Plz fix broken simulation! 19 cavalry vs 7 looters one heavy cavalry die!

Users who are viewing this thread

It was better than Native, but when you look at the formulas, you see many arbitrary values and tweaking them is just "magic".
In retrospect it's better to assign some kind of empirical strength values to troops (ditch the troop strength formulas) and then try to come up with a battle formula that also takes into account numbers and terrain and whatever else.
The empirical strength values could be arrived at by running real, 3D 1-on-1 fights in some kind of a troop league (in development, not in a released game). Something like ELO ratings for troops. :smile:
That makes sense
 
If we could select which troops to send in auto calc it would make it much more useful. I'm fine accepting that ramming 30 recruits into 30 looters, some recruits die, I don't want more advanced troops (who I don't need to level on looters anyways) to have chance to die randomly. Yes you can circumvent this by taking stacks of just recruits (or whatever) you want to ram into looters, but it would be more convenient and *GASP* USEFUL to choose, as typically you're getting a mixed bag of troops as you waddle around your many estates piling everyone into the murder stack. You often don't want troops going above tier 2 or 3 anyways, but 1 is a waste of food and just archer fodder.

I agree with other's feelings that if something is in the game it needs to be useful in some way.
 
Having auto-resolve as the inferior option is probably fine (since you ain't putting the legwork in), but the fact that it's so unpredictable makes it unviable.
It isn't unpredictable though. If you're even (eyeballing here) 5 or 10% superior on the power bar, you'll win virtually every single time. The thing autocalc isn't is safe because you can get wacky outlier results like the seven looters managing to kill a T5 horse archer among a 50 man party when they couldn't manage that in 10,000 live battles.
 
It isn't unpredictable though. If you're even (eyeballing here) 5 or 10% superior on the power bar, you'll win virtually every single time. The thing autocalc isn't is safe because you can get wacky outlier results like the seven looters managing to kill a T5 horse archer among a 50 man party when they couldn't manage that in 10,000 live battles.
[Wacky outlier results] ~ [unpredictable] in that context. Naturally, a T5 HA party will dispose of any potential amount of looters in both real battle and auto-calc, but there's absolutely no way to lose a T5 HA in RB (unless the player actively tries to get them killed via hold position + hold fire and letting looters throw deathstones), where as in Auto-calc batles, it's very much possible.

It ain't the ultimate evil and the root of all issues in the game, far from it. However, it makes auto-calc unreliable in situations it's specifically designed for
 
I guess something like this occured :
Due to the overwhelming difference between his army and its opponents, the army commander felt unecessary to show up on the battlefield.
Let my troops deal whith those petty looters, he thought.
The army advanced towards the looters
An elite cataphract, filled with a legitimate sense of superiority, and in the absence of his commander, broke off the main line and charged bravely at the looters.
Alas, by a sheer struck of luck, one of the stones thrown at him by the looters landed on the right temple his helmet. The speed of the stone being added to the opposite speed of the horse charging, rendered the blow so hard that the cataphract passed out and fell of his horse.
It was not long before the looters cut the throat of the unfortunate horseman laying on the ground. Another Goliath slained by a seemingly feeble Samson David.
Corrected, I have to study my bible stories

Improbable? Of course.
Possible ? Absolutely!

Saddened by this loss, the commander decided next time he would watch his troops more closely :wink:
Really entertaining, thanks! (y)
 
The system seems broken. In a mixed army I always loose tier 6 units like palantine guards. Looters would never come close.
Also if i decide to fight the battle, i just set all my troops to charge, and I never loose anything. But if I autoresolved it, I would have lost units.
 
I had a party off over 100+ Elite horse archer's and yet the auto reslove will still find a way to get some of them killed. Even if you told your horse archer's to hold ground and fire, there would be no chance in hell 50+ looter's would be able to get into range, to even throw a rock, before they (Routed) or straight up (died). The only reason to auto-resolve is used for, is so that you don't have to keep fighting small groups of units. But it doesn't work and you still have to fight them. A thread was made about this when the game first came out, and they haven't really done much and that was over a year ago.
 
I guess something like this occured :
Due to the overwhelming difference between his army and its opponents, the army commander felt unecessary to show up on the battlefield.
Let my troops deal whith those petty looters, he thought.
The army advanced towards the looters
An elite cataphract, filled with a legitimate sense of superiority, and in the absence of his commander, broke off the main line and charged bravely at the looters.
Alas, by a sheer struck of luck, one of the stones thrown at him by the looters landed on the right temple his helmet. The speed of the stone being added to the opposite speed of the horse charging, rendered the blow so hard that the cataphract passed out and fell of his horse.
It was not long before the looters cut the throat of the unfortunate horseman laying on the ground. Another Goliath slained by a seemingly feeble Samson David.
Corrected, I have to study my bible stories

Improbable? Of course.
Possible ? Absolutely!

Saddened by this loss, the commander decided next time he would watch his troops more closely :wink:
lmao... thank you for this. yea i usually save the game before even fighting looters so if the autoresolve harms my numbers, RELOAD!
 
