New Game Options: Leaderhead Levelups and No Positive Traits on Game Start

Not rue if this was answered yet (didn't read all the posts) but does the AI know what they're doing when they pick traits or are they just picking random traits that could possibly ruin their kingdom
 
They pick according to their personalities as opposed to strategically. All negative traits are, well, just that, and all positive traits should be pretty much positive so no selection should be much better or worse than another if we have our traits nicely balanced. And picking by personality will play into their strategies quite naturally if the leader has been well thought out in its design. A flavor evaluation, similar to the way leaders pick many other things including religions when on the Divine Prophets/Choose Religions options, are the methods of selection and the flavor values on each trait also represent the strengths and weaknesses of that trait (at least that's what they should be doing... giving ai weight according to the strengths of the trait.)

I did not want an evaluation heavy, situation reactional method for their decision making as we have in many AI choices like tech tree and civic decisions. Particularly since traits should be a result of the leader's personalities in the first place, and that's not always for the best. You the player get a bit of an edge there but whatever you pick is pretty much permanent so if you consider things too near sightedly, it will come back to bite you later. And vice versa, it'll take a while to see the benefit which could be a lot of wasted time potentially.

IMO, this whole system will see vast improvement as we hunker down and really work on further development on the traits over the course of the next cycle.
 
No but that's a step to take soon. We have identified, really, that ALL the National Commerce totals are really needing to be converted to a long int. That should be a project that should be done very soon. I believe it would be fairly easily done but should it be so 'easy' as to be done during the freeze? I suppose it counts more as a bug fix issue... so maybe it should be huh?

Your proposed adjustments won't make it any further problematic for the developing leader option than that limitation already is since either way the limit would be hit and cause that system to get a bit screwy anyhow.

OK, sounds good to me. I think that new stuff that screws around with the internals should wait until Koshling or AIAndy get back, otherwise things can remain broken for a long while, but adjusting that sounds like a good idea to do during Multi-Map development next version.

They pick according to their personalities as opposed to strategically. All negative traits are, well, just that, and all positive traits should be pretty much positive so no selection should be much better or worse than another if we have our traits nicely balanced. And picking by personality will play into their strategies quite naturally if the leader has been well thought out in its design. A flavor evaluation, similar to the way leaders pick many other things including religions when on the Divine Prophets/Choose Religions options, are the methods of selection and the flavor values on each trait also represent the strengths and weaknesses of that trait (at least that's what they should be doing... giving ai weight according to the strengths of the trait.)

I did not want an evaluation heavy, situation reactional method for their decision making as we have in many AI choices like tech tree and civic decisions. Particularly since traits should be a result of the leader's personalities in the first place, and that's not always for the best. You the player get a bit of an edge there but whatever you pick is pretty much permanent so if you consider things too near sightedly, it will come back to bite you later. And vice versa, it'll take a while to see the benefit which could be a lot of wasted time potentially.

IMO, this whole system will see vast improvement as we hunker down and really work on further development on the traits over the course of the next cycle.

I do not like the idea of The AI choosing traits based off of LH personalities. That gives the human, who is playing to win, an advantage as they will always choose the trait they see as better for their situation, as opposed to an RP trait. You may want to talk to Koshling about a way to do this without hurting the AI players more than they are already hurt.
 
My 2 cents here? Great job! Love the idea (of both!)
 
I do not like the idea of The AI choosing traits based off of LH personalities. That gives the human, who is playing to win, an advantage as they will always choose the trait they see as better for their situation, as opposed to an RP trait. You may want to talk to Koshling about a way to do this without hurting the AI players more than they are already hurt.

Really, any given trait should be equal. As a player, I won't be selecting traits with immediate needs in mind anyhow either. Being stuck with the selections you take kinda enforces that you think of it in the sense of a long term strategy anyhow and not take them as reactions to immediate situations. If you do, you might gain an immediate and needed edge but your overall game style strategy would suffer in the long run. So with those facts in mind, there really are no bad choices (or at least there shouldn't be and I think the current traits are somewhat adherent to that enough for now.) Therefore, there should be no huge evaluation but rather a selection based on the leader's general overall personality/1000 ft overview of the leaders general game strategy. And that's what we achieve with flavor based determinations.

