[BTS] SGOTM 14 - Kakumeika

[mabraham]

Originally Posted by shulec

I will call it "closed border teleportation".

When a unit is in foreign territory when borders are closed, they are "teleported" to the nearest neutral land territory. Once the land by Ning-hsia in Genghis' land is claimed as his territory, we can put a unit on the peninsula in Elizabeths land. We close the borders, and if there is an unoccupied land tile in the mountain island, the unit will be "teleported" there.

Clever.

Originally Posted by shulec

I started an archer up that direction to consider trying that. We can also send our wandering chariot into Mansa's lands and try to teleport him to the island east of his lands.

This requires the AI to own all territory nearby. Potentially a stack can be teleported in if need be.

Hmm a whole stack sounds like a large investment for a doubtful outcome at an unknown time. If we have city vision to know when a Genghis border will pop, we'll have enough map vision to know whether there are any non-mountain tiles in that particular pocket.

[bcool]

shulec!

Sun Tzu Wu... UP NOW
mabraham... on deck
Tachywaxon... in the hole
Walter_Wolf... waiting
Kaitzilla... waiting
frogdude... waiting
bcool... waiting
shulec... waiting

[bcool]

teleportation?
I like the out of box thinking on the teleportation trick, but closing borders with the northern witches would be a big set back on our goals for diplomatic victory. I think we can wait for airships...
new wizard probably location
That tiny barbarian city on the edge of the map near Monty does look promising for a wizard location, good find and suggestion on the workboat mabraham!

Academy or not?
I estimate, best case scenario, we'll have 35 turns until we complete Mass Media. Washington's commerce will be about 45 maximizing commerce and it will max out at about 60 or so. Let's assume an average of 52 commerce. Under bureaucracy which likely come soon we will produce an average of 78 commerce.

If we run close to 100% science once Oxford is finished then an academy is worth .5*78 ~39 research per turn for 32 turns (since it will take 3 turns to finish Oxford). That means an academy is worth about 39*32 = 1248 (before modifier science) So we have at least a 1.2 modifier due pre-requisites so 1248 * 1.2 = 1497 post modifier science. Alright, less than a great scientist bulb ~1700 post modifier science), but more flexible research. Plus if we have to tech past Mass Media the academy keeps on giving. Plus we don't have to worry about any of the academy's science being wasted in a bulb overflow situation. I think an academy is better than bulbing printing press. Printing Press will add about 4 commerce to washington in the short term. Not very significant.

So I think Great Scientist coming should be used for an academy. The rest of the great scientists will be bulbs
estimating using my spreadsheet GS will come...
T145 academy
T150 printing press
T155 astronomy (will self tech rest of astronomy and scientific method)
T157 save for Physics
T164 double bulb Physics?
free GS with physics
T166 double bulb electricity

I think it will help our tech pace if we put the passive science into a little bit of printing press, then astronomy. We know we will have to have some non-bulb science for those techs. And getting Printing Press a turn earlier is nice and Astronomy is nice too. I think we might want Isengard to start cranking out galleons (if we get astronomy before engineering). I want a galleon chain set up in case we need to get to the that island on the far side of the world.

[shulec]

Originally Posted by bcool

I like the out of box thinking on the teleportation trick, but closing borders with the northern witches would be a big set back on our goals for diplomatic victory. I think we can wait for airships...

That tiny barbarian city on the edge of the map near Monty does look promising for a wizard location, good find and suggestion on the workboat mabraham!

The big stack would be needed when we close the borders the second time if it works and reveals the wizard on the first try.

The closed borders probably only cost us diplo point. About 20 turns to wait until they are willing to open borders and 20 more for the +1 bonus, hopefully we would win before the second bonus point came. Also, Liz spammed some taoism this turnset. With an infiltration, we can easily convert her and she might stay in Taoism.

That city formed in the last 3 to 5 turns. Regardless, it would be a great place to hide a barb unit before the city was founded.

[bcool]

Originally Posted by shulec

The big stack would be needed when we close the borders the second time if it works and reveals the wizard on the first try.

The closed borders probably only cost us diplo point. About 20 turns to wait until they are willing to open borders and 20 more for the +1 bonus, hopefully we would win before the second bonus point came. Also, Liz spammed some taoism this turnset. With an infiltration, we can easily convert her and she might stay in Taoism.

That city formed in the last 3 to 5 turns. Regardless, it would be a great place to hide a barb unit before the city was founded.

testing reveals my ignorance...
we lose the positive open borders diplo but there is no negative diplo hit for closing borders.

Even if the AI are pleased with us however, you can not immediately open borders with them again. So there is some time before you can open borders again.

