Factional Alignment system: Creatures

[Keldaryth]

Well more or less. One thing though you now have 2 points in the same alignment system, how should they influence each other, and interact with the players alignment?

No, I don't.

I have a creature alignment which is used to determine player alignment

I then have an alignment coefficient (well, 2 or three) which determine how the attractiveness of a player's dungeon to new potential creatures of that type changes with changes in the player alignment. See this post with game mechanics examples for details on how this works.

[andrewbuck]

Just to clarify and give my understanding of this, which may or may not be correct, the system you laid out is...

Each creature has an alignment which are the 2 or 3 dimensional vectors you list (depending on which dimensionality we go with one or the other will be used). You then take a weighted average of the creatures a player controls using a set of coefficients as the weights (and possibly some "modifiers" for various actions), and use this to determine the player's alignment. Then, given the player's alignment, you are more likely to draw creatures whose alignment closely matches yours and less likely to draw creatures farther away (and by draw I mean what kind of creatures show up through the portal).

If this is basically what you are saying I think you have a pretty clever system and I think it would work pretty well. I think the only thing I would add (and you may have already planned on this) is that when you join a game you should pick a faction from some kind of menu which gives you a small number of creatures (probably a pre-determined set for each faction choice to keep things fair). This will bias you in the direction you want to go right at the beginning so you don't have to through back a bunch of creatures if you don't get the draw's you are looking for while you are in the neutral alignment.

-Buck

[Keldaryth]

Hi Buck - you're close, but I've made it a little more detailed:

You take an unweighted average of the alignments and factions that a player controls to determine a players alignment (and faction factors). You then modify the base probability that any given creature will appear by taking using the creature's alignment (and factional) coefficients to determine how much that alignment/faction mix entice or repel that creature.

Based on a players alignment you are more likely to draw creatures who enjoy living in a faction/alignment mix that you the player have created - however, the creature's faction/alignment may not be close to the current mix, which can lead to some pretty interesting creature lists.

In terms of starting a game there are a number of options:
+ A keeper starts with no alignment: this makes the game truly random and their gameplay will be dictated by the first creature to enter their realm.
+ A keeper can select a certain background or set of backgrounds which can have an effect on alignment, creature availability, spell effectiveness etc.
+ A keeper can select a certain set of starting material in terms of a combination of rooms and creatures, which will affect starting alignment
+ A combination of the middle two

I think we should also allow for a completely neutral 'draw' as well because it's an additional challenge that will add replayability to maps (Think HOMM where on some maps you can pick your starting town or just play randomly). If you've played the original majesty, option 3 is very much like their multiplayer/random maps, where you didn't HAVE to make things fair - to even things out amongst friends who weren't playing competitively they ranked the starting options as 'beginner, advanced, expert and master' so you had a rough indication of the difficultly of the start, so an experienced player could select a harder start to make for a more even game. Although the hardest start we have here would be completely random

One thing I would like to implement is certain rooms affecting alignment - torture chambers tending towards egotism and violence (or chaos and evil) for example, or certain rooms/monuments giving a factional bonus.

[Keldaryth]

Dungeon 4

So, just for fun, I'm going to mimic the process of attracting 15 creatures to a fully functional dungeon, using the two axis alignment - i.e. every creature you could attract is available (24 on the current list). This gives each creature a base of of 0.042 or a 4.2% chance to appear.
Starting with 3 kobolds, the first randomly generated creature is a Troll Shaman.

Keeper attributes after the appearance of the Troll Shaman:
Human 0 | Corpar 0.25 | Construct 0 | Undead 0 | Denizen 0.75
2 Axis Alignment Attribute: LC 0.125 | GE 0

At this point, the following creatures will not appear in the 2 axis alignment dungeon: Necromancer, Rogue, Maiden, Miner, Knight and Wight.
Creature most likely to appear: Minotaur/Minotaur Warlord
Creature least likely to appear: Warrior/Wizard

The next creature to appear is the Steel Construct

Keeper attributes after the appearance of the Steel Construct:
Human 0 | Corpar 0.2 | Construct 0.2 | Undead 0 | Denizen 0.6
2 Axis Alignment Attribute: LC 0.3 | GE 0

With a Child of the Deep and a Construct present, the following creatures will not walk into your dungeon: Wizard, Necromancer, Wraith, Rogue, Orc, Wight, Ghoul and Vampire. Interestingly the boost to the LC attribute has meant the Maiden, Miner and Knight will consider joining you.

