AI unblamable for Mictlans early dominion death !
 

As I just had to find out, the AI is completely unblamable for Mictlans early dominion death, which occures all to often before turn 10 !!!

To see why, check this screenshot:

http://www.shrapnelcommunity.com/thr...oDomSpread.png

Means - your only source of dominion is blood sacrifice.
What gives ONE temple check per slave sacrificed.

Normally, a nation gets dom spread from its prophet, its temple and its home province. I ran a test without prophet, only Dom-5 and capitol and temple therein (imprisoned pretender).
Effect: dom strength went 2,3,3,4 in the next turns.

Mictlan (LA) with a holy-1 priest set to do sacrifice the 1 blood slave the nation gets from its capitol site, the dominion strenght went as follows:
1,1,1,1,1,1

Even if the holy-1 priest or the sacrifice is not bugged, this is considerably unbalanced: To get the same number of checks, Mictlan would need to sacrify 3+ slaves per turn in the capitol alone - using a holy-3 priest and a blood hunter. But in a 'narrow' environment, this takes too long to set up, especially for the AI, what results in unnecessary early dominions death. Not to mention the fact that hunting for 2+ slaves per turn in the capitol will be thoroughly detrimental to your gold income ...

Only when I started to use a holy-2 priest and sacrify 2 slaves per turn, the dom strength started to rise: 1,2,2,2,3 ... .

I doubt, though, that the AI is aware of this - it does as its told, sets a priest to sacrify the 1 slave it gets .. and dies 8 turns later.

Re: AI unblamable for Mictlans early dominion dea
 

So, home province has conversion rate of 0? I thought that Mictlan was supposed to get *at least* Dominion spread from that.

Re: AI unblamable for Mictlans early dominion dea
 

This is ALL explained under Mictian on page 93 of the manual.

Basically EA/LA Mictian's only source of dominion spread r pretender (1/2 as effective as a normal pretender) and temple sacrifices, 1 sacrifice per priest level.

This is not bugged, but themed and has early drawbacks but could be a game winner if used correctly later on.

I see no problem with it in MP, anyone choosing Mictian should have an effective strategy to spread Dominion with blood sacrifice.

In SP, well the AI has issues so needs looking at, until then in SP I suggest people avoid playing against EA/LA Mictian.

Re: AI unblamable for Mictlans early dominion dea
 

Those who can read ... . The thread is about the AI ... and did I say it isn't in the manual?

But that it is written in the manual, does it make it absolute, correct-without-a-doubt? Btw., the game "was there first", and the manual is wrong on details sometimes anyway ... .

The problems are the following:

- holy-1 priest seem only able to 'hold' the dominions strength level at 1. With only 1 priest allowed per temple, this means you'll need a 2nd temple in a 2nd province to get more dominion (spread). More likely, an enemy dominion will have killed you by that time.

- the LA Mictlan site gives only 1 bloodslave, while the old site (Dom2 and EA) gives 3 IIRC => blood hunting needed from turn 1

- the AI will most likely use the first priest it buys, but not necessarily the prophet. Looks like it ends up with a holy-1 priest 90% of the time, and therefore dom strength 1 until it has conquered another province and build a temple there. And maybe, just maybe a 2nd holy-1 priest sacrificing there (where does he get his blood slaves from, though?) will be enough to push its dom strength to 2+ ...


Ergo: Prophet and home province should spread dominion for Mictlan (or similar blood nations) as well. Thematically it's hardly understandable that it isn't that way anyway. For players it will make little difference, as they have little trouble to optimize their MM to make sure they have enough dom strength, at least in the beginning. But for the AI it will be the difference between life and death within the first 8 turns !!

Re: AI unblamable for Mictlans early dominion dea
 

Interesting conclusion, Arralen. A radically different conclusion might be: "In pretender design, EA and LA Mictlan AIs should always choose an awake pretender with a decent strength of dominion"

After all, that would have the effect desired: more and better dominion checks in the home province in the early game for Mictlan, thus preventing dominion-death, and it would have that effect without changing any rules whatsoever but just adding a few guidance rules to the AIs pretender design.

