Dev Blog 17/05/18

Users who are viewing this thread

[parsehtml]<p><img class="frame" src="https://www.taleworlds.com/Images/News/blog_post_40_taleworldswebsite.jpg" alt="" width="575" height="290" /></p> <p>From the steppe empires of Central Asia to the feudal states of Europe, cavalry was an integral part of medieval armies. Whether providing logistical support or charging the enemy on the battlefield, horses proved to be invaluable assets in warfare. In this week’s blog, we will take a look at these elegant and powerful animals and discuss the improvements and changes we have made to horses in Mount & Blade II: Bannerlord.</p></br> [/parsehtml]Read more at: http://www.taleworlds.com/en/Games/Bannerlord/Blog/60
 
Yes  :mrgreen:

To be a bit more serious - If a suggestion is not formulated to a reasonable degree, it is much less likely be implemented. For one, developers cannot be sure what exactly the user desires. Does he want a meaningful simulation? Or just a superficial illusion? ("Please add religion" comes to mind.) Is the user just thinking aloud casually or did he care enough to think things through? (Identified issues) Has he provided any solutions to potential problems or specific examples of improvements for consideration? (A developer may not always hold all the answers and sometimes even seemingly obvious input can help - See Yaga's Tree discussion f.e.)

My inquiry was not intented to put anyone on the spot. I just like to get people thinking about their suggestions - either so that they realize that it is not worthwhile/viable or so that they increase the chances of their input having an impact. Old modding habits, i suppose.
 
Duh said:
I just like to get people thinking about their suggestions - either so that they realize that it is not worthwhile/viable or so that they increase the chances of their input having an impact. Old modding habits, i suppose.

Easier said than done  :wink:
 
Duh said:
My inquiry was not intented to put anyone on the spot. I just like to get people thinking about their suggestions - either so that they realize that it is not worthwhile/viable or so that they increase the chances of their input having an impact. Old modding habits, i suppose.

So then the mechanic Khergitlancer suggested about lying should not be that hard to implement since the devs already made the algorithim and code for the honor and reown loss.
 
Rainbow Dash said:
I mean do you expect someone that browses this forum to be able to think up of a complex math equation and write a whole script to determine AI behavior
The considerations a player could reasonably be expected to make when betraying another lord are
[list type=decimal]
[*]Why do I want to do this (short term & long term if applicable) and what do I stand to gain?
[*]What do I stand to lose? Could I lose my land, title/rank, or material wealth? Could I be exiled from my faction? What's the Honor/Renown cost?
[*]Is there anyone else nearby who could intervene? If so, who do I think they will help or can I safely beat them both at the same time?
[*]Is my intended target part of my family (direct relation means more/less depending on culture)?
[*]Are any of my allies likely to turn on me when they hear about this?
[*]Will any of my enemies like me more for doing this, and are they stronger than the allies I would lose? Would they reward me?
[/list]
That's for the scenario described (agreeing to let defenders walk away and then attacking them after they surrender). There are additional considerations for acts of betrayal in other situations, but let's keep it simple. The first consideration could be a simple opinion & recent event check, which could be handled by a tagging/association system. The reason being you don't want lords to slaughter each other just because they dislike one another, as then they'll continue hating each other and be stuck in an endless cycle of butchery. So, check their opinion, and if it's negative then you check for tags associated with the opposing lord. Say Lord Rudeman is tagged for all kinds of mean stuff, like raiding caravans and villages, but there's one village in particular that he razed to the ground. Lord Angrypants owns that village, and has laid siege to Lord Rudeman's castle, with Lord Rudeman inside of it. Lord Angrypants not only has a negative opinion of Lord Rudeman, but he wants revenge because Rudeman utterly destroyed his fief. Lord Ridealong is helping Angrypants siege the castle, and Ridealong doesn't like Lord Rudeman either, but all Rudeman did to Ridealong was raid a couple of his caravans. Lord Ridealong isn't going to make the decision to slaughter Lord Rudeman after his surrender, and he may actually like Lord Angrypants less for it depending on Ridealong's honor value, but because he doesn't like Rudeman he isn't going to do anything to stop it from happening (maybe he would if he had a personality tag, like "merciful").

