The largest stable companion parties in With Fire and Sword (spoilers)

[EDIT:  UPDATED THE FIRST POST TO COMPLY WITH THE NEW NPC INTERACTION MATRIX. THE REPLIES UP TO #12 ARE BASED ON THE OLD INTERACTION MATRIX.]

Designing the largest possible stable parties is a nice little mini game in Warband. The game has an elegant system of limiting the number of npcs present in the player’s party: the npcs do not get along with all the other npcs which means that the player has to make choices about whom to include in the party. The choices depend partly on how stable the player wants the party to be, i.e. how resistant the party is to low morale.

We start by giving some definitions:
a stable party  is group whose all members either
[list type=decimal]
[*]have no person they dislike in the group, or
[*] have at most one disliked person in the group while also having a person that they like
[/list]

a strictly stable party is a stable party whose members have no one they dislike in the group (equal to condition 1 of stable party).

A strictly stable party has the best resistance against low morale and the fastest speed on campaign map. A stable party that is not strictly stable is not quite as resistant against low morale but is still quite acceptable.

Any party with a member who either
[list type=decimal]
[*]doesn’t like another party member and has no friend to compensate, or
[*] has at least two persons that she doesn’t like in the same party
[/list]

is called unstable. Unstable parties are also possible to keep together but are much more fragile when faced with adverse morale effects, with unhappy members often trying to leave the party. Since they tend to have lower morale they also suffer from slower campaign map speed.

In regular Warband dislikes and likes were mutual, i.e. if A liked B then B liked A back. This is no longer the case in WFAS. With the new non-reciprocal system of likes and dislikes building the largest stable parties in WFAS is a much more interesting problem than in regular Warband.

The method and source data
I wrote a short program that reads in the relationships between the different NPCs and finds the largest stable parties. Since some might consider relationship information a spoiler I’ve put them behind spoiler tags.
The source data fed to program is the relationship matrix I dug up from the game’s .txt files:


see reply #13 for more details.

Note that at the moment of writing this differs from the data given by the Mount and Blade Wiki:

Only additional information used is that you can’t have both Mamai and Zagloba in the party at the same time (they have different storylines, and for generating parties that are suitable for frequent looting I used the wiki information about the NPCs who dislike robbing villages:

The results:

Disclaimer: the results depend on the correctness of the source relationship matrix and other source information. I accept no responsibility for any possible errors due to erranous source information or for any reason whatsoever. This was done purely for fun so use at your own risk:

Strictly stable

The largest strictly stable parties without Zagloba or Mamai:


you can freely add either Zagloba or Mamai to any of these parties, so in essence you have 8 possible 7 member strictly stable storyline parties:


Even strictly stable parties are not all equivalent when it comes to party morele. The absolute best in terms of morale are those who have most members with morale bonus from having a friend.

Stable

The largest stable parties:


There are 11 stable parties that have 9 members. All of them have either Mamai or Zagloba as a member so they are only accessible if you are doing either the Polish or the Cossack storyline, 6 choices for the Polish and 5 for the Cossacks.

If you are not doing either of these two storylines you can have at most 7 NPCs in a stable party (11 different parties):

Having Mamai or Zagloba allows for larger stable party as in addition to these storyline npcs it’s now possible to pull in extra members who like them!

If you insist on building a “no-qualms about robbing innocent villagers”-party your maximum stable size will drop to 7 for the Cossack storylines, to 6 for the Polish storyline and to 5 for others:

The largest stable Cossack raiding parties:

The largest stable Polish raiding parties:

Stable raiding parties for others:
NOTE: any subgroups of the stable groups are not necessarily stable themselves; this is because you might need to have all the members before the necessary friendships exist. This behavior differs from the strictly stable groups: any subgroup of a strictly stable group is itself strictly stable, so they are easier to build as you don’t need to worry about members leaving before you find the missing members.

