Scaling bandit party size with player level is bad design + solution

Do you want bandits to spawn dynamically based on settlement prosperity?

  • Yes.

    Votes: 28 80.0%
  • No.

    Votes: 1 2.9%
  • There are better ways to solve the issue.

    Votes: 6 17.1%

  • Total voters
    35

Users who are viewing this thread

So, here's the problem. I'm 400+ days into my campaign and banditry is absolutely wrecking Calradia. The more the game progresses, the harder it is for villagers, caravans and lords to live. Why? Apparently, there is a mechanic that scales bandit parties according to the player level. Essentially, this means that the entire world has to suffer more simply because some random player guy is leveling up. I can elaborate on a few issues that this mechanic brings up and why it's bad design:

1. Bandits are cheating units, while the rest of the parties have to work their way to power in more natural manners. Let's take lord parties as an example: notables produce units, and it takes time to train them, all that WHILE the numbers of notables/settlement depends on prosperity; thus, the number of units produced per settlement is reasonably dynamic. In order for lord parties to grow, they have to recruit directly from these settlements, and then take time to train their units in battles. Now, if you ask, how bandit parties evolve? They just puke units out of nowhere, and the more you level up, the harder it is for the AI lords to deal with it. Lord parties will always tend to be very weak mid to late game because they expend their precious troops (which have been produced dynamically) on cheated bandit parties.
2. The mechanic was (supposedly) designed to pose a challenge to you, but it does not achieve its goal. By mid-game you'll be running around with 100+ elite troops around the map, and nothing the bandits can cheat at you will be a challenge.
3. If you intend to police the world in order to make the roads safer for caravans, you're hopeless. That's because:
a. It gets boring and repetitive really fast;
b. Your actions don't matter at all. You can destroy as many parties as you want, but they will cheat more, stronger and bigger parties as you level up. At some point, you just give up and ignore bandits altogether.
4. Only the player can destroy hideouts, and that's horrible. I genuinely tried to reduce the bandit spawns by destroying as many hideouts I can, but I quickly understood that it's like playing whack-a-mole. More and more hideouts will pop around the world, and you alone cannot do anything to stop it from happening. Like in the previous case, it also gets repetitive and boring. The world map will eventually be filled with hideouts and the AI won't do anything about them.
5. Villager parties are hopeless. In the early stages of the game, you're used to seeing 30-40 big villager parties. In mid-game, these parties are down to ~15 in size, and it will only get worse. While villagers take time to recover and grow their parties, bandits don't care, because they're cheated.
6. Owning caravans gets too unfeasible. Early game, they're able to defend themselves reasonably well, but from mid-game, it starts to feel like the wheel of luck, because banditry is out of control.

All these point to bad or lazy design, but I thought about some changes that could be made to the system. The goal is to spawn bandits dynamically and make them play by the same rules, just as the other AI. In short: no cheating units.
- The player level will not matter anymore in bandit spawn behavior;
- Bandits will spawn dynamically, based on settlement prosperity. The idea is simple. Low settlement prosperity = higher chance to spawn a bandit party from that settlement. The lower the prosperity, the more and bigger parties are spawned. Higher prosperity settlements would still spawn bandits but at lower rates and smaller in size. This means that wars, and not cheating, will influence the security of a kingdom. If kingdom X has not been able to defend its settlements from kingdom Y village raids, then kingdom X will end up with more bandits on the roads. This is simple, dynamic and makes all the sense.
- Bandits will have to level up their troops in battle.
- AI lords will also clean up hideouts.

Optionally:
- Add the option to enlist settlement patrols for the AI, and the player;
- Bring back the manhunters, and allow them to raid hideouts. The manhunters would also spawn from settlements with low prosperity, but at lower rates than bandits.

@mexxico Sorry for the ping, but I feel that this cannot be left as it is. Please bring up the issue if you have the possibility.
 
Yea, I absolutely hated the bandit spam mechanic since I got my hands on first MB years ago. Bandit infestation is beyond reason, beyond control, beyond any interaction and is and always was wrecking gameplay in later stages of the game. Its only purpose is to give player easy leveling and income but at the expense of everything else in the game. It's definitely a case of lazy design, I agree with you.