The second consideration is pretty straightforward: how much honor will I lose for doing this, will I gain or lose renown (I can see it going both ways), and will my liege lord punish me for doing it ostensibly in his name (assuming you're at war)? Some lords might not care at all, they could have a very vindictive personality or be impulsive. Others may feel justified for doing it, like Lord Angrypants, but would refrain because they know that their liege lord wouldn't care for it at all. This is more nuanced, unlike the above, because instead of asking "do I have a reason to do this" we're asking "is my reason good enough to justify my potential losses?" This could, practically speaking, be decided by a weighted dice roll which is modified by personality type of the lord in question, the personality of their liege lord (if it matters), and the weighted significance of their reason (e.g. the murder of one of your family members is a very compelling reason, but raiding your village isn't), which we can call the Justification Value. Check the result of his dice roll against a normalized average of the lord's holdings and material wealth, modified by their liege lord's current opinion if applicable, and away you go, which we'll call the Potential Loss Value.

As an example, Lord Angrypants may be vindictive so he'll believe a wider range of offenses are sufficient justification for what he's about to do. So, what goes through his mind is "I hate Lord Rudeman, but my liege lord would be a little upset if I murdered all of Rudeman's soldiers after they surrendered. That said, he murdered a bunch of my serfs and burned their houses down, so I think this is fair." Angrypants would have a pretty high modifier on his dice roll for determining his final Justification Value. We take that value and match it up against Lord Angrypants' estimated wealth and the current opinion his lord has of him. Let's say Angrypants doesn't have much in the way of fiefs, just a couple of villages (one of which was burned to the ground not too long ago) and maybe a castle or fort nearby. Also, Lord Angrypants has been at odds with his liege lord for a while, so while their relations aren't bad they are somewhat strained, close to neutral. Lord Angrypants would upset his lord by betraying Lord Rudeman, and because he's already straining that relationship there's a good chance that his liege lord would punish him by taking away some of his land. Lord Angrypants doesn't have much land, so he's keen to hang on to what he has left, so he's unlikely to risk it. He's likely to lose a large portion of his fiefs, which would make him even poorer, and his liege lord would actively dislike him, so his Potential Loss Value is very high. There's still a chance he would follow through, depending on the Justification Value dice roll.

For the third consideration, we can make a similar calculation to what we already do when lords decide to attack each other on the map. Basically, check their surroundings for additional parties and determine if they are a threat. The relations change you receive with your liege lord for a betrayal could be delayed for weeks, but if someone is with you when you do it then their relations would change immediately. Some may support you, some may be indifferent (like Lord Ridealong), and others may try to stop you. There could also just be a hostile party nearby that would take advantage of the opportunity to bushwhack you.

The fourth consideration could be applied in the second step, where we calculated what losses we might face for betraying a lord. If that lord is family and our culture places great significance on family loyalty, then we could heavily weight the calculation against betrayal. This could be applied in a very simple way as a percentage modifier of the Potential Loss Value. If family is very important, then we could multiply the value by 2.0, whereas if family is inconsequential then we could multiply it by 1.0. By carefully choosing our modifiers for the Justification Value, we could still make it remotely possible for loyal family members to betray each other, but they would have to hate each others' guts before they would do it.

