Traits that are bugged in 1.2 with fixes. Part 1: Scarred, Farmer, Trader

[gregab25]

You can quite easily break the threshold, simply move your governor to another settlement.

I am not Kreskin so i cannot read CA's mind, but I don't think they meant for your governor to run around all the time to get a single virtue. Of course like I said this is just my opinion and your opinion is as valid as mine. We just disagree.

Another thing to support my reasoning though: Remember you want to get your virtues to a high level in part because they then have a chance to be passed down from father to son. However, in order to qualify for these inherited virtues the father must have a very high level of that virtue. Getting the Farmer and Trader virtues to the level they need to be passed down would be nearly impossible with the default triggers even if you did run your governor from province to province.

Personally I don't like the logic of changing the trigger conditions to give GoodFarmer points on buildings other than farms.

This is just an erroneous statement as that is not what is happening at all. If you apply the corrections i suggest you get points for GoodFarmer because you build Farms before other buildings, not because you build buildings other than Farms. That would just be silly.

The problem with the triggers lies in two areas: the most important being that the condition doesn't allow for the possibility that the appropriate building has already been built for that level for settlement, it's only then that the effects of the unbalanced thresholds become a problem. If you correct the first problem then the thresholds don't matter so much. Conveniently enough, there's already an official example shown in the handling of the BadMiner vice - simply add the condition that the Advisor is recommending a farm/trader is built, although like the BadMiner triggers this will probably require one trigger for every level of building.

Please clarify what the "appropriate building" you are referring to is and why it you came to the conclusion it is the "appropriate building"? Im trying to understand what you are saying. Perhaps you can post an example of how you think the trigger should look?

If you want all levels of farms, you can use '...>= farms'

I'm sorry but I just do not understand this. The >= sign is part of an equation to compare Farms to the previous subject which is "SettlementBuildingFinished." It is not there as a stand alone arguement to qualify Farms. You cannot just arbitrarily remove the subject "SettlementBuildingFinished" and use half of an equation.


Again, if you would try the corrections i suggested before commenting you could at least see that they work precisely as I explained them.

As far as how CA intended them to work, that is open for disagreement. I have tried to use what mathematical evidence that there may be in the files to support my arguements. No one can claim to know for sure what CA meant unless CA stated what they meant. So argueing about what CA intended is really just an exercise in futility. If you disagree with the supporting evidence i gave then just don't make the changes.

Quite frankly, I wish CA would respond with what they intended so we could quit the speculating and get on to fixing the problem. I especially would like them to comment if I am giving wrong advice, because i don't want to that, even if it meant a bunch of egg on my face.

If we knew what they meant it would probably take all of five minutes to fix the problem if the corrections i suggested do not fix it already.

[gregab25]

Heya Jambo , this is just a quick reply.

I think shieldbearer is for non roman cultures.

And also I sent you a PM.

I'll look at the rest of your post in more detail now.

[gregab25]

Jambo:

From the Ancilliary Data Section -

;------------------------------------------
Ancillary shieldbearer
Image helmeted_soldier_ancillary.tga
ExcludedAncillaries swordbearer
ExcludeCultures roman
Description shieldbearer_desc
EffectsDescription shieldbearer_effects_desc
Effect BodyguardValour 1

;------------------------------------------

If you remove the line ExcludeCultures roman I think it might work.

[tai4ji2x]

Quite frankly, I wish CA would respond with what they intended so we could quit the speculating and get on to fixing the problem. I especially would like them to comment if I am giving wrong advice, because i don't want to that, even if it meant a bunch of egg on my face.

the more i think about it, i don't think they'll ever tell us, since if they admit that something's screwed up, that would justify a patch... and we wouldn't want that, now would we?

[Jambo]

I repeated the battle above and got what i was looking for: a battle which awarded the BattleScarred and RomanHero trait, but NOT the heroicsaviour ancillary (meaning that the HP's lost weren't sufficient to award the saviour).

I was never awarded the shieldbearer ancillary and therefore the > sign doesn't appear to work for this condition.