Cheers,
Lohi

You know, I actually have no trouble keeping all 15. I used to have all 14, but not Zagloba or Mamai. Then I got Zagloba, as I always play for the Poles (despite being a Russian and Ukrainian ). Before Zagloba, Bakhyt and Oksana tried to leave. After Zagloba, Oksana took a liking to him, and stayed. I have high persuasion and maintain a high party morale. Looting villages is too much like cheating to me, as you get too much crap, too easy. So I manage to keep all 15 companions. Whenever Bakhyt leaves despite my persuasion, I simply reload. I save very often, so it's no problem for me.  Everyone else is sorta neutral about staying in my party, and they never speak up. However, if anyone else wants to do the same, but doesn't like to reload, I suggest you take up the Cossack storyline, so you can get Mamai. Bakhyt is the best companion in terms of Strength, so he is a very valuable fighter. Meanwhile, Oksana is fairly useless.


This is how I separate my party skills, BTW, which took me a few days of meticulous research:
Yelisey as both Tactics and Engineering companion. He starts out with the highest engineering, so he can double as the tactics guy (Intellect, after all, give you two point per level up, which is more than enough to advance two skills at once). I NEVER use Victor de la Buscador because he has high strength, which means that he can be a fighting companion. Yelisey, on the other hand, is two strength levels weaker (8 versus 10 I believe), so he is perfect for the role of a tactician/engineer.
Sarabun as the Surgery, Wound Treatment and First Aid companion. He's pretty self-explanatory, as he is the best doctor out of all the companions.
Ingri as the Trading companion. Again, self-explanatory, as she has the highest trade skill, and she is too weak to be of use elsewhere (7 strength). In OiM she use to start out with 7 Trading skill. That has changed in WFaS though...
Fatima as the Looting companion. She has the highest Agility, so that's why. And she is average at combat - starting out with the standard 8 strength.
Fedot as the Tracking, Spotting and Pathfinding companion. Again, a pretty easy choice, because he is simply the starting best at those skills.

And that is it, I believe.

I really admire this post, it's a great idea Lohi. I'm actually surprised someone hasn't posted this sooner. Kudos!

Aemilius Paullus said:

So I manage to keep all 15 companions. Whenever Bakhyt leaves despite my persuasion, I simply reload. I save very often, so it's no problem for me.  Everyone else is sorta neutral about staying in my party, and they never speak up. However, if anyone else wants to do the same, but doesn't like to reload, I suggest you take up the Cossack storyline, so you can get Mamai.

Click to expand...

<interesting analysis of companion roles snipped>

Aemilius Paullus, yes with persuasion and without military catastrophes it is possible to keep much larger NPC groups. However I (and many others) are interested in having a robust group that you don't need to persuade and which can take setbacks. That's why I defined what I meant with 'stable' in the beginning of my post: in essence the above party configurations should not result in anybody wanting to leave just because of the company they are in.

Of course you might still  need to use persuasion occasionally, for example if you face repeated disasters and insist on doing things that p*ss them off. Also some of the party configurations I listed above might not be stable until you get them all! 

Pendy, I'm glad you found it useful.

Hah, I was going to try and work this out by hand to practice for the LSAT, but you've taken away all my motivation. 

This list couldn't be accurate because Mamay dislikes Priest. 


The list without Mamy / Zagolobada may be accurate, though.

Do you have a list of the maximum number of companions with no dislikes at all? The only one I've personally confirmed is:

Sarabun
Priest
Fatima
Yelisei
Karlson
Fedot
Zagloba (when doing Polish story quests)

In Warband, I always preferred to have 0 dislikes in the party so that I didn't have to worry about anyone leaving if morale started to tank (from lack of battles, mostly). It seems a little easier to keep morale high in WFaS though, so having a dislike here and there may not be much of an issue.

Schmon said:

This list couldn't be accurate because Mamay dislikes Priest. 

Click to expand...

Are you 100% sure Schmon? As you can see, the source data that I used only had Priest disliking Mamay, not the other way round.

Wheem said:

Do you have a list of the maximum number of companions with no dislikes at all?

Click to expand...

No, but that would be easy to do, I'll just modify the source data to have no likes so that the program can't use them to compensate for dislikes. I'll have a look later, when I have something faster than a cell phone in my hands.

In Warband, I always preferred to have 0 dislikes in the party so that I didn't have to worry about anyone leaving if morale started to tank (from lack of battles, mostly). It seems a little easier to keep morale high in WFaS though, so having a dislike here and there may not be much of an issue.

Click to expand...

I found that isolated dislikes compensated by likes for the same characters have not resulted in anybody leaving, even in a period of inactivity. But I guess if you had a long period without battles and also happened to fail a quest  etc. even more robustness could be useful.