The fifth and sixth considerations could all be lumped in to the first and second, but I feel they're important enough to discuss on their own. Social status & reputation are quite important, and with the wide variety of personalities that would be present it's quite possible to win as many new friends as you would lose old ones. Some of the friends you make could even be in your enemy's faction, and this could be a way to ingratiate yourself with them before you switch sides. Maybe Lord Angrypants won't care that his lord will take away his land, because he knows that Lord Rudeman is widely disliked in his own faction and Angrypants isn't content with his current status at home. So, Lord Angrypants decides to kick Rudeman while he's down, jumps ship, and goes looking for greener pastures. Alternatively, some of Lord Rudeman's family might get really upset with Lord Angrypants for what he's done and raze both of his villages, siege his castle, and do the same thing to him.
tl;dr These could all be tied to different numerical values and then manipulated any number of ways with a dash of randomness thrown in to account for irrationality and emotions. In fact, I'd much prefer if it wasn't just a simple renown/honor calculation. I want the AI to have a reason for doing it, something more than "I have -55 relations with you."
 
Orion said:
Rainbow Dash said:
I mean do you expect someone that browses this forum to be able to think up of a complex math equation and write a whole script to determine AI behavior
The considerations a player could reasonably be expected to make when betraying another lord are
[list type=decimal]
[*]Why do I want to do this (short term & long term if applicable) and what do I stand to gain?
[*]What do I stand to lose? Could I lose my land, title/rank, or material wealth? Could I be exiled from my faction? What's the Honor/Renown cost?
[*]Is there anyone else nearby who could intervene? If so, who do I think they will help or can I safely beat them both at the same time?
[*]Is my intended target part of my family (direct relation means more/less depending on culture)?
[*]Are any of my allies likely to turn on me when they hear about this?
[*]Will any of my enemies like me more for doing this, and are they stronger than the allies I would lose? Would they reward me?
[/list]
That's for the scenario described (agreeing to let defenders walk away and then attacking them after they surrender). There are additional considerations for acts of betrayal in other situations, but let's keep it simple. The first consideration could be a simple opinion & recent event check, which could be handled by a tagging/association system. The reason being you don't want lords to slaughter each other just because they dislike one another, as then they'll continue hating each other and be stuck in an endless cycle of butchery. So, check their opinion, and if it's negative then you check for tags associated with the opposing lord. Say Lord Rudeman is tagged for all kinds of mean stuff, like raiding caravans and villages, but there's one village in particular that he razed to the ground. Lord Angrypants owns that village, and has laid siege to Lord Rudeman's castle, with Lord Rudeman inside of it. Lord Angrypants not only has a negative opinion of Lord Rudeman, but he wants revenge because Rudeman utterly destroyed his fief. Lord Ridealong is helping Angrypants siege the castle, and Ridealong doesn't like Lord Rudeman either, but all Rudeman did to Ridealong was raid a couple of his caravans. Lord Ridealong isn't going to make the decision to slaughter Lord Rudeman after his surrender, and he may actually like Lord Angrypants less for it depending on Ridealong's honor value, but because he doesn't like Rudeman he isn't going to do anything to stop it from happening (maybe he would if he had a personality tag, like "merciful").

The second consideration is pretty straightforward: how much honor will I lose for doing this, will I gain or lose renown (I can see it going both ways), and will my liege lord punish me for doing it ostensibly in his name (assuming you're at war)? Some lords might not care at all, they could have a very vindictive personality or be impulsive. Others may feel justified for doing it, like Lord Angrypants, but would refrain because they know that their liege lord wouldn't care for it at all. This is more nuanced, unlike the above, because instead of asking "do I have a reason to do this" we're asking "is my reason good enough to justify my potential losses?" This could, practically speaking, be decided by a weighted dice roll which is modified by personality type of the lord in question, the personality of their liege lord (if it matters), and the weighted significance of their reason (e.g. the murder of one of your family members is a very compelling reason, but raiding your village isn't), which we can call the Justification Value. Check the result of his dice roll against a normalized average of the lord's holdings and material wealth, modified by their liege lord's current opinion if applicable, and away you go, which we'll call the Potential Loss Value.