In Warband I always avoided to settle near lair spawn areas because of the economic impact and complete lack of options how to deal with them.

Wiking Conquest at last had an option to scale their spawn down.

I agree also with your proposition to make them non scalable. Fine, they serve early leveling and income purpose, but there is no need for bandit deathstars mid and late game when player is usually fighting noble armies.

Making them dependent on dynamic factors like prosperity or wars is also good idea in my opinion.
 
Last edited:
are you sure that bandits increase with player level? where/how did you discover this?
Only the player can destroy hideouts
this is wrong. when a lord enters a settlement, there is a chance for an issue to be resolved. if it was a hideout quest, it will be destroyed.
link here
Normally lords already do that. In some cases they clear hideouts (but not by entering hideout by entering a settlement which has clear hideout issue)

i don't like spawning due to low prosperity. if it were security, i'd agree.
also, wouldn't high prosperity fiefs be a better target for banditry?

a better way to challenge the player is to have bandit parties merge when they get close to each other (a mod does this) and to be able to upgrade their weapons when they defeat stronger (lords) parties.
warband had a mod that lets the player's soldiers to take equipment from enemies they kill.
can't see why bannerlord can't make it for all bandit parties but have it happen randomly/upgrade the bandits instead of only changing weapons and change the party name to indicate that it is a more powerful party than normal bandits.
 
Last edited:
It's good to know that lords can actually do something. My assumption was drawn from the fact that hideout numbers add up too much. Mexxico pointed out they do it less often than intended, but it would be disappointing if their solution will be limited at increasing lord intervention. In either case, bandits will still be cheated on the map.

Linking bandit spawn to security is also not a bad idea to take in consideration. My point about prosperity comes from the idea that poverty = banditry, as it is in real life. Poor and desperate people are more inclined to go rogue and make a living from illegal actions. Security can be understood as a variable that prevents these actions from happening. Either way, a formula combining these two values could be implemented to determine spawns.

are you sure that bandits increase with player level? where/how did you discover this?
I remember it from a post on the forums. I can't find it though.
 
I think this sounds interesting. I think the point about spawn rates being tied to prosperity could be a bit different.
  • Maybe more looters, specifically, should spawn when prosperity gets low. Makes sense since looters in lore are peasants turning to crime when there’s a lack of economic mobility.
  • But maybe all other types should work a bit differently? Perhaps there’s a higher chance to spawn one of more hideouts at X prosperity that will result in the “higher tier” bandits that want to prey on the more prosperous towns and related caravans (as others mentioned above). But when a player or lord resolves these specific hideouts, they don’t come back until a new X level of prosperity is met.
And other hideouts not connected to prosperity should still spawn bandit parties that range between Y and Z values that are fixed?
 
1. Bandits are cheating units, while the rest of the parties have to work their way to power in more natural manners
Only the player spends any real campaign time preparing ANYTHING. Bot lords just wander around like zombies until called to a army, yes they magically solve issues sometimes, which is good, but it's in no way comparable to the players campaign time spend to take care and prepare. I wish they did though, they should IMO spawn only recruits and BE FORCED to fight bandits until the power reaches certain threshold via upgrades before doing any raiding villages or joining an army! Can't ever reach it and faction wiped out? too bad so sad! Player needs to solve issues for their fiefs so we need more downtime before defeated enemies can return. They need harsh penalties for failing to pay troops/garrisons and removal of free money too and field loot = money.
BOTS ARE WORSE THE BANDITS!

I agree it's a lazy and dysfunctional mechanic to just poop out endless bandits more and more but I also enjoy and benefit from it :smile:
My clan is strong and gaining strength everyday, all other factions are weakening! I love that bandits are big enough to actual take the bots spawn templates :smile: I savor every "bot was captured by steppe bandits" good riddance!

And I need those big big parties to produce troops because any thing normal sizes will just surrender to me.

And Sometimes I want to just collect hundreds of bandits and stuff them in a garrison!

I wonder though, when you die and a new heir takes over does that reset the bandit size if it's tied to player level?
 