Creature most likely to appear: Any Construct
Creature least likely to appear: Maiden

The next creature to appear is: Miner
Keeper attributes after the appearance of the Miner:
Human 0.17 | Corpar 0.17 | Construct 0.17 | Undead 0 | Denizen 0.5
2 Axis Alignment Attribute: LC 0.3 | GE 0
Creatures who will not appear: Goblin, Necromancer, Wraith, Rogue, Naga, Orc, Wight, Ghoul, Vampire.
Creatures who will now consider appearing: Wizard
Creature most likely to appear: Miner
Creature least likely to appear: Wizard

Seventh creature to appear: Stone Golem
Keeper attributes after the appearance of the stone golem
Human 0.14 | Corpar 0.14 | Construct 0.29 | Undead 0 | Denizen 0.43
2 Axis Alignment Attribute: LC 0.43 | GE 0
Creatures who will not appear: Necromancer, Wraith, Rogue, Naga, Orc, Wight, Ghoul, Vampire.
Creatures who will now consider appearing: Goblin
Creature most likely to appear: Any Construct
Creature least likely to appear: Goblin

Eighth creature to appear: Warrior

Keeper attributes after the appearance of the warrior
Human 0.25 | Corpar 0.125 | Construct 0.25| Undead 0 | Denizen 0.375
2 Axis Alignment Attribute: LC 0.475 | GE 0.05

Creatures who will not appear: Goblin, Necromancer, Wraith, Rogue, Naga, Orc, Wight, Ghoul, Vampire.
Creature most likely to appear: Miner
Creature least likely to appear: Wizard

Ninth creature to appear: Minotaur

Keeper attributes after the appearance of the minotaur
Human 0.22 | Corpar 0.22 | Construct 0.22| Undead 0 | Denizen 0.33
2 Axis Alignment Attribute: LC 0.49 | GE 0.04

Creatures who will not appear: Goblin, Necromancer, Wraith, Rogue, Naga, Orc, Wight, Ghoul, Vampire
Creature most likely to appear: Miner
Creature least likely to appear: Wizard

Tenth creature to appear: Maiden

Keeper attributes after the appearance of the maiden
Human 0.3 | Corpar 0.2 | Construct 0.2| Undead 0 | Denizen 0.3
2 Axis Alignment Attribute: LC 0.44 | GE 0.13

Creatures who will not appear: unchanged
Creature most likely to appear: Miner
Creature least likely to appear: Wizard

Eleventh creature to appear: Troll Stone Priest
Keeper attributes after the appearance of the Troll Stone Priest
Human 0.27 | Corpar 0.27 | Construct 0.18| Undead 0 | Denizen 0.27
2 Axis Alignment Attribute: LC 0.45 | GE 0.12
Creatures who will not appear: unchanged
Creature most likely to appear: Miner
Creature least likely to appear: Wizard

Creature number twelve: Knight
Keeper attributes after the appearance of the Knight
Human 0.33 | Corpar 0.25 | Construct 0.17| Undead 0 | Denizen 0.25
2 Axis Alignment Attribute: LC 0.48 | GE 0.18
Creatures who will not appear: unchanged
Creature most likely to appear: Monk
Creature least likely to appear: Troll Shaman/Priest

Creature number thirteen: Monk
Keeper attributes after the appearance of the monk
Human 0.33 | Corpar 0.25 | Construct 0.17| Undead 0 | Denizen 0.25
2 Axis Alignment Attribute: LC 0.52 | GE 0.17
Creatures who will not appear: Goblin, Necromancer, Wraith, Rogue, Naga, Orc, Wight, Ghoul
Creatures who will now consider appearing: Vampire
Creature most likely to appear: Warrior
Creature least likely to appear: Vampire

Creature number fourteen: Minotaur
Keeper attributes after the appearance of the minotaur
Human 0.36 | Corpar 0.29 | Construct 0.14| Undead 0 | Denizen 0.21
2 Axis Alignment Attribute: LC 0.53 | GE 0.16
Creatures who will not appear: Unchanged
Creature most likely to appear: Monk
Creature least likely to appear: Vampire

Creature number fifteen: Stone Golem
Keeper attributes after the appearance of the stone golem
Human 0.33 | Corpar 0.27 | Construct 0.2 | Undead 0 | Denizen 0.2
2 Axis Alignment Attribute: LC 0.56 | GE 0.15
Creatures who will not appear: Goblin, Necromancer, Wraith, Rogue, Naga, Orc, Wight, Ghoul, Vampire.
Creature most likely to appear: Warrior
Creature least likely to appear: Wizard