Additionally, the AI should not be all rigged for strategy decisions but also for Role Playing flavor too. It helps with game immersion to, as a player, feel like the only one here who really thinks over all the angles without having emotional impulses guiding you. If the leaders come across more as people... leaders with feelings... rather than all of them being computerized strategic processing modules like you'd play against in chess, then the player can feel more a part of a dynamic and alive game world. So some flaws that may be evidenced in this method is not necessarily a flaw in game design.

I can understand other ways of seeing this but I don't think the argument that we should design more reactive selection mechanisms holds enough value in this particular AI case. I'm not unable or resistant to designing such a mechanism so much as I flat out don't believe its the right way to develop them in this case.

My 2 cents here? Great job! Love the idea (of both!)

Thanks Prime!
 
Really, any given trait should be equal. As a player, I won't be selecting traits with immediate needs in mind anyhow either. Being stuck with the selections you take kinda enforces that you think of it in the sense of a long term strategy anyhow and not take them as reactions to immediate situations. If you do, you might gain an immediate and needed edge but your overall game style strategy would suffer in the long run. So with those facts in mind, there really are no bad choices (or at least there shouldn't be and I think the current traits are somewhat adherent to that enough for now.) Therefore, there should be no huge evaluation but rather a selection based on the leader's general overall personality/1000 ft overview of the leaders general game strategy. And that's what we achieve with flavor based determinations.

Additionally, the AI should not be all rigged for strategy decisions but also for Role Playing flavor too. It helps with game immersion to, as a player, feel like the only one here who really thinks over all the angles without having emotional impulses guiding you. If the leaders come across more as people... leaders with feelings... rather than all of them being computerized strategic processing modules like you'd play against in chess, then the player can feel more a part of a dynamic and alive game world. So some flaws that may be evidenced in this method is not necessarily a flaw in game design.

I can understand other ways of seeing this but I don't think the argument that we should design more reactive selection mechanisms holds enough value in this particular AI case. I'm not unable or resistant to designing such a mechanism so much as I flat out don't believe its the right way to develop them in this case.

But not every trait is equal in all situations. For instance, if you are the only civ on your continent it is rather pointless to take Aggressive or Protective as you won't be using them too much. Likewise if you have no coasts it is not a good idea to take Seafaring. If you are hostile with many civs and are going to be at war for a while takign Philosophical may not be the best choice. Do you see what I mean by that? The AI should know that, and not be hardcoded into following the flavor values.
 
Well, you make some good points so I can see some value in coding in some special case tweaks to the evaluation system where its possible to easily evaluate those things (for example, modify the result on seafaring based on the percentage of your cities that are coastal with 50% being the zero point etc...)

We'd have to consider each one this way and give a way to do so where its not entirely a hard-coded to specific trait sort of affair. That'll involve another trait tag or two, perhaps one to aide with each special situation to indicate that this trait applies to that special evaluation.

But for the core basis of trait selections, it still makes sense to go about it based on the leader flavors compared to the trait flavors - this helps to also make sure they tend to have consistency in valuing the same general types of benefits overall.

I have also begun to consider that the C2C team should begin considering adding additional flavors to the flavor structure. I don't think the current setup there is really comprehensive enough for the added scope C2C brings to the table. But this thread isn't the best place to begin a discussion on that - though its related its a whole matter of its own to consider and affects a LOT of different game sections.
 
Maybe then for V28 something should be added to the hover text for Leveling Leaderheads that says "The AI may not use this well" or something, in order to let people know what they are getting into.
 
The chances of it being any more crippling than a usual mismatch on their predefined traits if the options aren't utilized is no greater though. The chance exists just as strongly that a seafaring leader ends up with a landlocked nation in the normal manner of play. The only difference here is that the player could be smart enough to choose to avoid such a mismatch.

So this, I think, is a fairly minor imbalance factor in relation to other factors we know we have still. The actual strategic warfare shortcomings of the ai is far more impacting to the AI's play than this is.

I will be adding a message on the hover however to state that "This option is in beta testing and could suffer from some imbalance." And mostly that would be due to the fact that the traits as we have them haven't been hashed out anywhere to the degree we all want to see them developed yet, particularly in light of the developing leaders.

But in the meantime, yes, we should start compiling a list of obvious bad choice scenarios that should play a factor so that the evaluations can be tweaked accordingly.

