SGOTM 07 - Murky Waters

Erkon · Apr 26, 2008

LowtherCastle said:
My mind is running rampant. Here's another thought. In our Alpha bee-line, we get Writing many turns before the AIs. AIs far to the South will be trapped so we are free to steal our 2nd (...and 3rd? (...and 4th )) worker without fear of reprisal, especially if we can decide which AI to attack early on. We can pick on a low-commerce AI, for example.

Now tell me what I'm forgetting.

EDIT: Gnejs when were you completing Writing relative to the worker build, when you went warrior-wb-wb-wkr?

You forget that all workers will be escorted :cry:

But if the opportunity arise, I thinks it's a good idea (against low-commerce AIs).

LowtherCastle · Apr 26, 2008

Erkon said:
Will a camped warrior prevent forest creep?

Will forest ever creep to a tile with unrevealed horsies?

Mais oui, Monsieur.

Don't you remember the Smurkz SG6 complaints about their jungle-crept rice? It's one of the few times I remember that Gyathaar, as map creator, felt obliged to address players' concern with code and explanation.

klarius · Apr 26, 2008

LowtherCastle said:
My mind is running rampant. Here's another thought. In our Alpha bee-line, we get Writing many turns before the AIs. AIs far to the South will be trapped so we are free to steal our 2nd (...and 3rd? (...and 4th )) worker without fear of reprisal, especially if we can decide which AI to attack early on. We can pick on a low-commerce AI, for example.

Now tell me what I'm forgetting.

It's by no mean certain that the AIs in our game are such slowpokes. In fact even on my map I had the case once that Gandhi offered open borders before I had writing.

We also have to know first who our neighbors are. There are some who like dogpile wars very much (even in the overall friendly category).

LowtherCastle · Apr 26, 2008

Erkon said:
You forget that all workers will be escorted

But if the opportunity arise, I thinks it's a good idea (against low-commerce AIs).

I remember lots of unescorted workers baring their sweaty muscles at my drooling warrior. And speaking of escorts, we wouldn't even have to escort any workers home...

LowtherCastle · Apr 26, 2008

klarius said:
It's by no mean certain that the AIs in our game are such slowpokes. In fact even on my map I had the case once that Gandhi offered open borders before I had writing.

We also have to know first who our neighbors are. There are some who like dogpile wars very much (even in the overall friendly category).

Fair enough. Even if we just steal 1 worker right before or at Alpha, it stills saves us the trouble of building it and it's good timing for pre-chopping or chopping the GLib, right?

Erkon · Apr 26, 2008

LowtherCastle said:
Okay, let's take it a step farther and turn it into a 'live apprenticeship.' Whoever gets to Alphabet immediately uploads the save (let's just hope we don't get any fkedup invitations to trade a tech in the IT) and we discuss/learn how to adress the situation, step by step, trade by trade, deal by deal (with every single step followed by a save and upload to the thread)...

I like, I like. Very MM'ish. I would then prefer everyone to do their homework and think how they would have addressed the trades once they get to see the save, else I'm afraid klarius will present a complete plan that is flawless and the peanut gallery just nods :lol:

Perhaps it will take a long time, which may be a drawback.

One other significant drawback is that we must ensure we don't accidentally cheat (i.e. accidentally perform any irreversible actions).

Is there any drawbacks to deny a trade IBT? They won't demand any tech, so the trade would be repeatable? I guess there are factors, but are they significant?

klarius · Apr 26, 2008

LowtherCastle said:
Mais oui, Monsieur.

Don't you remember the Smurkz SG6 complaints about their jungle-crept rice? It's one of the few times I remember that Gyathaar, as map creator, felt obliged to address players' concern with code and explanation.

Well, but forest will not creep on horses.
It's only possible if the resource can naturally appear on the feature.
Rice can be in jungle, but horses can't be in forest.

Code:

    if ((getBonusType() == NO_BONUS) || (GC.getBonusInfo(getBonusType()).isFeature(iI)))

LowtherCastle · Apr 26, 2008

Erkon said:
I like, I like. Very MM'ish. I would then prefer everyone to do their homework and think how they would have addressed the trades once they get to see the save, else I'm afraid klarius will present a complete plan that is flawless and the peanut gallery just nods

Perhaps it will take a long time, which may be a drawback.