[shulec]

General Comments on my turnset:

Plans went pretty much as expected:

• I was able to trade for maps to see most of Shaka's lands. I didn't get to the western part of his lands.
• I begged once at pleased and was surprised Gandhi denied me.
• I sold our map for about 50 gold and spent about 15 gold for maps.
• Elizabeth spread Taoism to several cities.
• We didn't get spontaneous spread of Taoism to any of our cities without religion.
• Isengard came out maybe 1-2 pop behind due to my build order mistake.
• Stole compass, drama, optics.
• We are seven turns from stealing Civil Service from Trojan horse.
• We are not far from an infiltration of the north witches, thus enabling us to steal Engineering.
• We should consider moving the spy out of Boston if we are more than 6 turns from infiltration of the north witches.
• I think we need to get the chariot in the hub to Mansa Musa's territory and closed border teleport him to the island in the ocean between Shaka and Mansa
• I changed the plans for growth in Culture bridge to get a caravel asap.
• We did not go to -2 for the spy found in Trojan Horse, but it probably reset the decay for that diplo hit.
• GP are on target as expected.
• I was able to redo trades with Mansa, Genghis, Gandhi and Elizabeth. Shaka has more gpt out there, but I don't want to chance a diplo hit with the north witches for a new trade. I am unsure how the diplo hits work. The current trade now is unbalanced, but I fear a diplo hit with renegotiation. Note that the turn I renegotiated with Mansa, he had just become the north witches worst enemy. We got no diplo hit for that.

I will try to update the test game tonight. I may need someone to finish the last 2 turns to make sure everything is correct. The accuracy of the most recent test game made it an awesome tool! Thanks bcool!

@ Tachy- yeah, I have been sleep deprived the last four nights, but I can do well on a few hours sleep. I have my first day off in about 20 days tomorrow and my son has no school! Guess who's sleeping in after finishing the test game! If I fall asleep before finishing the test game, I will do it tomorrow morning in about 24 hours.

[mabraham]

Originally Posted by bcool

Academy or not?
I estimate, best case scenario, we'll have 35 turns until we complete Mass Media. Washington's commerce will be about 45 maximizing commerce and it will max out at about 60 or so. Let's assume an average of 52 commerce. Under bureaucracy which likely come soon we will produce an average of 78 commerce.

If we run close to 100% science once Oxford is finished then an academy is worth .5*78 ~39 research per turn for 32 turns (since it will take 3 turns to finish Oxford). That means an academy is worth about 39*32 = 1248 (before modifier science) So we have at least a 1.2 modifier due pre-requisites so 1248 * 1.2 = 1497 post modifier science. Alright, less than a great scientist bulb ~1700 post modifier science), but more flexible research. Plus if we have to tech past Mass Media the academy keeps on giving. Plus we don't have to worry about any of the academy's science being wasted in a bulb overflow situation. I think an academy is better than bulbing printing press. Printing Press will add about 4 commerce to washington in the short term. Not very significant.

A bulb is 1500 + 80*3 = 1740 right now, and probably growing at about 2*3=6 per turn as we regrow a 10-odd cities from whipping. By T166 a bulb might be around 1850. There's room for at least one bulb on SciMeth, and probably another on Electricity, so I think all of our existing Gscientists will get full value if used as bulbs.

Valuing the academy at ~1500 depends on us coming up with ~100 per turn for ~30 turns. We have 800 now. Where's 2200 going to come from? An average of 17 merchant specialists per turn would do the job, but Gems and Silver are running scientists to pop their Gscientists for a few turns yet. CB, Marble and Stone can support 4, 5 and 4 when they get to full size in 3-8 turns, but I think we were hoping Stone would produce hammers on stuff. With renegotiated trades and hoped-for map and tech trades we might be able to sustain 100% for 30 turns, but I think it will be close.

Of course, if we need to tech stuff like Nat, Gunpowder or anywhere deeper to get the wizard then the academy has much longer in which to show more profit.

Thinking further, if the academy is a straight boost of 1500 and our base beakers are 375/turn (so 450 after single pre-req multipliers, which is generous because Astro has none), we'll produce something like 450*35 + 1500 =15,750 over those 30 turns, before factoring in the Rep bonus on those merchants we plan to run. Is 16K plus 6*1800 from bulbing (as postulated by bcool = 26.5K) enough to get PPr, Lib, Astro, SciMeth, Physics, Elec and MM? (Assuming Lib gets Radio)

Side point: Because Astro has no prereqs, if two bulbs are less than 1.2 times the size of Astro, it's actually more efficient to double bulb and overflow Astro wastefully, so that our natural beakers get the prereq multiplier. Which techs in that tree have multiple relevant prereqs? PPr? SciMeth? If so we should minimize bulbs on them if possible.