So far we have:
  • Seafaring should not be picked on primarily landlocked nations but should have greater weight for those nations that are mostly coastal anyhow.
  • Aggressive and Protective should be diminished somewhat in value if there are no other civs in contact with this one. (This one is iffy. Barbs pop up all over the place and in the long run, impeding these two might impede the later game for that civ. Additionally, some very isolated peoples ARE very hostile by nature so it may not be necessary to have this special case considered.)
  • Some traits are better for nations that manage to maintain peace, others better for nations who are often at war or have already lost cities and enmities with neighbors are high. A weight tweak should perhaps be given to this particular condition - likelihood of conflict.
Some (more) specific thoughts on these:

Generally, leader personalities that would choose to invade often would also have high desire for military based flavors already so to some extent this is already inherrent in the current system. But we could add a frequency of war count, a frequency of being invaded or a # of cities lost count and use those to further influence some of the more military vs non-military civics. But again, I wonder, is being ultra defensive a matter of personality based paranoia than it necessarily is about the situation?

I find in evaluating some players who tend to choose a trait like Protective, a desire to spend traits to make sure they would be ready if trouble comes with the sense that trouble can't be predicted. Perhaps weighting in this way would tend to somewhat backfire as it would go off of what has happened, when the truly prepared are prepared even though nothing may have happened... yet.

I believe we'd find in a multi-player setting, that even humans aren't evaluating the immediate perceived needs for trait selections so much as their overall strategy in how they feel they can come out on top in general. So where it's not absolutely necessary, perhaps it should not be a special case like this. I, personally, would probably lean towards choosing Aggressive, even if there aren't any other civs in reach at the moment because I know there WILL eventually be no matter what. And I would find the benefits to my military worthwhile even without a player enemy because there's always animals/barbs, and the prepared army to deal with any intruders at any time.

Anyhow, so far, I'm only seeing the first case as truly necessary to address. But I'm sure more may be brought up.
 
I was having a "minor revolt" every 10-20 turns until I was able to choose a trait with less revolutionary sentiment, I have not had any trouble at all since then.

I have never been able to select a second tier in any trait yet, when does that become possible?
 
I was having a "minor revolt" every 10-20 turns until I was able to choose a trait with less revolutionary sentiment, I have not had any trouble at all since then.

I have never been able to select a second tier in any trait yet, when does that become possible?

When we add the code for it? :rolleyes: Currently this is only supporting adding new existing traits. The new tiers are what I'm writing up in terms of balance and stats.

Speaking of which I should have the next three trait suggestions ready soon.
 
I was having a "minor revolt" every 10-20 turns until I was able to choose a trait with less revolutionary sentiment, I have not had any trouble at all since then.

I have never been able to select a second tier in any trait yet, when does that become possible?

The coding is there but the trait designs themselves are not. Next cycle, this is a goal to hash those out.
 
I'm not exactly sure why, but the AI is able to choose the trait "Barbarian" as one of it's negative traits, where the player is not. As Barbarian is an entirely positive trait all around, it's hardly surprising that every single AI civ chooses barbarian as it's first negative trait. In my current game using this option, all five of the civs I have contacted chose this trait as their negative.
 
hmm... not sure why the player wouldn't be able to select that trait but the point is well made that barbarian is not well balanced to other negative traits. I would imagine another along the same lines of difficulty would be the Hunter/Gatherer trait. Unfortunately, anything I do about those would be a bit of programming for a temporary fix as in the end, all traits should be equally balanced to each other (well... positive to positive and negative to negative anyhow).

But I could certainly do something about it and probably should implement that quick fix for now anyhow.

Thanks for the feedback! That's a very helpful report!
 
Near as I can tell, the "Barbarian" trait is meant only for the Barbarian Civ. It grants bonuses to nearly everything except science, culture, and finances to help the Barbarians remain competitive. The simple fix is to make sure that nobody can select this trait as a positive or negative, keeping it solidly within the barbarian's realm.

Also, was it your intention that everyone be able to select 3 positives and one negative by tribalism? In my current game, that is the case. For reference, my main civ is Scientific, Creative, Financial, Megalomaniac, every AI civ has Barbarian as their negative except the vikings, which chose Revolutionary. Some interesting things have popped up, though. Hitler chose Protective, Charismatic, then Industrious, suggesting a Hitler that just wants to keep his people safe and successful. Who knows how this will develop? Boudica went for Aggressive, Protective, and Nomadic, because they are clearly trying to collect as many promotions as possible.

Here's the full breakdown of what I have seen so far, in game.

