0 prowess there is 10% chance to get captured (10 / (10 + 5 + 85). With brave the chance would be 20 / (20 + 10 + 85) = ~17%.

 

[Wethospu]

If this is the case then the tooltip for knight effectiveness is wrong and needs to be updated, because it currently states as an example that with 200% knight efficiency a knight fights as if his prowess is doubled, and it is the prowess that determines the damage and the toughness.

Of course, should the tooltip be wrong this would hardly be the first time, but lacking evidence to the contrary I believe the default position should be that it probably does do what it says.

Here is some of my combat testing https://docs.google.com/spreadsheets/d/1A17P_fl-MQARkYm2XH1t_1Sp6ZYoE-XTXCfu1c16mR8

On sheet combat sample 2: columns Q and S are recorded knight strength and caused damage (based on save file data). Then on columns AK and AL is my calculated values.

Currently I have the +75% knight effectiveness only applied to damage dealt. Applying it to toughness would significantly affect remaining strength because that would reduce losses by 43%.

It's not a definite proof but better than nothing.

 

[Wethospu]

About knights event during battle I haven't too closely looked how they work in-game. But according to files there is:

• 2000 weight that nothing happens
• Stalwart leader increases this by 1000 for player and 300 for AI.
• Wounds add 10%, 20% or 30% depending on wound level.

• 30 weight for berserker ripping head off.
• Can only trigger with berserker trait and target enemy knights with 80% or less prowess.
• Stalwart leader increases this by 1000 for player and 300 for AI.
• Wrathful trait multiplies by 2.
• Giant trait multiplies by 10.
• Wounds add 25%.
• One legged, disfigured, one eyed or maimed multiply by 0.25.

• 30 weight to become berserker.
• Also kills enemy knight with 80% or less prowess if available.
• Can only trigger with Tenet warmonger doctrine and without Craven, Berserker or Calm traits.
• Stalwart leader increases this by 1000 for player and 300 for AI.
• Wrathful or Giant trait multiplies by 5.
• Impatient trait multilpies by 3.
• Sadistic, Brave, Ambitious and Norse culture multiplies by 2.
• Content multiplies by 0.5.
• Compassionate, Temperate, Lazy and Patient multiplies by 0.25.
• Wounds multiply 0.5, 0.5 or 0.25 depending on level.
• One legged, disgured, one eyed or maimed multiply by 0.25.

• 100 weight to gain a wound level.
• Can't trigger if already has max level wound.
• Can only be credited to an enemy knight with a 25% more prowess (if none available then no one gets the credit).
• Prowess multplies by (40 - prowess) / 40, minimum 0.1.
• For example 10 prowess multiplies 0.75 so about -25% chance to get wounded.
• Outnumbering the enemy reduces chances. Multiplied by 1.4 * enemy strength / own strength.
• For example double army causes about -30% chance to get wounded.
• Brave multiplies by 2.
• Craven multiplies by 0.5.

• 40 weight to gain a maim.
• Can't trigger if already has one legged, disfigured, one eyed or maimed,
• Can't trigger when vastly outnumbering the enemy (5x size).
• Can only be credited to an enemy knight with a 25% more prowess (if none available then no one gets the credit).
• Prowess multplies by (30 - prowess) /30, minimum 0.1.
• For example 10 prowess multiplies 0.66 so about -33% chance to get maimed.
• Outnumbering the enemy reduces chances. Multiplied by 1.4 * enemy strength / own strength.
• For example double army causes about -30% chance to get maimed.
• Brave multiplies by 2.
• Craven multiplies by 0.5.

• 30 weight to get killed.
• Can't trigger when vastly outnumbering the enemy (5x size).
• Can only be credited to an enemy knight with a 25% more prowess (if none available then no one gets the credit).
• Prowess multplies by (30 - prowess) /30, minimum 0.1.
• For example 10 prowess multiplies 0.66 so about -33% chance to get killed.
• Outnumbering the enemy reduces chances. Multiplied by 1.4 * enemy strength / own strength.
• For example double army causes about -30% chance to get killed.
• Brave multiplies by 2.
• Craven multiplies by 0.5.
• Wound level 3 multiplies by 4.

Chance for an outcome can be calculated by "weight of outcome" / "sum of all weights".

I think this event triggers every 5 days in a battle. I don't know does it trigger for every knight or just one per side. Probably for all knights because with some prowess there is only like 5% chance for something to happen.