The system seems broken. In a mixed army I always loose tier 6 units like palantine guards. Looters would never come close.
Also if i decide to fight the battle, i just set all my troops to charge, and I never loose anything. But if I autoresolved it, I would have lost units.

You're more likely to lose troops to looters if you set the all to charge. Because that one guy with a fork is going to stop one of your horses in their tracks, making them vulnerable to a pile on for a moment or two - which is really the only way looters can kill a high tier unit (I've even had it happen to me).

The auto resolve doesn't really take into account the things a human player would do for troop safety, such as stand still and not issue any commands and let your archers kill the 3 looters required for them to flee.
 
It ain't the ultimate evil and the root of all issues in the game, far from it. However, it makes auto-calc unreliable in situations it's specifically designed for
The autocalc isn't meant for the player to avoid fighting (in live battles, I mean) small groups of bandits. For a long while there was 'feature' of it that would try to guarantee you'd lose troops to small groups of bandits, even if you grossly outnumbered them. It was removed. The earliest iteration -- where autocalc looters would never kill your troops -- was also patched out as a bug.

TW apparently wants to punish the player who autocalcs those fights. If you say that doesn't make sense and just serves to make the game feel especially grindy -- I don't disagree.
 
The autocalc isn't meant for the player to avoid fighting (in live battles, I mean) small groups of bandits. For a long while there was 'feature' of it that would try to guarantee you'd lose troops to small groups of bandits, even if you grossly outnumbered them. It was removed. The earliest iteration -- where autocalc looters would never kill your troops -- was also patched out as a bug.

TW apparently wants to punish the player who autocalcs those fights. If you say that doesn't make sense and just serves to make the game feel especially grindy -- I don't disagree.
If autocalc results were "realistic" (close to a real outcome), the player would have little incentive to fight battles live, as mistakes could make battle results worse than autocalc's. So the results must be somewhat worse than live battles. How much worse is a matter of debate and balance, but don't open new exploits.
 
150 tier-5 crossbowmen versus 20 looters… If I take the field—even if I click "attack" and then go AFK for the entire battle—there is literally zero chance of any of the looters wounding (much less killing) any of my troops. The looters will be erased long before they can do anything meaningful, even with my total absence of input. But with 'simulated battles,' it isn't uncommon for one of those crossbowmen to be killed.

The OP is right; 'simulated battles' are simply broken. Anyone who says differently is selling something.

If autocalc results were "realistic" (close to a real outcome), the player would have little incentive to fight battles live, as mistakes could make battle results worse than autocalc's. So the results must be somewhat worse than live battles.
This may be true, but it shouldn't be an issue in "large numbers of elite troops versus tiny number of scrubs" battles. As others have pointed out, making autocalc dysfunctional in order to coerce players to manually fight every band of looters turns it into a grind game. There's no benefit to anyone to spread the game over too much grind, like Bilbo's butter.
 
Last edited:
I had a casualty = game is broken.
The purpose of the autoresolve feature is for the player to avoid sitting through boring, unfun battles they already know the outcome of.

However, if actually using autoresolve punishes the player by killing troops that definitely would not have died in a non-autoresolve battle, then the player will not want to use autoresolve for its intended purpose.

And what we find is that a huge amount of players do not want to use the autoresolve feature because of this. Therefore, the feature is broken.
If autocalc results were "realistic" (close to a real outcome), the player would have little incentive to fight battles live, as mistakes could make battle results worse than autocalc's. So the results must be somewhat worse than live battles. How much worse is a matter of debate and balance, but don't open new exploits.
@MadVader while definitely true, and there should be some risk in any battle that's remotely close, there should be no risk involved in a curbstomp battle of highest-tier units vs. lowest-tier units.

If you have 50 T1 vs. 50 T1 and autoresolve it, you should have an equal chance of losing the battle and a high chance of wounded/injured (influenced by tactics and medicine ofc).

But if you have 50 T6 vs. 50 T1 in autoresolve, you should have an incredibly low likelihood of losing even a single unit. We all know how the battle would go, and the player has to sit through an enormous amount of battles like that. Just let us skip it without being penalized.
 
Last edited:
The purpose of the autoresolve feature is for the player to avoid sitting through boring, unfun battles they already know the outcome of.

Is it?

And what we find is that a huge amount of players do not want to use the autoresolve feature because of this. Therefore, the feature is broken.

Meh. You win some, you lose some. All war is a gamble. If you wanted a predictable outcome, you should be watching a movie.
 
Explain what other reason the autoresolve exists for than for the player to skip uninteresting battles they don't want to sit through.
Meh. You win some, you lose some. All war is a gamble. If you wanted a predictable outcome, you should be watching a movie.
Do you think manually fighting a battle of 100 cataphracts vs 10 looters is going to have an unpredictable outcome?

No. And that's exactly why the player should be allowed to skip battles like that without losing troops. Precisely because we don't want to deal with predictability for so long.