Only the player spends any real campaign time preparing ANYTHING. Bot lords just wander around like zombies until called to a army, yes they magically solve issues sometimes, which is good, but it's in no way comparable to the players campaign time spend to take care and prepare. I wish they did though, they should IMO spawn only recruits and BE FORCED to fight bandits until the power reaches certain threshold via upgrades before doing any raiding villages or joining an army! Can't ever reach it and faction wiped out? too bad so sad! Player needs to solve issues for their fiefs so we need more downtime before defeated enemies can return. They need harsh penalties for failing to pay troops/garrisons and removal of free money too and field loot = money.
BOTS ARE WORSE THE BANDITS!

I agree it's a lazy and dysfunctional mechanic to just poop out endless bandits more and more but I also enjoy and benefit from it :smile:
My clan is strong and gaining strength everyday, all other factions are weakening! I love that bandits are big enough to actual take the bots spawn templates :smile: I savor every "bot was captured by steppe bandits" good riddance!

And I need those big big parties to produce troops because any thing normal sizes will just surrender to me.

And Sometimes I want to just collect hundreds of bandits and stuff them in a garrison!

I wonder though, when you die and a new heir takes over does that reset the bandit size if it's tied to player level?
I don't have an issue with bandits spawning from their hidey holes but I do think there needs to be some kind of a cool down timer on the respawn of bandits from new hideouts imho it should take at least a few days before it begins producing bandits at it's normal rate.
 
I don't have an issue with bandits spawning from their hidey holes but I do think there needs to be some kind of a cool down timer on the respawn of bandits from new hideouts imho it should take at least a few days before it begins producing bandits at it's normal rate.
I think we also need to have some proactive way of preventing bandits so we can keep our fiefs and areas of interest clear while we go off to wage war. Although I don't mind the bandits I don't like that they just appear forever no matter what and you can't create a safe area for you fiefs for a long time or pay for patrols or anything. It be much more interesting if you could really effect how they spawn or cause them to migrate to get away from you and such too, cause some areas to overun while others prosper in safety. Of course I get the idea TW will shake it's head on the player having more control or options in anything at this point.
 
I spend alot of time trying to manage my fiefs properly and it "grieves" me that for all that I've done for them, my people still turn to banditry.

I think high prosperity should lead to less bandits because jobs, jobs, jobs. If the people are content, there's no reason to turn to crime. Agreed that there will always be bandits but the degree should dynamically correlate with prosperity.

Also, groups of bandits should merge together to form a great host with a leader. Some of the wanderers in the game are already criminal in nature so perhaps one of them could become the new leader.

On the subject of poor design, I'm afraid the whole game is riddled with it. Alot of features don't work together or are put in place arbitrarily.

Take for example, armor. It's mind boggling how after so many iterations, lamellar plate boots are actually weaker than suede.

Or the very primitive autocalc function. There are just so many examples that have been documented excessively in this forum.
 
Interesting OP. Just posting here to mark the thread to see any interesting design proposals in my feed.
I have strong opinions, but they are not based on facts, so I can't contribute.
 
I like this idea, one thing I hated about a lot of Bethesda games was the entire world would scale with you rather than specific zones. I think Bandits serve their purpose to impact the world economy which is quite intricate, but it does get annoying that they become *too* impactful in mid-late game and can grind down AI Lord parties, minor factions and pretty much any villager/caravan party they run down.

I'd like to suggest that lower security and high prosperity should be the weights to spawn more Bandits, and bandit parties should only come from hideouts but there needs to be a cap of hideouts per zone with cooldowns. Effectively, it should be possible to totally pacify all bandits for a duration of time. Lower security makes sense due to the (in my head) there won't be many patrols, maybe same thing for lower garrison, it should not ever get to 100% or 0% - but some band between 10-70% (arbitrary)

High prosperity may be harder, but I think a very wealthy city would attract raiders + bandits as well, so there should still be a chance even when security and garrison are topped off for bandits to spawn anyway. Parties should be created and capped per hideout - something like no more than 10 parties per hideout and no more than 3 hideouts per "region" - and to counter the ability for the player to cheese this by just wiping everyone the parties should always continue to consolidate. So if you neglect a region you could arguably see a bandit horde up to 200 members. That would obviously pose an extreme danger similar to the current system, but only due to a ton of neglect (and in war-torn border regions)
 
