When do we get a warlords epic?

[T-hawk]

(And also World of Warcraft)

Nooooo! I've lost more RL friends to that game. They go in... and they Don't. Come. Out. Ever.

[Iustus]

Are there other show-stopping problems for which a patch is needed?

The biggest problem I have seen is the Trebuchet unit seems a bit too powerful, for its time. It is the unit to use to take cities that contain longbowmen. It relegates macemen to escort units. It is to medieval combat what modern armor is to modern combat.

That said, it is not a god unit, useful against everything. It just becomes the main unit you use to take cities, rather than macemen.

I am not sure I would call this a showstopper. It probably needs some tuning, but the game is completely playable.

---

A couple other minor issues which I do not think are showstoppers:

- If a city using the governor goes into starvation, the governor gives up on food and maximizes production, causing the city to starve faster and lose more than the one population point it had to lose

- AIs will seem to manipulate the vassel state mechanics to get a 'big brother' to join a war when they are decisively losing, creating a domino effect of constant war until you either accept one as a vassel state yourself (ending the war) or make peace by other means. On occasion, you can get into wars with allies who accept vassalege from those you are at war with. AIs should probably only take a vassel who is at war if they are willing to declare war with that civ themselves.

-Iustus

[ThERat]

AIs will seem to manipulate the vassel state mechanics to get a 'big brother' to join a war when they are decisively losing, creating a domino effect of constant war

I played that Mongol scenario and honestly, that domino effect was really annoying. It felt almost like Civ2 where you would eventually fight all Civs. You would attack a Civ until it is pretty small, it would become a vassal of another bigger nation which in turn would declare war on you immediately. Stupid feature really...I was so annoyed, I didn't touch that scenario any more.
I don't know whether I would be that bad for a 'normal' game in warlords, but in theory it can happen there as well.
So, I think a patch for vassal states is surely needed.

[Iustus]

I wanted to point that out, as the major complaint of vassel states. That said, vassel states are an option that you can turn off. Even so, it might make sense to do an adventure with vassel states on, just to see how many people run into the issue.

If you are willing to conquer by accepting vassel states, rather than completely eliminating a civ, this also becomes a non-issue. So a varient where you have to accept captitulations would also eliminate this issue.

-Iustus

[Kylearan]

Hi,

And also World of Warcraft

It has been nice to know you, Sirian. Good bye, and see you again maybe in a couple of years...

(Like T-hawk, I've lost a RL friend to WoW. One of my closest friends, whom I've known for over 20 years, and who I haven't seen or even heard from for over 1,5 years now thanks to that damn game. )

-Kylearan

[Rowain]

Hi,


It has been nice to know you, Sirian. Good bye, and see you again maybe in a couple of years...


-Kylearan

i doubt it WoW is far too much easy-mode to hold Sirian for a long time.


About Warlords:
Besides the bugfixes (which should go into Vanilla too) Warlords has 2 great advantages:
a) Chariots are now a cheap and very good Counter to Axes
b) Cossacks and Redcoats are brought inline and no longer the overpowered units they are in original CIV

other huge change: Normal Barbs are similar to raging Barbs in vanilla(I guess to make the Great Wall usefull)

Trebs: well they are good for attacking Cities but weak everywhere else(base strengh is 4 and costs 60 hammers Cats have str 5 and cost 40).

About vassalstates: I had my enemy turn to a vassal of my ally but this did not put me into war against my ally it forced me into peace with my enemy.

As far as I understand it only if your Ally becomes the vassal of your enemy you have to fight 2 AI instead of 1. (The vassal adopts the war/peace standing of his master not the other way round).
Personally I do find this quite good. No longer can you pay a weaker Civ to attack your enemy (to remove his forces from your borders to attack him then) without a risk.

[Iustus]

About vassalstates: I had my enemy turn to a vassal of my ally but this did not put me into war against my ally it forced me into peace with my enemy.

As far as I understand it only if your Ally becomes the vassal of your enemy you have to fight 2 AI instead of 1. (The vassal adopts the war/peace standing of his master not the other way round).

Hmm, that is not my experience, so I checked the code in the warlords SDK.

This is the actual agorithm:

Code:

the vassal declares war on everyone the master is at war with

AND

if the vassal capitulated, then the vassel makes peace with everyone it was at war with

else (the vassal voluntarily sought a master), then the master declares war with everyone the vassal was at war with

What this means is that what I originally said was true:

If you are at war with Examandia, they will offer to capitulate to you, to become your vassal. If you accept, then they will make peace with everyone they are at war with, and make war with everyone you are at war with.