As an example, Lord Angrypants may be vindictive so he'll believe a wider range of offenses are sufficient justification for what he's about to do. So, what goes through his mind is "I hate Lord Rudeman, but my liege lord would be a little upset if I murdered all of Rudeman's soldiers after they surrendered. That said, he murdered a bunch of my serfs and burned their houses down, so I think this is fair." Angrypants would have a pretty high modifier on his dice roll for determining his final Justification Value. We take that value and match it up against Lord Angrypants' estimated wealth and the current opinion his lord has of him. Let's say Angrypants doesn't have much in the way of fiefs, just a couple of villages (one of which was burned to the ground not too long ago) and maybe a castle or fort nearby. Also, Lord Angrypants has been at odds with his liege lord for a while, so while their relations aren't bad they are somewhat strained, close to neutral. Lord Angrypants would upset his lord by betraying Lord Rudeman, and because he's already straining that relationship there's a good chance that his liege lord would punish him by taking away some of his land. Lord Angrypants doesn't have much land, so he's keen to hang on to what he has left, so he's unlikely to risk it. He's likely to lose a large portion of his fiefs, which would make him even poorer, and his liege lord would actively dislike him, so his Potential Loss Value is very high. There's still a chance he would follow through, depending on the Justification Value dice roll.

For the third consideration, we can make a similar calculation to what we already do when lords decide to attack each other on the map. Basically, check their surroundings for additional parties and determine if they are a threat. The relations change you receive with your liege lord for a betrayal could be delayed for weeks, but if someone is with you when you do it then their relations would change immediately. Some may support you, some may be indifferent (like Lord Ridealong), and others may try to stop you. There could also just be a hostile party nearby that would take advantage of the opportunity to bushwhack you.

The fourth consideration could be applied in the second step, where we calculated what losses we might face for betraying a lord. If that lord is family and our culture places great significance on family loyalty, then we could heavily weight the calculation against betrayal. This could be applied in a very simple way as a percentage modifier of the Potential Loss Value. If family is very important, then we could multiply the value by 2.0, whereas if family is inconsequential then we could multiply it by 1.0. By carefully choosing our modifiers for the Justification Value, we could still make it remotely possible for loyal family members to betray each other, but they would have to hate each others' guts before they would do it.

The fifth and sixth considerations could all be lumped in to the first and second, but I feel they're important enough to discuss on their own. Social status & reputation are quite important, and with the wide variety of personalities that would be present it's quite possible to win as many new friends as you would lose old ones. Some of the friends you make could even be in your enemy's faction, and this could be a way to ingratiate yourself with them before you switch sides. Maybe Lord Angrypants won't care that his lord will take away his land, because he knows that Lord Rudeman is widely disliked in his own faction and Angrypants isn't content with his current status at home. So, Lord Angrypants decides to kick Rudeman while he's down, jumps ship, and goes looking for greener pastures. Alternatively, some of Lord Rudeman's family might get really upset with Lord Angrypants for what he's done and raze both of his villages, siege his castle, and do the same thing to him.
tl;dr These could all be tied to different numerical values and then manipulated any number of ways with a dash of randomness thrown in to account for irrationality and emotions. In fact, I'd much prefer if it wasn't just a simple renown/honor calculation. I want the AI to have a reason for doing it, something more than "I have -55 relations with you."


+1

Somethinng like this is not very hard to implement, and maybe have it so that the lying mechanic could be limited to some debauched lords.
 
Deal = surrender the town & your garrison can leave with their arms and equipment.
Your watch towers report either 1 the enemy have cleared their siege lines outside the exit gate or 2 the enemy have not withdrawn from the agreed exit gate.
Your choices if the enemy have not withdrawn are either: 1 refuse to surrender the town until the enemy withdraw from your exit gate or 2 march out at a disadvantage trusting the enemy to honour their agreement (possible but not guaranteed).
You know from the in game encyclopaedia that your opponent lacks honour, but you march out regardless. An ambush battle ensues and your forces are defeated. You are given a choice to flee on horseback with a considerable loss of honour and renown or join your troops in captivity. The enemy Lord also loses some renown and a lot of honour as well as being flagged for execution if subsequently captured by your faction.
If you decide not to flee, there is a random chance that your captor wants no witnesses to his ill deeds, resulting in your execution - game over.
You had the option to refuse to surrender the town. You had fair warning about your opponent’s character. You had the option to flee. You also had a random chance that you wouldn’t be executed. Is that a sufficient safety net or will you still feel annoyed that you have been executed and dumped out of the game?
 