Final creature list:
3x Kobolds
1x Troll Shaman
1x Troll Priest
1x Steel Construct
1x Miner
2 x Stone Golem
1 x Warrior
2 x Minotaur
1 x Maiden
1 x Knight
1 x Monk

Conclusions
The player's alignment crept towards Lawful Good, and generally it seemed alignment was usually the deciding factor when determining if a creature showed up. Part of this is because the three starting Kobolds have Denizen 1 score, which hangs around for a long time. I am going to make Kobolds Denizen 0, so that faction should have a bit more of an impact on creature selection, as there was a very large (possibly overly large) amount of faction blindness.

Things that worked:
+ Wizards really don't like constructs
+ Naga are very touchy and hard to get
+ Goblins really like constructs.
+ Undead don't want to be around constructs
+ Miners are typically the most pragmatic of the humans
+ Some undead don't mind hanging around humans

Possible issues:
+ Orcs are difficult to get for a Children of the Deep player (might be the faction issue above).

[andrewbuck]

Yeah, I think kobolds should be considered neutral as every player always gets three of them, and they are re-spawned after a time delay if they are killed. Other than that I think the system is pretty workable, I still don't fully understand how you determine the blocks like:

Creatures who will not appear: Unchanged
Creature most likely to appear: Monk
Creature least likely to appear: Vampire

but it is obviously possible. As for the starting faction choice, I think we should have four factions (LG, LE, CG, and CE) as well as a neutral one in the middle as your choices. The idea of answering some questions about your background is an interesting one (the game Daggerfall did something like this), however I think that should just suggest one of the four starting factions for you (or at least suggest one). It could maybe then have some effect on exactly where you start too, I guess. For example, if you answer all the questions with "evil" answers and then end up choosing a "good" faction then you would start with a score that is mostly good, but not as good as if you just picked it without the questions. Also I think the questions should be optional as it would get annoying to have to do them every time you joined a game. Maybe we could make the questions apply only in campaign mode and just choose a pre-selected one before multiplayer games.

-Buck

[Keldaryth]

If you look at the very simple results in dungeons 1 and 2 you'll notice that after the wizard rocked up the goblin ended up with a negative probability of appearing. This was because of the factional dislike a goblin holds for the human faction - his factional coefficient for humans is -1, meaning that each percentage increase in the proportion of humans in your dungeon leads to an equal percentage decrease in the willingness of a goblin to enter your dungeon.

Since at that point the proportion of human faction creatures in the dungeon was 0.25, this meant that the goblin had a -0.25 modifier to his base probability of appearance which was 0.2, giving a total probability of -0.05 before alignment modifiers - which in this case were insufficient to give the goblin any chance of being attracted in.

I simply went through all the calculations for each creature to determine which creatures had no probability of appearing, and put them on the list of creatures that would not appear.

The creatures most and least likely to appear are those with the highest and lowest mathematical probability of appearance after all modifiers have been applied. In the case of dungeon 2, the creature most likely to appear would be the wizard, while the creature least likely to appear (but who could still appear) is the gold golem - it's purely based on the maths, and I thought it would be a meaningful bit of information to convey without posting up a table (which I'm not sure I can do on the forum) 24 rows long each time I did a new set of calculations - although if you guys want me to I'm happy to do that.

[Keldaryth]

These tests will start with a different faction creature and see what the three axis system lists generate. I will then repeat this using the same starting creature on the two axis system to see what differences there are in results (if any)

Dungeon 5

Test: Fifteen creature generation using the 3 axis alignment system

Note: This example does not count the 3 starting Kobolds as generated creatures

1. Necromancer
2. Vampire
3. Necromancer
4. Wraith
5. Ghoul
6. Ghoul
7. Wraith
8. Wraith
9. Skeleton*
10. Skeleton*
11. Vampire
12. Ghoul
13. Wraith
14. Wight
15. Wraith

*Out of interest I assumed 2 skeletons were summoned at this point

Lessons: Undead really don't play well with others, which is currently unsurprising. The faction lines seem to be working.

Notes: For a brief period of time the rogue and the orc were attracted to the dungeon, probably based on the chaotic freedom.
Potentially more creature might be needed to allow for mixing with the undead faction, or perhaps not. Potentially the Orc and the Rogue's numbers might need to be tweaked.