One opportunity that this global bandit infestation poses is to introduce factions and spawns that counteract them (although never decisively).
Someone proposed manhunter parties, but any bandit-unfriendly minor faction would do. Maybe a knightly order dedicated to law and order, or the military wing of a merchant league, freelancing bounty hunters, whatever fits the lore. This would automatically make the map more dynamic as the player would regularly run into parties fighting bandits and could join one side or the other (with consequences) and some bandit pressure would be taken off lord parties and civilians.
 
one thing I hated about a lot of Bethesda games was the entire world would scale with you rather than specific zones
Exactly what i was thinking when reading!
Maybe a knightly order dedicated to law and order, or the military wing of a merchant league, freelancing bounty hunters, whatever fits the lore.
This is really cool! Maybe the ol' generic Manhunters alongside these new ones would be cool. Or maybe the deserters can be brought back and fight with the bandits as well...
I didn't know that they scaled up with the player level, i'm definitively not a big fan of level scaling, especially in this occasion with so much potential to tie-in with another existing systems.
One thing I'd like would be for the bandits to play a larger role in a player bandit playthrough, with options for the player to use the back alley's, recruit bandits parties, maybe some black-market shenanigans, it could be cool =D
 
I notice often (in 1.5.9) that there will be a large amount of both bandit hideouts and bandits parties around villages, but not any quest to destroy them in any villages. This is kinda lame as at the same time elsewhere I will have multiple "help with brigand" quests in areas that actually don't have that much parties there, making the quest annoying (let alone 2-4 times). I don't know what the game's inner working are but it's pretty crappy to not have these quests in villagers that very blatantly DO have a problem with brigands and DO have MANY "bandit base near..." This, I assume means that NPC lords don't have the ability to clear these hideouts, since they are only resolving issues.

I also notice the "GANGLEADER associate kidnaped" seems to be the most comment hideout quest, but it's actually good to leave it since it gives +security, however I wonder if it blocking other hideout quests from appearing.... but at the same time I know if I clear it another "gangleader ....blah blah" will likely pop up soon and not any other.

Gangleaders suck, TW if you're not adding all the criminal stuff from dev blogs, we don't need gangleader notables and stuff. It's just salt in the wound.
 
Scaling up bandit parties with the player's level is ok, early on. It only becomes a problem when bandit parties become massive, on the order of bandit armies. Taleworlds needs to institute a cap on bandit party size. I think a max size of 30 seems reasonable.
 
This would make sense... I think currently bandit spawns scale up with player level/clan tier? Later in the game there's just too many small bandit parties everywhere, ruining the immersion... I wish there were bigger groups of bandits instead, personally. Since bandits serve no purpose but a way for the player to farm/train his units, I wouldn't implement this system without something extra... I think it would make sense for them to spawn based on prosperity if they also will be able to raid villages and try to raid a settlement if it's really low in garrison/ security. Why else would the low prosperity towns spawn bandit parties if they aren't going to also try to raid and get some money. Maybe even implement a semi-crime system where gang leaders gain more power if the town prosperity is low, and it would be cool if that also affected some other things... Anyway, with such mechanics added I think it would be really great, and since you have an internal mechanic for spawning bandits, and not player level etc, I imagine it won't be too hard to balance either.

Also we need more bandit types, for example looters don't upgrade into any bandit and it's kinda stale to see such basic repetitive parties... Some better equipped looters (maybe even a couple of short bows) would be a nice change of pace.

Scaling up bandit parties with the player's level is ok, early on. It only becomes a problem when bandit parties become massive, on the order of bandit armies. Taleworlds needs to institute a cap on bandit party size. I think a max size of 30 seems reasonable.
I don't agree. Less bandits = no more farming for lords and player, no more training of troops. It's only a problem currently because there's many small bandit parties and they all pile up and they raid caravans together etc. If you had big bandit parties instead, and less small bandit parties, then the big bandit party would be slower and wouldn't cause as many problems. And it's nice to see a big party of looters just waiting to be taken. Late-game you feel like ignoring small party of bandits because they're just not worth your time, when you have over 100 troops to train, so everyone would just auto-resolve battles instead of wasting their time fighting 30 mere looters which pose no challenge at all, so it's really not a good idea. Also, if the bandits would raid villages, which really makes sense, we could finally reduce the amount of annoying lords raiding villages all the damn time (lower their chances to do that - for merciful traits etc at least).
 