NPC99 said:
Deal = surrender the town & your garrison can leave with their arms and equipment.
Your watch towers report either 1 the enemy have cleared their siege lines outside the exit gate or 2 the enemy have not withdrawn from the agreed exit gate.
Your choices if the enemy have not withdrawn are either: 1 refuse to surrender the town until the enemy withdraw from your exit gate or 2 march out at a disadvantage trusting the enemy to honour their agreement (possible but not guaranteed).
You know from the in game encyclopaedia that your opponent lacks honour, but you march out regardless. An ambush battle ensues and your forces are defeated. You are given a choice to flee on horseback with a considerable loss of honour and renown or join your troops in captivity. The enemy Lord also loses some renown and a lot of honour as well as being flagged for execution if subsequently captured by your faction.
If you decide not to flee, there is a random chance that your captor wants no witnesses to his ill deeds, resulting in your execution - game over.
You had the option to refuse to surrender the town. You had fair warning about your opponent’s character. You had the option to flee. You also had a random chance that you wouldn’t be executed. Is that a sufficient safety net or will you still feel annoyed that you have been executed and dumped out of the game?

Theres still some improvements to make to something like this.

Thanks for reminding me that you can now execute lords.

Now for the improvements:

-Maybe make it so if you are caught lying you are marked for execution. This applies to AI lords too.

-Why does fleeing a battle make you lose honor?

-remove the "random chance" thing about how he wants everyone to be slaughtered. This promotes save scumming.

 
Rainbow Dash said:
I mean do you expect someone that browses this forum to be able to think up of a complex math equation and write a whole script to determine AI behavior for a game that isin't released yet?

Dude, what're you trying to say about the people that browse this forum? Don't consider our standard you.
 
KhergitLancer80 said:
Callum replied to my question people:

I talked with a dev about it and he said that he doesn't think that Warband was ignoring that velocity. As far as he can remember, the code was considering the shooter agent's velocity. The agent can either be mounted or not, doesn't matter. So in Bannerlord that velocity computation will be %100 realistic.

All the best,

Callum
In Warband shooters velocity may be considered but it's not added to the velocity of arrow. I suspect it's considered during damage calculations. Unfortunately it looks like its the same in Bannerlord or at least on last gameplay with horse archery. Arrows lag behind the shooter when firing at 90 degrees witch suggest the horse velocity was not added.
snxQepu.gif
 
+M4rm4n

Well, my question was ''specifically about'' its addition to arrows velocity. Other than that ofc the velocity of horse isnt ignored, horses arent galloping on the same spot after all  :lol:
Plus, the gif you took is very hard to be read. In fact, I even can say it goes a bit oblique rather than straight if I push myself so hard.


I think I am going to stick with their words on this, considering my overall trust to their great passion and because this is what I want to believe in :lol:

EDIT: Wow, I am a knight at arms now, I have a fancier lance !
 
KhergitLancer80 said:
+M4rm4n

Well, my question was ''specifically about'' its addition to arrows velocity. Other than that ofc the velocity of horse isnt ignored, horses arent galloping on the same spot after all  :lol:
Plus, the gif you took is very hard to be read. In fact, I even can say it goes a bit oblique rather than straight if I push myself so hard.


I think I am going to stick with their words on this, considering my overall trust to their great passion and because this is what I want to believe in :lol:

EDIT: Wow, I am a knight at arms now, I have a fancier lance !
Oblique path of arrow is only illusion because of changing perspective. The point being that if arrow inherits velocity of horse it should fly straight.
 
Back
Top Bottom