(More generally, if it was possible to set "guidance rules" for pretender design on a nation basis and made possible to mod this, we'd probably end up with much more capable AIs in general than we have now)

Re: AI unblamable for Mictlans early dominion dea
 

Combine this with "Mictlan is specifically designed for big bless effects - and suck without it" and what do you have ? A contradiction !! [img]/threads/images/Graemlins/Cold.gif[/img]

Re: AI unblamable for Mictlans early dominion dea
 

PDF says:-
Combine this with "Mictlan is specifically designed for big bless effects - and suck without it" and what do you have ? A contradiction !!

OR

A strategic choice...

Re: AI unblamable for Mictlans early dominion dea
 

What strategic choice ? Risking dominion death with even a dormant pretender, or not being able to put out an effective bless strategy ?
Plus the fact that such a mm level in slave sacrifice look awfully bad to me..
Please do stg, make the Prophet spread dominion, else of what use is a prophet ?

Re: AI unblamable for Mictlans early dominion dea
 

Actually there is a pretty reasonable compromise, try an awake dom 9 f9w9 moloch. Your scales will be very bad, but don't try to spread dominion a lot and it won't hurt that much. And you will expand at lightning speed between your pretender from turn 2 on, and a new expanding squad of sacreds every turn.

Re: AI unblamable for Mictlans early dominion dea
 

Yes I agree, just run a few test pretender/scales and Mictian IS a awesome nation to play.

U do not after compromise, u can have dominion 10 (every temple check is at 100%, ie...automatic success), a lvl 9 bless, with two lvl 4 blesses with scales of turmoil 3, sloth 3, heat 3, death 3, fortune 3 (avoid misfortune with Mictian, it could kill u), magic 3 with a awake pretender!?!?!

U can do what quantum says, alternatively u could build temples/priests and try to inflict yr bad dominion on yr neighbours, cancelling out there good dominion, yr blessed armies could follow up behind yr dominion...

With strong priests lvl 3 and lvl 2 Mictians need for blood sacrifice is not much of a weakness and becomes very powerful mid/end game.

Now Johan Karlsson needs to explain this to the AI... http://forum.shrapnelgames.com/image...es/biggrin.gif

Re: AI unblamable for Mictlans early dominion dea
 

Great: Exactly ONE choice of pretender/playstyle - thats exactly what we all want.

And its even one the AI cannot really do at all atm. And to remind everyone - I started this thread to talk about the AI and why Mictlan usually dies before turn 10 if played by it. http://forum.shrapnelgames.com/images/smilies/wink.gif

Re: AI unblamable for Mictlans early dominion dea
 

You kind of gave off a different image, there. By saying the AI's 'unblamable', it sort of implies the players couldn't do much better.

Re: AI unblamable for Mictlans early dominion dea
 

And I recognized the issue as you described it and explained how a simple change in AI pretender design would void the issue without having to change the underlying game mechanics. http://forum.shrapnelgames.com/images/smilies/happy.gif Now, if you want to change the mechanics out of principle, that is one thing, but it isn't needed, when the real problem isn't the mechanics but the AI's playing capabilities, and the latter can be remedied without too much work.

Now, if the AI in general designed kick-*** pretenders, and this would absolutely screw them up, worsening their situation because they really need those extra points from being dormant or awake to be competitive, there might be an issue; However, as the AI is extremely incompetent in pretender design as is, with regards to choice of pretender, scales, magic, and basic dominion, that hardly seems to be the case.

[I mean look at it - even if you set an AI to impossible such that it has points coming out of the wazoo, it still takes bad scales for the nation, sometimes to the point of crippling it, bad paths, and the wrong choice of pretender for the strategy a human player would have pursued if he had been silly enough to choose those paths in the first place, not to mention nonsensical choice of awakening].

As for those talking about players - well, players just have to make a strategic choice - something that the AI is chronically incapable of in Dominions 3.

Re: AI unblamable for Mictlans early dominion dea
 

I think the solution is simply for the home province to generate dominion. Not even the prophet needs to do so. It makes sense thematically because it's holy ground and is the only place dominion spreads from besides the god and blood sacrifices. You can also rationalize it by saying that the magic site itself in the Mictlan home province radiates the power of blood and so spreads dominion. Good grief, a change like this is simply not disloyal to the Mictlan legacy, especially if you remember that in the past, Mictlan priests all had higher levels and were more effective at blood preaching.

I said there would be some Mictlan dominion headaches in another before the game even came out.

=$= Big J Money =$=