But if we can't autoresolve a curbstomp battle without losing troops, we're forced to deal with 5 minutes of predictability, as opposed to 1 second of predictability. Do you get my point?
 
Explain what other reason the autoresolve exists for than for the player to skip uninteresting battles they don't want to sit through.
IMO it's just a out of place inherited thing from warband and it's too bad an entire skill is wasted on it. Basically it has no point, it's just garbage. The only use is to circumvent other stupid bad designs, such as" looters rout too soon in live battle and recruits can't catch them" or "I perposfully want more wounded troops for medicine". This is 100% not intended by TW though, it just happens to have a use because of these other ****ed up parts of the game. IMO if it's in the game it should be made to be useful for the player and should in fact be changed for reliably defeating weak parties with less time wasted or at least let the player choose who to send into the battle.

See in warband if your leader was KO'd the battle abruptly ended and you had to use auto calc to finish it. Fighting with the leader KO'd was a big bad punishment for the player and had famously bad outcomes. You ****ed up, you're taking losses now, try harder/smarter next time.

In bannerlord, you don't get kicked out of battle when the leader is KO'd, you can finish it with live combat, albeit with the questionable bot tactics...
Auto calc is 'presented' as an actual option because has a skill tree devoted to it, which is frustrating because in fact it always safer to do live battle.
In short, left over 'punishment' feature even though the normal punishment is removed.....
 
It's a convenience feature, and isn't uncommon in other similar games...

The Total War series has had an auto calculation feature since at least Rome 1 from 2004. As people have mentioned above... In the early game, it's helpful to fight every battle, but as you build your strength, every battle becomes a micro nightmare so later on, it provides a convenient way around the grind.

But, as others have also noted, it must have risk, otherwise it just becomes a lazy exploit that kills the game. You can have your short cut, but you have to be prepared for the possibility that it might cost you. In the Total War series it became an ongoing joke. Because it's a calculation with probabilities, murphy's law said you'd always lose the troops you didn't want to lose and be left with the junk you didn't mind losing.

It's also exploitable. I've recalculated/reloaded an auto-siege over and over to see the spread of results. You end up with a bell curve of casualties - the majority of results clustered around the same place, but occasionally you end up with a result well outside the bell curve, even better than you could get if you managed the battle yourself.

I'm sure the devs could tinker with it and change the probabilities, but it seems to me like a pretty mundane thing to waste their time on. It might not be perfect, but it certainly isn't broken. Th rest is subjective.
 
The case of elite units dying in objectively safe and overwhelmingly favorable auto-calc battles has been a meme since long before Bannerlord.

uqqy7aj8v3g21.png

n4kkpb3l74c21.png


Though in this instance, I actually like that auto-resolving battles is dangerous and unpredictable, because it promotes active gameplay from the player. There is a high chance you'll be punished for using auto-calc, so instead of just simulating the best part of a Mount&Blade game (the combat), you are incentivized to take an active role in the struggle.

And that is good game design.

Auto-calc should remain how it's been for over a decade: A very unwise risk taken by the player, often resulting in disaster, thus prompting them to take a hands-on role in future fights. And a source for funny memes.
 
Though in this instance, I actually like that auto-resolving battles is dangerous and unpredictable, because it promotes active gameplay from the player. There is a high chance you'll be punished for using auto-calc, so instead of just simulating the best part of a Mount&Blade game (the combat), you are incentivized to take an active role in the struggle.

And that is good game design.
The best part of Mount & Blade is not the 500th time in a campaign that you've steamrolled a party with less than 5% of your strength. TW might imagine it encourages active gameplay but it doesn't -- it encourages people to **** right off and play something else that's less out-and-out grindy.

That's because TW doesn't actually play this game, so they think a number of things are fun when they really aren't.

It isn't even like it actually forces the player to do a damned thing except wait a minute or two to resolve the fight 'live.' You can literally hit "Join your troops" and tab out, then come back when you hear cheering and (assuming your party has the bare minimum necessary) never take losses against parties you greatly outpower. It is 100% a waste of time that could be saved with an autocalc that had a built-in threshold for party power required to cause casualties.

Of course, that ties back into the intended playstyle (mixed quality parties, semi-frequent wipes, etc.) vs. the typical player playstyle (max-tier, max size party, never losing, few casualties). You can get by never fighting looters after the very start of the game, without caring too much about your troops' survival and (IMO) that is a much more fun way to play but it isn't by any means the most common way to play Bannerlord.
 
Last edited:
The best part of Mount & Blade is not the 500th time in a campaign that you've steamrolled a party with less than 5% of your strength. TW might imagine it encourages active gameplay but it doesn't -- it encourages people to **** right off and play something else that's less out-and-out grindy.
This was a problem with Warband already.
However, now we have the surrender mechanics, which SOLVES the problem of trivial battles, though I don't know if it covers all the overwhelming odds cases and it should work for nobles too.
 
Last edited:
Back
Top Bottom