Dungeon 6

Test: Fifteen creature generation using the 3 axis alignment system

1. Orc Berzerker
2. Naga
3. Troll Shaman
4. Orc Berzerker
5. Orc Berzerker
6. Naga
7. Minotaur Warlord
8. Orc Berzerker
9. Goblin
10. Goblin
11. Troll Stone Priest
12. Minotaur
13. Orc Berzerker
14. Naga
15. Minotaur Warlord

Notes: Children of the deep appear to work as a cohesive faction

Dungeon 7
Test: Fifteen creature generation using the 3 axis alignment system

1. Diamond Golem
2. Stone Golem
3. Gold Golem
4. Steel Construct
5. Diamond Golem
6. Miner
7. Troll Shaman
8. Minotaur
9. Warrior
10. Troll Stone Priest
11. Minotaur
12. Steel Construct
13. Warrior
14. Gold Golem
15. Miner

Notes: A golem list can very quickly turn into an alignment based list around lawful or orderly behaviours. They appear to be one of the best races for creating a mixed faction strategy.

Dungeon 8
Test: Fifteen creature generation using the 3 axis alignment system

1. Wizard
2. Maiden
3. Monk
4. Wizard
5. Vampire
6. Monk
7. Maiden
8. Maiden
9. Wraith
10. Miner
11. Monk
12. Wraith
13. Ghoul
14. Necromancer
15. Maiden

Comments: There appear to be a lot of damsels in distress in this haunted castle. The humans might not like it but, as intended, the dead can come and haunt the living.

[svenskmand]

Well more or less. One thing though you now have 2 points in the same alignment system, how should they influence each other, and interact with the players alignment?

No, I don't.

Actually you do. You have a creature alignment point, and a point to determine attractiveness. Please learn the basic terminology of linear algebra (which is what you are using), there is vectors/points and matrices. It ruins the otherwise good discussion when you claim that I am wrong when I am in fact right, because you mix up the terminology.

I do not want to be rude here, but a discussion cannot take place if we are not using the same terminology, as this just gives confusion.

Also could you make a short post describing how exactly you compute the probabilities for a creature to enter you dungeon? I would like to formalize it and make a wiki post about it.

[Keldaryth]

Well more or less. One thing though you now have 2 points in the same alignment system, how should they influence each other, and interact with the players alignment?

No, I don't.

Actually you do. You have a creature alignment point, and a point to determine attractiveness. Please learn the basic terminology of linear algebra (which is what you are using), there is vectors/points and matrices. It ruins the otherwise good discussion when you claim that I am wrong when I am in fact right, because you mix up the terminology.

I do not want to be rude here, but a discussion cannot take place if we are not using the same terminology, as this just gives confusion.

Also could you make a short post describing how exactly you compute the probabilities for a creature to enter you dungeon? I would like to formalize it and make a wiki post about it.

Coefficient

This is what I am using which you are referring to as a second point. Assuming I am understanding your arguments correctly, your basic assumption is that this system is using geometry - which isn't precisely accurate. While you can conceive of creature alignments existing in a geometric space of n dimensions, the interactions between those alignments are not simply the vectors between the creature's alignment point and the alignment point of the creature or player the creature is interacting with.

Assuming I am understanding your definitions correctly, the system for calculating creature attraction to a players' dungeon is closer to using matrices, where one matrix has the coefficients for creature attraction to the various variables under consideration: each faction as a proportion of the total dungeon population and each alignment axis value.

Personally, I find it easier to think of this system as using a number of continuums rather than a n dimensional geometric space, and using simple linear algebraic equations which work of those values.

However, I will freely state that I stopped studying maths at High School and I may indeed be using non-standard terminology to describe the system I am proposing and am simply going by my understanding of the terminology you are using - i.e. my understanding of your meaning of your terminology

I will work on an in depth post for you and andrewbuck in terms of this system.

[andrewbuck]

I think svenskmand is being a bit over-antagonistic. I don't think Keldaryth has been perfect in his use of terminology, but I think that is somewhat of a moot point as he has given enough examples of how his system works that we can effectively reverse engineer it from his example numbers and trial runs. The system obviously works on at least a basic level, there may be some obscure effects we haven't seen yet due to numerical instabilities and the like, but barring very weird stuff like that it appears the system will work.