Wheem said:

Do you have a list of the maximum number of companions with no dislikes at all? The only one I've personally confirmed is:

Sarabun
Priest
Fatima
Yelisei
Karlson
Fedot
Zagloba (when doing Polish story quests)

Click to expand...

Isn't Priest one of Zagloba's dislikes? I wish somebody would do as good a list of likes and dislikes for WFAS as xyp had done for the Brytenwalda mod.

Essay said:

Hah, I was going to try and work this out by hand to practice for the LSAT, but you've taken away all my motivation. 

Click to expand...

Sorry for that ...I waited quite long that somebody would do it, but in the end decided to "waste" one evening myself. The program was rather easy to write but I find it hard to verify which source data has all the relationships (who likes who etc.) correct.

[edit: somehow I didn't notice that you planned to do it by hand!]

Wheem said:

Do you have a list of the maximum number of companions with no dislikes at all?

Click to expand...

Here are the largest strictly stable (i.e. not even compensated dislikes allowed) party configurations based on the Filou's relationships matrix:


so 6 seems to be the maximum possible size for the strictly stable parties if the source information is correct.

Lohi-soturi said:

Wheem said:

Do you have a list of the maximum number of companions with no dislikes at all? The only one I've personally confirmed is:

Sarabun
Priest
Fatima
Yelisei
Karlson
Fedot
Zagloba (when doing Polish story quests)

Click to expand...

Isn't Priest one of Zagloba's dislikes? I wish somebody would do as good a list of likes and dislikes for WFAS as xyp had done for the Brytenwalda mod.

Click to expand...

I never had Zagloba say anything about any of my companions, though I didn't have that game going for too long (had to reformat recently because I foolishly tried a Windows repair install from an old disc with a previous version service pack ).

Lohi-soturi said:

Wheem said:

Do you have a list of the maximum number of companions with no dislikes at all?

Click to expand...

Here are the largest strictly stable (i.e. not even compensated dislikes allowed) party configurations based on the Filou's relationships matrix:

Spoiler

Code:

[,1]	[,2]	[,3]	[,4]	[,5]	[,6]	 [,7]	 [,8]
Tepes	Tepes	Tepes	Tepes	Viktor	Viktor	 Sarabun Varvara
Ingri	Ingri	Ingri	Ingri	Varvara	Varvara	 Fatima	 Fatima
Bakhyt	Bakhyt	Viktor	Viktor	Fatima	Fatima	 Priest	 Priest
Viktor	Viktor	Zagloba	Zagloba	Priest	Priest	 Karlson Karlson
Zagloba	Mamay	Nogai	Fatima	Fedot	Yelisei  Fedot   Fedot
Nogai	Nogai	Fedot	Fedot	Yelisei	Algirdas Yelisei Yelisei

so 6 seems to be the maximum possible size for the strictly stable parties if the source information is correct.

Click to expand...

Thanks, though it seems really limited if all the source info is correct. It feels like one "needs" a doctor, and Sarabun has by far the highest starting medical skills but there's only one available no-complaining setup with him in it. And that group also doesn't allow for either of the story line companions.

So it might be best to just put up with a little bit of bickering, and try to keep party morale from slipping too much (and/or invest in a bit of persuasion).

Wheem said:

I never had Zagloba say anything about any of my companions

Click to expand...

Wheem’s post about Zagloba not complaining about Priest made me doubt the correctness of the source data that I had used. Since there seemed to be no reliable information about NPC interactions in WFAS available I decided I need to do some datamining with the game .txt files to find  out what the companions feel about each other. It was a pain to figure out how to decipher the files but armed with the excellent Mod Viewer .NET GUI by themendios and my trusty text editor I think I managed to figure it out after a few hours toil:




In short: the relationship matrix I had found on the Paradox forums:



seem to have been a little bit too restrictive as Zagloba and Mamai actually don’t seem to have any dislikes (and nobody seems to dislike them!) and also Priest seems to dislike no-one. This means there are bigger stable groups available than what the previous data led to believe. I will soon post the new info..

Cheers, LS.

hello,

I've just updated the first post to reflect the new interaction matrix.

This is fantastic. Thanks!