Last edited:
I like this idea, one thing I hated about a lot of Bethesda games was the entire world would scale with you rather than specific zones. I think Bandits serve their purpose to impact the world economy which is quite intricate, but it does get annoying that they become *too* impactful in mid-late game and can grind down AI Lord parties, minor factions and pretty much any villager/caravan party they run down.

I'd like to suggest that lower security and high prosperity should be the weights to spawn more Bandits, and bandit parties should only come from hideouts but there needs to be a cap of hideouts per zone with cooldowns. Effectively, it should be possible to totally pacify all bandits for a duration of time. Lower security makes sense due to the (in my head) there won't be many patrols, maybe same thing for lower garrison, it should not ever get to 100% or 0% - but some band between 10-70% (arbitrary)

High prosperity may be harder, but I think a very wealthy city would attract raiders + bandits as well, so there should still be a chance even when security and garrison are topped off for bandits to spawn anyway. Parties should be created and capped per hideout - something like no more than 10 parties per hideout and no more than 3 hideouts per "region" - and to counter the ability for the player to cheese this by just wiping everyone the parties should always continue to consolidate. So if you neglect a region you could arguably see a bandit horde up to 200 members. That would obviously pose an extreme danger similar to the current system, but only due to a ton of neglect (and in war-torn border regions)
I like this concept.
  • Prosperity and security should be main factors for bandit spawn
  • Level based difficulty and spawns should go away, it brakes the mid and late game and promotes bandit farming and grind, this is not fun gameplay
  • Mid and late game the priority shouldn't be randomly hunting bandits but fighting wars and create prosperity, I like the idea of cooldowns, this mechanic would result from a focused effort to fight organized crime, clean hideouts, hunt bandits, increase patrol, even quests to find fight gangs and kill their leaders. Bandits should behave more like a faction rather than a cheating mechanic... simulating number of parties and number of individuals is important...

Edit to the text below: Crap, I assumed that you can calculate the value of items possessed by bandits, I haven't played the game in months, I don't remember exactly if bandits actually acquire new items by attacking other parties, raiding and such. If not, you can't directly apply some things I proposed, maybe simulate it somehow, perhaps on the value of items acquired by leveling..., I don't know.
Bandits should behave more like a faction, items and primary resources should be traceable and be produced with reason, based on prosperity, regional production...

Hmm, it just came to my mind that we are viewing the bandit spawn from the "police" perspective. Maybe seeing it, or rather calculating it, from the "bandits" point of view would be a better option.
Why not have a standard bandit spawn value or maybe even a standard bandit spawn/numbers growth rate... Then, based on this value you would apply modifiers or variables such as profitability of crime, which would basically consist on the value of the resources bandits gather in a region or per hideout/spawn. This variable alone, if I'm thinking right, would take into account prosperity and security, between many others.
  • If a region is prosperous and has low security, bandits would gather highly valuable resources in more quantity, and so, bandit numbers would grow the most.
    • This is also has an auto cap, the number of bandits would eventually stagnate and drop because the number of resources bandits can gather is limited, and prosperity would drop.
    • Having a bandit spawn growth rate based on item values simulates a cooldown effect, where bandit numbers would take some time to increase if low, they first need to acquire items.
    • Also, implementing bandit factions in which they fight themselves would simulate competition, decreasing their ability to focus on new resources gathering, they would just exchange resources between bandit factions (create one spawn per faction, for each spawn calculate total resource value gathering).
      • The incentive (in formulas) to fight each other should be less than the incentive to gather new resources, otherwise they would just constantly fight each other. This would prove a good mechanic because when the prosperity of a certain region decreases and/or security increases, or the growth in the total value of the resources is 0, they would turn to other sources of value (each other).
  • If a region is prosperous and has high security, bandits would gather valuable resources but less frequently, and thus in less quantity, therefore drooping the growth rate of their numbers.
  • If a region is poor and has low security, bandits would gather resources in low or modest quantity and the value of the items would be low or modest, their numbers would grow at a steady rate but not for long since the volume of trade and resources is low. Total number of bandits would also be lower than in high prosperity of regions.
  • If a region is poor and has high security, bandits would not gather many resources and not obtain much value from them, and so, bandit numbers would be the lowest of the other options because the bandit number growth rate is very low.
I revised what I said earlier and the variable "profitability of crime" should/could be the growth rate of the total value of resources possessed by a bandits/ a bandit spawn. And would be the same as the "bandit spawn growth rate", then you could have a standard limit minimum growth rate so that you couldn't totally eradicate them, or if the number of bandits reaches a minimum, ignore the growth rate if 0 or negative and maintain the minimum value of bandits, or give a boost to the growth rate beyond a certain value.