Otherwise, they will eventually try to find another master, lets say Whyatia, who will declare war with you if they accept Examandia as their vassel.

This is what causes the endless war domino effect.

I will look more into what factors the AI looks at when deciding whether to accept a voluntary vassal in the next installment!

-Iustus

[Iustus]

As promised, I looked into that very condition.

It turns out it is a rather simple problem, I think...

When an AI is deciding whether to accept a voluntary vassal, it does make a check, to confirm it is willing to declare war on everyone with which the vassal is at war.

The problem is that it specificly does not do the power comparison.

That is to say, it specifically omits the power comparison when deciding if it is alright to go to war.

This basically means that:

if (potentialVassel.Attitude > potentialMaster.DeclareWarRefuseAttitudeThreshold)
AND
if (potentialWarOpponent.Attitude <= potentialMaster.DeclareWarThemRefuseAttitudeThreshold)
then
AcceptVassel

Well, some leaders, such as Alexander, have a very low threshold for war:

Code:

<DeclareWarRefuseAttitudeThreshold>ATTITUDE_ANNOYED</DeclareWarRefuseAttitudeThreshold>
<DeclareWarThemRefuseAttitudeThreshold>ATTITUDE_PLEASED</DeclareWarThemRefuseAttitudeThreshold>

Which means they will almost always accept a vassal that brings them into war.

The easy 'fix' is to instruct the comparison to take power into account. This is a trivial change, changing one 'false' into a 'true', but it still will not take into account multiple wars. Each war is compared separately, the combined power of everyone the potential vassal is at war with is not compared vs ours, just our power vs each separate opponent. (Note, for this, vassals and permenant alliances do count all as one opponent).

easy fix:

Code:

if (GET_TEAM(eMasterTeam).AI_declareWarTrade((TeamTypes)iTeam, getTeam(), [b]true[/b]) != NO_DENIAL)

The best fix is probably to correctly compare our power vs the sum of the power of everyone the potential vassal is at war with, perhaps modified by some fraction (ideally from the leaderhead, but that might involve adding a new field). This is actually not that hard to do, but would be obviously a bigger change.

best fix:

Code:

bool bAccepted = true;
[b]int iSumDefensivePower = 0;[/b]
TeamTypes eMasterTeam = GET_PLAYER((PlayerTypes)iI).getTeam();
for (int iTeam = 0; iTeam < MAX_CIV_TEAMS; iTeam++)
{
    if (iTeam != getTeam() && iTeam != eMasterTeam && atWar(getTeam(), (TeamTypes)iTeam) && !atWar(eMasterTeam, (TeamTypes)iTeam))
    {
        if (GET_TEAM(eMasterTeam).AI_declareWarTrade((TeamTypes)iTeam, getTeam(), false) != NO_DENIAL)
        {
            bAccepted = false;
            break;
        }
[b]        else
        {
            bool bLandTarget = AI_isLandTarget((TeamTypes)iTeam);
            
            // do not use getDefensivePower, that would count vassals twice
            // could add an assert here to confirm that every vassal of iTeam is also at war with us
            iSumDefensivePower += (GET_TEAM((TeamTypes)iTeam).getPower() / ((bLandTarget) ? 2 : 1));
        }[/b]
    }
}

[b]if (bAccepted)
{
    // could add a leaderhead based modifier to iSumDefensivePower here, if desired
    if (iSumDefensivePower > getPower(true) + GET_TEAM(eMasterTeam).getPower(true))
        bAccepted = false;
}[/b]

if (bAccepted)
{
    GC.getGameINLINE().implementDeal(getID(), ((PlayerTypes)iI), &ourList, &theirList);
}

And that is probably way more C++ than anyone wanted to see!

-Iustus

[Rowain]

Ah If the Vassal seeked a master voluntarily then his master joins the war!

So in my game(s) he was forced into vassalage by my ally which resulted in peace and in yours( and others games) they sought a Master to protect them and so the master declared war to save his new vassal.

Thanks for clarifying that.

The first report I read about this mentioned an ally which turned to be an enemy guess it was only a peacetime-ally then.

[Ug the Barbarian]

Update from Civfanatics website is particularly relevent:

Warlords Patch Coming in October
posted by Thunderfall at Sep 15, 2006 02:09 AM
Firaxis software engineer Alex "alexman" Mantzaris commented on the forum yesterday that a Civ4 Warlords patch is in the work but it likely would not come out before mid-October, due to scheduling conflicts with Quality Assurance (QA).

He also said that vassal state bugs will be fixed, but don't expect major implementation changes like being able to drop your vassal state, which is working as designed. One patch change to the vassal system is that "the new master will only accept the vassal if he also wants to declare war on you".