Well, we could just set a time period when anyone who wants to joins in and we do it fast. I think we're all in pretty much the same time zone (not sure about Morpheus11).

I'm hoping klarius will present his flawless plan and the peanut gallery just nods... :lol:

. just kidding. I get your point. That was implicit in my idea. You're not going to learn as much if you don't try the exercise yourself first. I'm sure klarius will be so kind as to put his plan in spoilers...

LowtherCastle · Apr 26, 2008

klarius said:
Well, but forest will not creep on horses.
It's only possible if the resource can naturally appear on the feature.
Rice can be in jungle, but horses can't be in forest.

Code:

if ((getBonusType() == NO_BONUS) || (GC.getBonusInfo(getBonusType()).isFeature(iI)))

Works for me. :goodjob:

Maybe that's what's messing jesusin up. Did you hear that? Your horse can't appear in the forest, only out in the open. :cool:

klarius · Apr 26, 2008

I just tried a pyramids variant on klarius-LC map.
Gandhi was first to writing (T37) and Hatty declared T47.
4 axes, 2 spears and 2 archers coming from India against my 1 warrior before I even made OB with her.

LowtherCastle · Apr 26, 2008

I got the pyramids pretty fast but ALpha was too late to get a DoW with Hatty. There are a lot of choices for how to play the pyramids. Not sure what's best.

Erkon · Apr 26, 2008

I have a problem. With klarius korner. Every second post from klarius contains nuggets of wisdom. What shall I do? Link them all (as I started to do) or copy all relevant text from klarius into a bigger post? The latter requires less effort from me, while the current system enables tracking directly to the context. Please advice me...

Now to something completely different: I read somewhere a post that described the inherit friendliness between AI. It was a picture with different colors for each AI and they were sorted in columns from -2 to 7 or something. Does anyone remember that post? I'm afraid it was lost in the crash a while ago / transition to a new server. I read it in a GOTM-thread (spoiler/pre-game/results). The post explained why certain AI are annoyed with each other at +1 (regardless of random element).

klarius · Apr 26, 2008

Erkon said:
Now to something completely different: I read somewhere a post that described the inherit friendliness between AI. It was a picture with different colors for each AI and they were sorted in columns from -2 to 7 or something. Does anyone remember that post? I'm afraid it was lost in the crash a while ago / transition to a new server. I read it in a GOTM-thread (spoiler/pre-game/results). The post explained why certain AI are annoyed with each other at +1 (regardless of random element).

Well, I don't remember this post. But I did a spread sheet to this effect for SGOTM 3.
I will update this once we know which leaders are our neighbors.
I attach the old version

.

Gnejs · Apr 26, 2008

LowtherCastle said:
I got the pyramids pretty fast but ALpha was too late to get a DoW with Hatty. There are a lot of choices for how to play the pyramids. Not sure what's best.

Made a try on klarius modified map, never working or trading the horse as usual.

Pyramids T39 via heavily emphasizing production: worker-warrior-Pyramids-Library. Research was Masonry-Wheel-Hunting-AH-Writing (T44)-Alpha (T63)
Nobody had Alpha, Cyrus didn't have writing. T64 pops a GE. T66 Gandhi has Alphabet.
PA with Hattie 640 AD (T142). She is the worlds score leader by a huge margin having 6 cities. We have a huge tech lead: Constitution, Liberalism, Printing Press, Astronomy, Chemistry, Corporation, over the closest rivals on our continent. Everybody else is even more backwards.
When emphasizing science I bring in 419 beakers in Beijing, emphasizing production gives 360 beakers and 36 raw hammers.

Some observations:
Delaying Alphabet causes the research pace of the whole continent to slow down. Even though I was almost 20 turns late I could still trade for all the basic techs. I could not bribe Hattie into a common war though, but it was no problem to get the 40 war turns in time for Communism anyway.

Conclusion:
I am not impressed by the benefit of the Pyramids. The happiness bonus is temporary and doesn't matter after Globe Theatre. The research bonus doesn't seem to make up for the 18 turn delay of Alphabet, and we will anyway get Constitution for Representation when we get the PA or soon thereafter. The early culture boost is nice though.