When I get the spreadsheet and am able to see the exact calculations being carried out I will understand it better, but as of now I think this is a good enough candidate system to be implemented in the code. It does not necessarily have to stay that way, I have rewritten sections of the code before and will likely do it again, but if we keep arguing forever the game will never be done. Portals are a key aspect to making a playable game; you can play the game as it is now but once your creature's die you are finished. One of the reasons portals have never been added is due to the fact that we had no agreement on how they should work. Everyone had proposed various ideas, but no one every backed their idea up with numbers and equations to the point that it could be tested and implemented. Keldaryth has done this and for this reason I think we should run with it and see where it gets us.

-Buck

[svenskmand]

I, as everybody else (except Keldaryth), is just trying to understand Keldaryths system. And I get a better understanding of it if I see the formal definition of the system (that is how you do all you calculations formally) I do not like to try to figure out how your system works from examples, I like to see the formal definitions and THEN see the examples, but that is just me - note I am not blaming Keldaryth for the form of his presentation of his system.

But what really annoys me is when I ask a question about some part of his system, formulated using standard mathematical terminology, and Keldaryth then just bluntly says that my question does not make sense because he does not understand my terminology. For example the two points/vectors that he uses, the first is the alignment of the creature, and the second is used for determining the attraction of the creature to a keeper. Then Keldaryth claims that the second point is not a point, but coefficients, even thought it is in fact also a point, and it is most natural to look at it as a point, especially because there is no syntactically difference between the first and the second point.

I can try to give an analogy to illustrate my annoyance with your misunderstanding of my questions even though this analogy is somewhat flawed. Say you are building a garage for a limousine service and when I ask you questions about your construction. and I use the term "cars" to refer to the limousines, then you just say that my question does not make sense, because I am not using the word "limousine" in place of the word "cars", even though a limousine is also a car. So what I would like is that you accept that other people does not have the same view of you system as you do, and therefore would probably ask you questions seeing your system from different points of view, meaning that the questions would still be perfectly valid if you just try to understand them just a bit more, from another point of view.

This is what I am using which you are referring to as a second point. Assuming I am understanding your arguments correctly, your basic assumption is that this system is using geometry - which isn't precisely accurate. While you can conceive of creature alignments existing in a geometric space of n dimensions, the interactions between those alignments are not simply the vectors between the creature's alignment point and the alignment point of the creature or player the creature is interacting with.

Assuming I am understanding your definitions correctly, the system for calculating creature attraction to a players' dungeon is closer to using matrices, where one matrix has the coefficients for creature attraction to the various variables under consideration: each faction as a proportion of the total dungeon population and each alignment axis value.

I am not assuming that you are using geometry. I am assuming that you are using linear algebra (or that you system can be expressed in terms of linear algebra), which you also now say it can almost be.

I think svenskmand is being a bit over-antagonistic.

I need to ask questions to understand how his system works. And one also need to challenge his system to understand how it works, and check for any weaknesses.

[Keldaryth]

Regardless, a spreadsheet can be found here.

Should a post still be required it will have to wait for another day. I need sleep.

[svenskmand]

Thanks, but I would also like if you could make a post of how you formally do the calculations, but do get some sleep first

[Keldaryth]

But what really annoys me is when I ask a question about some part of his system, formulated using standard mathematical terminology, and Keldaryth then just bluntly says that my question does not make sense because he does not understand my terminology. For example the two points/vectors that he uses, the first is the alignment of the creature, and the second is used for determining the attraction of the creature to a keeper. Then Keldaryth claims that the second point is not a point, but coefficients, even thought it is in fact also a point, and it is most natural to look at it as a point, especially because there is no syntactically difference between the first and the second point.

The second 'point' as you call it, although I understand that it can be viewed as a geometric point, is intended to be three coefficients. I can appreciate that given they have a range between -1 and 1 as the first 'point' it can seem logical to view them as a point, and I am asking you to not do that as that is not how they are intended to be viewed. The reason that everything has a range between -1 and 1 is because that is what I believe will make it easiest for creature calculations as demonstrated in the spreadsheet which I have now posted, and I hope it will help make things clearer. I hope you can appreciate that creating an entire system from scratch and getting it and enough examples to see if it works is not the easiest process to undertake or explain. I make things that work and sometimes I don't know exactly how - I just know they do. Finding a way to explain it to you over the internet in a fashion *you* can understand isn't necessarily the easiest thing in the world and frankly, while I can understand your frustration, asking someone to 'learn basic terminology of linear algebra' without 'intending to be rude' is actually quite rude. It would be a bit like demanding that you learnt Chinese in order to explain a nuclear physics to me if I only spoke Chinese - and I would hope that if a contributer from Japan, China or other non-english speaking background popped up, you wouldn't insist they went away and mastered the English language before contributing.