Kingdom/Faction/Clan policies, personal traits, perks and rvents could affect the bandit growth rate in a specific region.

The growth rate could be calculated once per day. The spawn number could be the whole number of the result of the growth or the rounded number.
E.g.:This doesn't account for bandit deaths
Day​
1​
2​
3​
4​
Total value of possessed resources​
1000​
1010​
1144​
962 (some party with 15 died)​
Bandit number growth rate​
0​
1010-1000
1000
= 0.01 or 1%​
1144-1010
1010
=~0.133, 13.3%​
962-1144
1144
=-15.9%​
Total Number of Bandits​
100​
100*(1+0.01)=101
spawn 101-100=1​
101*(1+0.133)=114.433
spawn 101-114=13​
Negative growth
= 0 spawns
Total
114-15= 99​

The growth rate can be easily adjusted, increase/decrease effect.
E.g.:
normal 349*(1+0.12) = 390.88;
decreased 349*(1+0.90*0.12) = 386.692
increased 349*(1+1.10*0.12) = 395.068

The number of bandits that spawn could be spread during time instead of just popping up. Spawn the computed number of bandits in x day, during different times of the day, during several days, in a single party, by adding up members to existing parties...

I realized that maybe by altering numbers of bandits based on growth rate of the value of total resources possessed by bandits might create a problem when the growth rate is 0 or negative for a long period of time, when you are actively killing bandits for example. During that time there would be no bandit spawns. (Hmm, but when you stop pursuing bandits they also stop running and return to "work", also, while they are running for their lives prosperity increases because they cause less harm to the economy.)

(if the above is a problem) To solve this there could be a standard growth rate based on the prosperity of the region (this would substitute the need for an arbitrary minimum growth rate that I mentioned earlier), it would simulate the sense of opportunity bandits have and the potential profitability of crime.
It would require some testing to figure out the exact formula.

Edit to the text below: I just realized that this gives basically the same result as calculating growth rate. But you can still use this as an alternative to compute, also the concept underlined is still usable.

I also thought of an alternative/addition to the growth rate of numbers. What could be done is figuring out a the amount of value necessary for one particular bandit to spawn. Then, from the total value of items bandits possess, it would be determined the number of new spawns.
Different types of bandits would have different value needed to spawn. Different regions could have lower value needed per bandit and favor certain bandit types based on the lower price and probability, which should be based on the lore.

E.g.: Admitting all types of bandits have the same value needed to spawn.
Day1234
Total value possessed by bandits or prosperity, as suggested in text below50 00054 00061 33368 745
Value needed to spawn one bandit500500500500
Number of bandits100108; spawn 8122.666; spawn 14137.49; spawn 15

Again, the number of spawns could be spread in time instead of single moment.

Based on the system above you could also calculate the bandit spawns based on "prosperity" instead of the "total value possessed by bandits" (if it is even possible to calculate). Only basing it on prosperity might not account for security or the effort you put on cleaning up bandits.
A complementary variable to this option is also basing the spawn growth rate on bandit death rate, this would simulate security.
If death rate numbers are higher then it means there is some action in reducing crime in the area. It could be used to reduce bandit growth rate in the respective region.

I hope that any of this can help make some change.

In a real world scenario, crime has many causes. Searching real world data could help finding the right formula.
Social-disorganization-causes-of-crime-a-selection-of-factors-Source-Reconstructed-by.png
12117_2007_9013_Fig3_HTML.gif
40503_2019_81_Fig3_HTML.png
The-crime-rate-imprisonment-rate-and-GDP-per-capita-at-1990-price-of-the-United-States.png


For more info:
 
Last edited:
Back
Top Bottom