As a result, I can recommend the following GeneralHPLostRatioinBattle conditions for the following traits (as CA intended, but got wrong):

For BattleScarred/RomanHero:

GeneralHPLostRatioinBattle < 0.7

For heroicsaviour:

GeneralHPLostRatioinBattle < 0.3

For shieldbearer (the easiest to get):

GeneralHPLostRatioinBattle < 0.8

Now on to test Greg's Good/Bad Farmer and Good/Bad Trader traits... :)

Regards

[Jambo]

Ah, well spotted mate! Maybe GeneralHPLostRatioinBattle > 0.2 for the shieldbearer will work just fine then! ;)

[Jambo]

Remember the biggest bug of them all:

The fact that traits obtained from the battle map are giving twice their supposed amount towards the trait's threshold - this means super fast climbing of the traits, e.g. gaining the first two levels of GoodCommander, BattleScarred, RomanHero after just one battle.

To counter this, double the threshold of traits you can receive on the battlefield, particularly the combat related ones as these can really offset the game's balance in favour of the human. Alternatively, look for the next release of the CVP. :)

[therother]

The fact that traits obtained from the battle map are giving twice their supposed amount towards the trait's threshold - this means super fast climbing of the traits, e.g. gaining the first two levels of GoodCommander, BattleScarred, RomanHero after just one battle.

Sorry, trait newbie butting in, but has this been confirmed as a bug in the code, rather than in the traits file? Could there perhaps be another explanation? For instance, have you doubled the "Standard_Battle_Any_Victory_VnV_Trigger" to "Affects GoodCommander 4 Chance 100" to see if your general becomes a Great Commander (Threshold 8) after one battle? I ask as after reading a few threads about the scarred issue, I was surprised to get only the first level (Been in the Wars) after a test battle for something else (the effect of Command).

[gregab25]

Sorry, trait newbie butting in, but has this been confirmed as a bug in the code, rather than in the traits file? Could there perhaps be another explanation? For instance, have you doubled the "Standard_Battle_Any_Victory_VnV_Trigger" to "Affects GoodCommander 4 Chance 100" to see if your general becomes a Great Commander (Threshold 8) after one battle? I ask as after reading a few threads about the scarred issue, I was surprised to get only the first level (Been in the Wars) after a test battle for something else

Sorry, beyond the obvious Scarred trait bug, i had not heard about a problem with the GoodCommander or RomanHero traits. So I don't know much about them.

Briefly looking at the triggers reveals there is one trigger for GoodCommander and it seems in order.

The RomanHero Trait has 3 seperate triggers 2 with a 20% chance and one with 15% i think. So, it is not surprising that your general would receive this trait quickly.

I did not see any doubling of the triggers left by the omition of the condition Culture Type Non Roman. Which was why you got tested twice in the case of the Scarred trait.

Maybe Jambo has more information.

[Jambo]

It's particularly noticeable with the combat traits (or rather easier to test). Try a battle with a fresh naive general (no GoodCommander trait). If he wins and the odds make him elligible to receive the 1 Chance 100 in the GoodCommander trait, what you find is that he skips Confident Commander (+1) which has threshold 1, and goes straight to Good Commander (+2) which has threshold 2.

On occasion it's actually possible to get straight to Superior Attacker/Defender (threshold 4) from just one battle. The battle has to be a crushing victory with battle odds of >=0.5 to less than 1.5. According to the character_traits file, this awards GoodAttacker/Defender 2 Chance 100. Due to the doubling bug this gives 4 Chance 100 instead.

This isn't for autoresolved battles, only those fought out on the battle map.

[Pode]

Bugger. I was really hoping double tests for romans would be the explanation for the double awarding "feature". There's no good way to mod around this one. Doubling the threshold prevents the AI from leveling up appropriately, and God knows it needs all the help it can get. Halving the odds for each comabt trigger would be slightly better, since it would tend to keep AI and player generals mired down in mediocrity, but where's the fun in that?

[Jambo]

double post somehow....




EDit: may as well use this slot.

Pode, there's a cunning strategy that the CVP uses, which helps this unfortunate doubling problem. Basically, it uses the BadCommander traits as a means to combat the meteortic rise that's possible with the GoodCommander trait. Essentially, the higher a general is in the GoodCommander trait, the greater the BadCommander award he'll receive should he lose a battle. For instance, a general at level 4 GoodCommander will have a bigger fall in the trait should he lose a battle than a mediocre commander at level two would get.

[Simetrical]

The results are in. I modded the file as I said, then built a Shrine to Ceres in Arretium. The governor gained no traits. Then I built a Communal Farming—and the governor gained Grower. Bingo. Question answered.

But gregab25 might be a little confused at this. After all, he's been playing for a while with his mod, and he's been gaining GoodFarmer traits. If I'm right, then he should have been setting the trait trigger so that it only rewarded him for building the first level of farms and no other buildings—i.e., it should be extremely unlikely that he'd gain even Grower. So what gives?