I am trying to build something and you are trying to understand it. I get that. I have also stated that I wished to test the system a bit before putting the nuts and bolts on display. I'd rather be happy that I have corrected what I think are the big gaping errors in a piece of work before letting everyone else loose on it. I still haven't fully thought through the implications this system could have for the game and ways it could be used to determine creature behaviour and interaction as well as the implications for rooms and other pieces of dungeon dressing, the skinning and lighting of a players dungeon etc. etc. etc.

Anyway. Hopefully now that there is a spreadsheet you will be able to poke around to your hearts content and get the answers you seem to want.

I would like to point out that I am not taking offense at what you have said - or if I have, I'm over it. I'm pretty blunt at times, so encountering it in others isn't something I'm allowed to be bothered by, and in any case, this is the internet - it's hard enough to communicate without 70-90 percent of the non-verbal (and in come cases verbal) cues we'd otherwise look for.

[Keldaryth]

Thanks, but I would also like if you could make a post of how you formally do the calculations, but do get some sleep first

If you opened the spreadsheet and looked at the cells - all the calculations are in there. You're a mathematician, I have faith in your ability to figure them out

[svenskmand]

I have looked at it, but I cannot see all the values that you use in you posts above, lets for example look at the Knight:

Knight
A noble commander in the service of Lord Darius
Faction: Human 1
Alignment (2 Axis): LC 0.9, GE 0.8, (Knights are schooled to uphold the law and work for the benefit of all)
Alignment (3 Axis): AE 0.7, OC 0.9, PV -0.6, (As above, although the knight is trained to protect those weaker than themselves with martial training)
Faction Coefficients: Humans 1, Children -0.8, Constructs -0.7, Undead -1 (I will protect the people of Evariar from those that would do them harm)
Alignment Coefficients:
2 Axis: LC 0.9, GE 0.9, (Knights like orderly societies that treat people well)
3 Axis: AE 0.8, OC 0.9, PV 0 (As above. Although a knight values peace, s/he will go to war as necessary)

I cannot find the points (0.9,0.9) and (0.8,0.9,0) but I can find the points (0.9,0.8) and (0.7,0.9,-0.6). But it will probably be easier if you could make a short post explaining the calculations, but it is up to you.

Also I use OpenOffice so it might be that it does not show all the things in you office spreadsheet.

[svenskmand]

It would be a bit like demanding that you learnt Chinese in order to explain a nuclear physics to me if I only spoke Chinese - and I would hope that if a contributer from Japan, China or other non-english speaking background popped up, you wouldn't insist they went away and mastered the English language before contributing.

Ok this is maybe how you see it. Let me then use this example to show how I see it: I know Chinese and nuclear physics. You do not know Chinese, but you do know nuclear physics, you have also discovered say a way to do clean an easy fusion. Now you try to explain it to me in Chinese and I do not understand everything that you are saying, as you speak Chinese poorly. Then I ask a question in Chinese and you understand it very differently from what the original semantics of the question where, and then bluntly says that my question is nonsense.

Then hopefully you will understand my annoyance in such a situation? So the point is that you should not try to communicate with me in a language I am familiar in, if you are not.

I am trying to build something and you are trying to understand it. I get that. I have also stated that I wished to test the system a bit before putting the nuts and bolts on display. I'd rather be happy that I have corrected what I think are the big gaping errors in a piece of work before letting everyone else loose on it. I still haven't fully thought through the implications this system could have for the game and ways it could be used to determine creature behaviour and interaction as well as the implications for rooms and other pieces of dungeon dressing, the skinning and lighting of a players dungeon etc. etc. etc.

That is completely understandable, then I will just wait with questions until you have formalized your construction, and is ready to explain it to the rest of us on the forum. But I just thought that you wanted a discussion about your stuff, while you where making it, because you posted it online and worked on it online, but that was not the case then.

[Keldaryth]

Ok this is maybe how you see it. Let me then use this example to show how I see it: I know Chinese and nuclear physics. You do not know Chinese, but you do know nuclear physics, you have also discovered say a way to do clean an easy fusion. Now you try to explain it to me in Chinese and I do not understand everything that you are saying, as you speak Chinese poorly. Then I ask a question in Chinese and you understand it very differently from what the original semantics of the question where, and then bluntly says that my question is nonsense.

