Experiences With "Flavors"?

[Ozymandias]

Ok so an issue with not disabling tech trading skewed my results... Indeed 0% means the civ will avoid it and research it last, and 50% means no effect, so hats off to Plotinus

Many thanks to you all! although the bit that I highlighted above is rather disappointing ...

... Quite awhile back, I started musing on what modern militaries might look like if certain Civs hadn't gone extinct. I noticed certain similarities throughout history - where loose versus close order formation militaries developed (e.g., steppe bow-armed cavalry and hoplites fighting in narrow mountain passes) and had hoped to guide certain militaries in different directions (large continental powers contesting the same mass - say, Eurasia - tend not to build aircraft carriers; seafaring powers never deployed the sort of "land behemoths" that the USSR did pre-WW2). So flavors won't ultimately help me there, which is a pity, as this would have been a far more elegant solution than combining Era=None techs with unit producing Improvements etc.

C'est la guerre.

Best To All,

Oz

 

[dexters]

how do we populate the flavors.

ie: I'm trying to mod 'ALL' AI civs to build a certain building, I populate the left column as flavour 'allCiv' 100% relation to flavor 'SuperBuilding' then flag my building as superbuilding flavour?

or is it the other way around? ie: superbuilding flavor on left column 100% relation to AllCiv *flavor

 

[Ozymandias]

I'm trying to mod 'ALL' AI civs to build a certain building, I populate the left column as flavour 'allCiv' 100% relation to flavor 'SuperBuilding' then flag my building as superbuilding flavour?

That is correct.

Best,

Oz

 

[DJ Bonebraker]

....Although that doesn't always work out... I have several buildings in the FF mod flagged as "AI" (and made all AI factions and other AI Flavors have the AI Flavor or be 100% good towards it), and sometimes they STILL don't build the damn things...

...Sometimes I think AI stands for "Absolute Idiot" in this game.

 

[wolf_brother]

So, in conclusion... the Flavors function seems to be worthless Correct?

 

[tom2050]

Here is a bit of what I have observed, and what seems to be happening during test plays... of course, tough to tell if it's a driving force or not in the game.

It seems that flavors only real use is when a tech is not required to advance, therefore you can make some civ's research it, and some will not ever. I suppose it works for normal tech branch as well, to make some civ's research certain techs instead of others, but it also seems that they will ignore the flavor (since most tech's seem to over-ride flavor settings unless the tech is empty).

It seems flavors for improvements works a bit, but even 100% does not make a civ build that structure right away, they may still wait quite a bit, building other things instead like units. 100% building flavor probably gives that building a very slight higher chance to be built instead of another building, but it seems the AI will not consider a building at all if it needs to build units, workers, or settlers, etc.

Tom

 

[Ozymandias]

Flavors aren't worthless, but all they really seem to do to effect gameplay is to prioritize which of several Techs an AI Civ will research first. Combining this with Alexman's weightings ( http://forums.civfanatics.com/showthread.php?t=45559 ) it should be possible to steer the AI, to some limited but critical degree, in its research decisions.

IMHO the worst bummer is that "0%" simply means the AI will simply research a Tech last instead of not at all.

Best,

Oz

 

[tom2050]

IMHO the worst bummer is that "0%" simply means the AI will simply research a Tech last instead of not at all.

Best,

Oz

True, but I don't believe the AI will research a tech with 0% if it is not required for advancement. I was under the impression that if a tech required to advance had 0% the AI is forced to research it so it can progress through the game.

Tom

 

[Ozymandias]

True, but I don't believe the AI will research a tech with 0% if it is not required for advancement. I was under the impression that if a tech required to advance had 0% the AI is forced to research it so it can progress through the game.

Tom

Ah! Now that's good to know

 

[tom2050]

And just to add, if a tech IS required to advance to the next era, and has a flavor of 0%, it will not be researched until absolutely last. In other words, all other possible techs will be researched first until the AI is forced to research the 0% tech in order to progress along the tech tree.

I remember the discussion on that somewhere in threads from long ago, but have no idea where that thread is now ... (since I'm the complete opposite of an Archivist of any kind, I get lost in my own neighborhood).

Tom

 

[Virote_Considon]

I should add that even with a 0% flavor and not being required to advance the next era, the AI will still research it (first!) if it deems the tech is useful enough. I tell you, it's hell trying to get the right civs "naturally" into the right governments!

 

[Ozymandias]

I should add that even with a 0% flavor and not being required to advance the next era, the AI will still research it (first!) if it deems the tech is useful enough. I tell you, it's hell trying to get the right civs "naturally" into the right governments!

That's why I referenced Alexman's formula above - I think the two have to be used in tandem. E.g., AI is presented with 2 worthless Techs, which it evaluates as having equal value; hopefully Flavor will be the tie-breaker, and the Tech after that will offer something valuable, making the AI research it before the other 0-value Tech it hopefully ignored.

Best,

Oz

 

[dexters]

Research is an interesting discussion, but what I'm trying to tie flavors to is to buildings.

 

[Virote_Considon]

Hmm... That reminds me, I don't seem to recall there being any "which improvement will the AI build" tests... Which improvements they'd build would likely influence the AI in the same way as research weights...

 

[Ozymandias]

Research is an interesting discussion, but what I'm trying to tie flavors to is to buildings.

Indirectly: my "steering" approach could be used to channel the AI to a Tech that allows whichever bldgs to be built.

-Oz

 

[Ozymandias]

Hmm... That reminds me, I don't seem to recall there being any "which improvement will the AI build" tests... Which improvements they'd build would likely influence the AI in the same way as research weights...

... Feel like doing some testing?

 

[tom2050]

That is a very interesting idea... I never had any thought of doing such a thing, perhaps some testing should be started, as I would participate. It could also give a great deal more info to how flavors really work with buildings, which could provide some great insights.

Tom

 

[Steph]

Starting with the hypothesis that "flavors really work" seems to be very optimistic...

 

[tom2050]

Starting with the hypothesis that "flavors really work" seems to be very optimistic...

The context I meant it in, was not that flavors are wonderful and joyous, and can even fill up your gas tank when you run out of fuel... but more as in to figure out how they work (take the really out of there), and if at all in situations with buildings (which they do seem to, but very little).

Tom

 

[dexters]

To clarify further, I'm modding the epic game and is trying to allow the AI to prioritize trade, especially the building of harbours.

I've run one debug game so far with CivAll tied to 50% TradeNetwork (harbours), and it seems to work, in the sense that civs which shares the harbours trait (militaristic, seafaring, commercial) will build it often, and infact, seems to build it in every city (for England which is seafaring and commercial) with the added prodding of the flavour. So the effects seem to stack. The civs that are least likely to build harbours are the ones where neither of the civilization treat are checked, so we only have the 50% flavour at work.

I'll bump it up to 100% and it should help out the AI a bit more.