LowtherCastle · Apr 26, 2008

Do we know when we get PA using klarius' modified game without the horses?

Erkon · Apr 26, 2008

There are two reasons for sticking to the current PPP (no pyramids). The first is the risk aspect, such as the risk of someone DoWing us, or the risk to miss the pyramids. Even if the risk is manageable, the consequence from either is that we may miss out on laurels.

The second reason is the timing. I think beakers earned early are much more valuable than beakers earned late. I've updated post#4 with a Turn vs Date list to show the year for each turn. Assume we run 6 specialists on average from 1000 BC to 1200 AD (constitution), then we gain 105*6*3=1890 base beakers = ~4000 real beakers, which is very close to the cost of Communism, and equals approximately 10 turns of research in Gnejs' game. I don't think that is enough, considering the overall slower tech pace.

My conclusion is that we stick to the PPP unless someone comes up with more favorable results than Gnejs.

LowtherCastle · Apr 26, 2008

Well, 640 ad is a good result. So beaker up or down, something worked very well in Gnejs' approach. klarius' 450ad result was with the horses. So it strikes me as more complicated than just beaker analysis. One diffference that stands out to me is that the current PPP approach appears to rely more on trading with AIs so that when we get our PA, other AIs also have CHemistry. We beat them down only because we know how to use them better. Gnejs' approach appears to slow down their research early on and not be so dependent on them toward the end. That's intriguing.

Erkon · Apr 26, 2008

Erkon said:
Regarding the AH breakpoint: I think we can accomplish this with LC's Mid Turn Break (~T21-22?). It's too far into the game for Gnejs. There are other breakpoints that are of similar importance, such as learning Writing and Alphabet. We will need to stop at these three occasions anyway, but I can't manage to fit the turn set lengths to them , so we'll have to rely on Mid Turn Set updates...

Doh! I just noted that AH&Worker is on T20. This is of course very suitable for a breaking point for first turn set :blush:

This would also give LC a better opportunity for planning his MM & worker action.

In my test game with my map I get Writing on T29, Library on ~T42 and Alpha on T45 (pop5, one angry)

Gnejs · Apr 26, 2008

LowtherCastle said:
Do we know when we get PA using klarius' modified game without the horses?

Just tried this. Started great, was able to trade for basically everything without giving up Alpha. But then things went bad. First of all there was no religion founded on our continent until Confu, which almost everyone had except Hattie who founded christianity. No pacifism, less great people. Even more annoying was the fact that I popped 4 artists... :mad:

Got the PA very late, around 820 AD. Random also had it that Hattie had 5 cities this time.

LowtherCastle said:
Well, 640 ad is a good result. So beaker up or down, something worked very well in Gnejs' approach. klarius' 450ad result was with the horses. So it strikes me as more complicated than just beaker analysis. One diffference that stands out to me is that the current PPP approach appears to rely more on trading with AIs so that when we get our PA, other AIs also have CHemistry. We beat them down only because we know how to use them better. Gnejs' approach appears to slow down their research early on and not be so dependent on them toward the end. That's intriguing.

Here is a save a couple of turns after the PA, when the research has catched up.

View attachment s7t AD-0680.Civ4SavedGame

This time the Pyramids strategy gave the best result. But there is so much random in these results that I don't know what to believe. Could anyone else run a couple of games and see how the two strategies compare? Perhaps we should also modify another of the test maps.

PS. Both these games are a little bit sloppy once I have reached Alpha. There is definitely room for improvement.

Gnejs · Apr 26, 2008

Erkon said:
Doh! I just noted that AH&Worker is on T20. This is of course very suitable for a breaking point for first turn set This would also give LC a better opportunity for planning his MM & worker action.

Are you saying I should go 20 turns now? :crazyeye:

SGOTM 07 - Murky Waters

Deity

Deity

... !

Deity

Deity

Deity

... !

Deity

Deity

... !

Deity

Deity

... !

Deity

Deity

Deity

Deity

Deity

Deity

Deity

Similar threads