Then hopefully you will understand my annoyance in such a situation? So the point is that you should not try to communicate with me in a language I am familiar in, if you are not.

From my perspective, this is the same thing. The point I'm making is that in this situation the common language is the only way we have of attempting to communicate. Yes there will be frustrations, yes there will be misunderstandings, but to say 'learn this so I can understand you' comes off (on the net) as rude because I am the one with the method of clean and easy fusion. Still, like I said, it's largely irrelevant, because if I were in your position I would probably say the same thing out of frustration, so yes I can understand it. As I said, I too can be very blunt (as you have experienced). I get the strange feeling we're more alike in our temperaments than we might otherwise think, so I also get that your main concern is making the best game possible, which is admirable, no matter how much we may clash heads.

I am trying to build something and you are trying to understand it. I get that. I have also stated that I wished to test the system a bit before putting the nuts and bolts on display. I'd rather be happy that I have corrected what I think are the big gaping errors in a piece of work before letting everyone else loose on it. I still haven't fully thought through the implications this system could have for the game and ways it could be used to determine creature behaviour and interaction as well as the implications for rooms and other pieces of dungeon dressing, the skinning and lighting of a players dungeon etc. etc. etc.

That is completely understandable, then I will just wait with questions until you have formalized your construction, and is ready to explain it to the rest of us on the forum. But I just thought that you wanted a discussion about your stuff, while you where making it, because you posted it online and worked on it online, but that was not the case then.

I don't mind queries, but I have to say the problem with working on things is that I don't always know *how* to describe what I do, and often I will try to explain it using whatever frame of references are currently available, only to find a different frame of reference is better. You could almost say my explanations subscribe to the buddhist theory: I find a way that fits even if I have to try multiple times and throw out the old ones. I guess the thing is that this 'system' is still less than a week old, so what I was looking for was comments on the effects, rather than a comment on the actual process. For example, if the system hadn't worked we'd see it in the output, i.e. a list full of knights and wights, or a list mainly comprised of golems and undead. I suppose I should have flagged that I was looking for analysis of the output.

The main reason I wish to test both 2 and 3 alignment configurations is to see if one works 'better' than the other in terms of game output, i.e. which one (if any) gives random creature listings which make sense given the game parameters - although even then you could easily say (as you have) that you'd prefer to look at it from a formula-first perspective.

Anyway. I'll see if I can do a formulae first posting soon from a more theoretical side, since I think that would benefit everyone.

EDIT: P.s. svenskmand, it's taken me a while, but I just figured out your icon. Way cool. *facepalm* I is dense at 6 in the morning sometimes.

[svenskmand]

As I said, I too can be very blunt (as you have experienced). I get the strange feeling we're more alike in our temperaments than we might otherwise think, so I also get that your main concern is making the best game possible, which is admirable, no matter how much we may clash heads.

Yes that is bound to happen in any discussion where each party does not agree OR completely understand the other parties ideas. And yes as long as we make the best game possible/ever then we cannot do it better

I suppose I should have flagged that I was looking for analysis of the output.

The main reason I wish to test both 2 and 3 alignment configurations is to see if one works 'better' than the other in terms of game output, i.e. which one (if any) gives random creature listings which make sense given the game parameters - although even then you could easily say (as you have) that you'd prefer to look at it from a formula-first perspective.

Yes that would probably have been better

Anyway. I'll see if I can do a formulae first posting soon from a more theoretical side, since I think that would benefit everyone.

I will look forward to it

EDIT: P.s. svenskmand, it's taken me a while, but I just figured out your icon. Way cool. *facepalm* I is dense at 6 in the morning sometimes.

Heh, if you look at the [url=opendungeons.sf.net]wiki's[/url] front page you can see that the logo is just the O and D from OpenDungeons, But yes a facepalm might be in place for not noticing

[Keldaryth]

Oh I see it NOW. I guess I never paid enough attention to the fact that the 'O' in OpenDungeons was shaped like a keyhole.

Incidentally, the ACTUAL spreadsheet has been uploaded, so please redownload it - it's better formatted than the working copy I uploaded earlier.

And thank you all for being too polite to tell me it was a piece of crud - because it was

[svenskmand]

Oh I see it NOW. I guess I never paid enough attention to the fact that the 'O' in OpenDungeons was shaped like a keyhole.

Incidentally, the ACTUAL spreadsheet has been uploaded, so please redownload it - it's better formatted than the working copy I uploaded earlier.