The answer is that he's confused ≯ with ≤—"not greater than" with "less than or equal to." He assumed that they were equivalent, when they're not. What's the difference? None, if you're only dealing with elements in an ordered set. gregab's belief was that both terms represented numbers, so that's fine—all real numbers are elements of the same ordered set. If a real number isn't greater than another, it has to be either less than or equal to.

But if you read my (admittedly a bit technical) explanation of how the condition in question actually functions, you may have realized that not all elements in question need to be in the same ordered sets. If the ordered set in question is hinterland_farms and I build a dirt road, the road isn't greater than or equal to the farm—but it's not less than it, either, since they aren't in the same ordered set. ~(Roads ⋚ farms) in the same way that ~(i ⋚ 42). (Dang, I wish Unicode specified a "not greater than, less than, or equal to" symbol.)

gregab25 therefore thought that the ≺= operator was bugged, so he changed it to NOT >. Unfortunately, this made every non-farm building increase his traits, tricking him into thinking that his interpretation of EDCT.txt was right.

Maybe you just made a typo but to be sure i'll correct you. According to my approach, building any nonfarm building equal to or LESS THAN the farm level would cause the governor to gain a level.

Of course, you're right. Typo.

Sorry, trait newbie butting in, but has this been confirmed as a bug in the code, rather than in the traits file? Could there perhaps be another explanation?

I don't think so. There aren't any duplicate triggers that could explain the observed phenomena.

I ask as after reading a few threads about the scarred issue, I was surprised to get only the first level (Been in the Wars) after a test battle for something else . . .

It's not certain that you'll gain any Scarred trait at all after battle—most trait triggers have a percentage chance associated with them. The trigger could have been checked twice, but only returned positive once.

-Simetrical

[gregab25]

the more i think about it, i don't think they'll ever tell us, since if they admit that something's screwed up, that would justify a patch... and we wouldn't want that, now would we?

Hehe, are you being tongue'n'cheek tai4ji2x? I'm not that cynical.

Besides didn't your momma teach you can catch more flies with sugar?
Hehe.

[Simetrical]

If we have an arithmetic comparison, the two sides of the comparison must be members of ordered sets. According to gregab25's interpretation of the game files, in "SettlementBuildingFinished >= farms", "SettlementBuildingFinished" and "farms" are both variables that the game sets as numbers depending on the situation. "SettlementBuildingFinished" would become the number equal to the level of the structure built in the trigger condition, while "farms" would become the number equal to the level of the farms in the settlement.

But Sinner raises an objection. "Farms" is already used by the game as a building name. That is, the first level of the hinterland_farm tree is known as "farms". Building names are evidently valid arguments in export_descr_character_traits.txt, as evidenced by all sorts of triggers. gregab25's interpretation would require that "farms" be used in EDCT.txt as an entirely separate variable—but how would you check anything relating to the structure "farms" if the game reserves that word for a special purpose?

The answer to this is that even if "farms" is reserved for special use, the SettlementBuildingFinished condition under gregab25's interpretation wouldn't accept a building name as an argument. Therefore, the game would know not to parse the term as a building name, but rather as a special reserved term. In this, it would be distinct from SettlementBuildingExists.

So what's Sinner's approach? Seemingly, the same one I stated before: that the variable SettlementBuildingFinished returns the building name, not the level number, and ">= farms" means what you'd think it means, that the previous argument is checked as to whether it's at or beyond the level "farms" in the ordered set of hinterland_farms. Thus, the trigger would work if and only if the building built was a farm of some description.

It should be very easy to test this. First of all, set the threshold of Grower to 1. According to gregab25's approach, building any nonfarm building equal to or greater than the farm level would cause the governor to gain a level of GoodFarmer; according to Sinner's, it wouldn't. I'll run the test in a minute. I'm betting that Sinner's right, but we'll see.

-Simetrical

[gregab25]

According to gregab25's approach, building any nonfarm building equal to or greater than the farm level would cause the governor to gain a level of Growing

Maybe you just made a typo but to be sure i'll correct you. According to my approach, building any nonfarm building equal to or LESS THAN the farm level would cause the governor to gain a level.

Not greater than... EQUAL OR LESS THAN.

[gregab25]

BTW, I still cannot believe you guys are still so vehemently argueing without running the tests first. That's pure folly!

Run the tests first and I'll be happy to admit that I am wrong if I am!

Cheers.