Pericles: Spiritual, Industrious, Scientific, Barbarian (For some reason, he is the score leader, even though I am more technologically advanced and have 5 cities to everyone else having 1.)
Boudica: Aggressive, Protective, Nomadic, Barbarian
Hitler: Protective, Industrious, Charismatic, Barbarian
Ragnar: Aggressive, Protective, Imperialistic, Revolutionary
Augustus Caesar: Aggressive, Industrious, Progressiveness, Barbarian
Elizabeth I (Myself): Scientific, Financial, Creative, Megalomaniac
 
Sounds about right. The selections will come very slowly after this point. Yes, I intended one extra selection beyond the usual 2 positive, 1 negative by the end of prehistoric so those playing without No Positive Traits on Gamestart would have a selection by the end of the first era. But they will come slower and slower as the game progresses.

Those picks are pretty in line with the programming since I'm seeing the military flavor leaders go for the military flavor traits there. Might be nice to add some additional flavors to help us to refine this a bit next cycle too. One for Defense, for example, could be helpful to give Protective and perhaps Defensive if that one is introduced, a differentiation from simply 'military' so that leaders like Hitler could be set to consider those Turtle traits a little less valuable and other leaders that would prefer that approach be given a means to gain some weight towards it.

The only issue with doing that is that it takes a fairly major reconsideration of all the leaders thereafter and... whoah that's a lot to consider!

As for the Barbarian trait... I'll look into it a bit more. SO stated he was wanting other leaders to have access to it too. I could always simply assign the flavor values on it so that it suggests its benefits aren't quite as good as they are for now too.
 
Here's the thing about Barbarian Trait, IMO:

If you really look at history, you GOT to remember your starting at 50,000 BC, you are not actually starting as the Leader you picked OK, you need to get that picture out of your head first, IMO again.

Back then weren't ALL leaders considered, Barbarian(s)? I believe no one actually got along with even their own neighbors back then. (I might be wrong).

I haven't gotten that much into it yet, so i canNOT really say how it is going to work?? But if its what you have written down, then it seems actually, as TB said "Reasonable." ie meaning alot of them have Barbarian. Again making sense then. BUT i need to know i guess is, if the Barbarian Trait can be transposed once you hit the Ancient Era, and MUST take another Negative Trait, now i see that as a substitute then, or am i off track here?? Can it be done this way, TB??
I am really only concerned about GAMEPLAY, not really RL, but it would be nice if it could, IMPO.
 
Here's the thing about Barbarian Trait, IMO:

If you really look at history, you GOT to remember your starting at 50,000 BC, you are not actually starting as the Leader you picked OK, you need to get that picture out of your head first, IMO again.

Back then weren't ALL leaders considered, Barbarian(s)? I believe no one actually got along with even their own neighbors back then. (I might be wrong).

I haven't gotten that much into it yet, so i canNOT really say how it is going to work?? But if its what you have written down, then it seems actually, as TB said "Reasonable." ie meaning alot of them have Barbarian. Again making sense then. BUT i need to know i guess is, if the Barbarian Trait can be transposed once you hit the Ancient Era, and MUST take another Negative Trait, now i see that as a substitute then, or am i off track here?? Can it be done this way, TB??
I am really only concerned about GAMEPLAY, not really RL, but it would be nice if it could, IMPO.

Are you by any chance confusing the trait "Barbaric" with "Barbarian"? Barbarian seems to have been intended as a trait only for the Barbarian civ. Here are the trait's modifiers.

Barbaric:

-15% culture
-20% GPP
+15% Military Production
Free Combat 1 for Melee Units
+2 crime/turn if city's population equals (something? I can't understand what the math is saying when the traits affect crime)


Barbarian:

+1 Health
+2 Happiness
+4 Free Units
Free Military Units Equal to 35% of the population
-30% Civic Upkeep
+50% Military Production
Free Aggressive Promotion for Melee
Workers Build Improvements 50% Faster
Can Upgrade Units Outside Borders
+25% XP Needed For Promotion (The only negative assocated with this trait)


If everyone was taking Barbaric, I wouldn't really think much of it except to wonder why the lack of variety. Given how absolutely awesome Barbarian is, it isn't even remotely surprising that every AI civ took it. I would take it too, if it were an option for me. :)
 
I could make a boolean tag that limits Barbarian to just that... Barbarians. Otherwise it doesn't currently make much sense as either a positive or a negative promotion. I'll do that today then.
 
I just started a new game, and I noticed that now the AI Civs don't seem to be taking a negative trait at all, unless the negative is just failing to display when I mouseover on the scoreboard.
 
Top Bottom