And thank you all for being too polite to tell me it was a piece of crud - because it was

The keyhole is inspired by the Open Source Initiative and the D is made into a devil because DK and DK2 was all about evil and devilish looking creatures

I am looking at the new spreadsheet, and yes it is better A trick when you do the dot product of two vectors like (a_1,a_2,a_3,...,a_n) and (b_1,b_2,b_3,...,b_n), and you want to compute a_1b_1+a_2b_2 + ... + a_nb_n, which is the dot-product, then you can use the SUMPRODUCT function which takes two arrays as input like SUMPRODUCT(A1:A10, B1:B10), at least it is called that in OpenOffice but there must be a similar function in Office.

[Keldaryth]

I am looking at the new spreadsheet, and yes it is better A trick when you do the dot product of two vectors like (a_1,a_2,a_3,...,a_n) and (b_1,b_2,b_3,...,b_n), and you want to compute a_1b_1+a_2b_2 + ... + a_nb_n, which is the dot-product, then you can use the SUMPRODUCT function which takes two arrays as input like SUMPRODUCT(A1:A10, B1:B10), at least it is called that in OpenOffice but there must be a similar function in Office.

And already you've blown past my knowledge of maths

Ah well, I can learn. Thanks. I may look into it should I have to do it again, but for the non-mathematically inclined the current format might make more sense. Of course, for actual coding, I'd suggest going with the most efficient systesm

[Keldaryth]

These tests will start with the same creatures from dungeons 5-8 and see what the two axis system lists generate. This can then be used to compare against the 3 axis system.

Dungeon 9

1. Necromancer
2. Wraith
3. Vampire
4. Ghoul
5. Ghoul
6. Necromancer
7. Necromancer
8. Wraith
9. Skeleton*
10. Skeleton*
11. Necromancer
12. Wraith
13. Ghoul
14. Wraith
15. Necromancer

Notes: no wights, but they still had a 10% chance of appearing. I put in two skeletons at 11 and 12 as I did last time. No rogues or orcs were ever thinking of joining this community.

Dungeon 10

1. Orc Berzerker
2. Goblin
3. Ghoul
4. Goblin
5. Goblin
6. Naga
7. Naga
8. Troll Shaman
9. Wraith
10. Ghoul
11. Orc Berzerker
12. Ghoul
13. Goblin
14. Necromancer
15. Orc Berzerker

Results: The undead are happy to haunt the Children of the Deep as well. Checking the maths, they are actually LESS likely to show up in the 2 continuum system than the 3 axis system. I'm not entirely certain which aspects of the model are causing this, but it's just a side note at this point should people feel it to be problematic.

Dungeon 11
1. Diamond Golem
2. Steel Construct
3. Steel Construct
4. Diamond Golem
5. Gold Golem
6. Warrior
7. Knight
8. Gold Golem
9. Stone Golem
10. Gold Golem
11. Gold Golem
12. Stone Golem
13. Maiden
14. Miner
15. Monk

Once again the Constructs very much lend themselves to a mixed faction list - typically with humans, although Trolls are still an option.

Dungeon 12
1. Wizard
2. Monk
3. Cleric
4. Wizard
5. Cleric
6. Cleric
7. Miner
8. Miner
9. Warrior
10. Vampire
11. Warrior
12. Miner
13. Monk
14. Knight
15. Necromancer

Notes: Once again the undead like to hang around humans.

[Keldaryth]

On balance, I'm not seeing massive flaws in either system. The numbers are slightly different, but both seem to give a similar result. I personally believe the three continuum (axis) system will be better in the long run for building other game systems around it (unless we change the 2 axis labels to something more meaningful), and while my gut tells me it's the best system for balancing potential game function with simplicity, I have yet to prove this - obviously either to myself or anyone here.

Thoughts?

[andrewbuck]

I think the three axis system would probably be alright. My objection to having too many axes is based on the fact that the number of possible creature personalities grows exponentially with the number of axes; however three is still small enough to avoid having too many types (especially since we will like have something on the order of 50 creatures in the finished game). The other potential problem is the issue of being able to assign meaningful values to each to the axes for each creature, if you think you can come up with appropriate numbers for all of these values then I have no objection to the three axis system. It seems like svenskmand wanted to go with a fairly high dimensional system so this is an argument for the three axes system over the two axis one. Ultimately, I would be fine with either 2 or 3, it makes little difference from a coding perspective. I will leave the final to the two of you since it will be you that will likely be filling in values for each of the creatures.

-Buck