[bcool]

Since astronomy doesn't get any prerequisite bonuses, then double bulbing it gives us 3120 research. So each great scientist there is worth 1560 unmodified science (there is no 1.2 prerequisite modifier on astronomy as mabraham indicated)

It is slightly more efficient to double bulb astronomy than it is to bulb something else with a prerequisite. 1800*2 / 1.2 = 3000 (plus we could build an observatory in Washington at least earlier too)

tech costs
PPr 2496 (1.2 modifier)
astro 3120 (no modifier)
Sci Method 3744(1.2 modifier)
physics 6240 (1.2 mod)
electricity 7020 (1.2 mod)
lib 2184 - 330 already done (1.2 mod)
Mass Media 5616 (1.2 mod)

so 26970 / 1.2 + 3120 = 25595 research required
so yes 26.5k is enough

CB, Marble and Stone can support 4, 5 and 4 when they get to full size in 3-8 turns, but I think we were hoping Stone would produce hammers on stuff. With renegotiated trades and hoped-for map and tech trades we might be able to sustain 100% for 30 turns, but I think it will be close.

Fur City will also contribute some merchants pretty quickly...

[mabraham]

Originally Posted by bcool

Since astronomy doesn't get any prerequisite bonuses, then double bulbing it gives us 3120 research. So each great scientist there is worth 1560 before (incorrectly on my part) assumed 1.2 modifier science.

tech costs
PPr 2496 (1.2 modifier)
astro 3120 (no modifier)
Sci Method 3744(1.2 modifier)
physics 6240 (1.2 mod)
electricity 7020 (1.2 mod)
lib 2184 - 330 already done (1.2 mod)
Mass Media 5616

so 26970 / 1.2 + 3120 = 25595 research required
so yes 26.5k is enough

The 26.5 assumed a 1.2 modifier was already applied to the natural beakers, so we should compare it with 26970. That's still acceptably close.

So we are on track to get MM in ~35 turns if we can raise the necessary gold. If we have to Lib something else (i.e. Elec), then that will hurt quite a bit, but it does make the academy more desirable.

So long as we're vigilant about getting gold, I think getting the academy is prudent.

[bcool]

I think we might want to start very slowing building galleys instead of building wealth in some of our merchant cities. We are probably going to want several galleons and we will want them soon after we get astronomy.

So the galley builds will convert to galleons long before we finish the galleys at 2 hammers/turn. I think the galleons sooner will be worth more to us than the ~150 gold those cities might build for us.

[bcool]

Originally Posted by mabraham

The 26.5 assumed a 1.2 modifier was already applied to the natural beakers, so we should compare it with 26970. That's still acceptably close.

So we are on track to get MM in ~35 turns if we can raise the necessary gold. If we have to Lib something else (i.e. Elec), then that will hurt quite a bit, but it does make the academy more desirable.

So long as we're vigilant about getting gold, I think getting the academy is prudent.

Your research estimate assumed that our base science rate wouldn't significantly increase from where it is now. I think that it will rise significantly from the merchants you mentioned, switching off the mines in the capital for commerce tiles, the cottages maturing and bureacracy! So even with my mistake I think 30-5 turns is reasonable as you suggest already...

[Tachywaxon]

Originally Posted by shulec

I am pretty sure that you do get a positive diplo "you voted for us" for a city vote. Can someone test this? I can test it after my turnset, but don't have time now.

Spoiler:

<MemoryType>MEMORY_VOTED_AGAINST_US</MemoryType>
<iMemoryRand>20</iMemoryRand>
</MemoryDecay>
<MemoryDecay>
<MemoryType>MEMORY_VOTED_FOR_US</MemoryType>
<iMemoryRand>20</iMemoryRand>

Like other types of decay, it is probabilistic and you were right to suspect a fast decay. It doesn't have to do with AP resolutions intervals or whatever a related property.
Looking into the spoiler and I already see the probability of memory loss is higher than forgetting spy diplo hit.
In a nutshell (IIRC), you have a dice of 20 faces where 0 value (0 to 19) represents his memory loss.

So, 19/20= 95% chance to keep the AP "You voted for us" for each turn.
Already after 13 turns, the probability of keeping that diplo bonus lowered to ~51%.
After 20 turns, it reached 36%.
From what I tested, I perceived an interval of 5 to 17 turns. Still a small pool though.

I am sure I have overdone the explanations...I think you are good with stats.

================================================== ===========

Just to take an ol' quote about diplo hits and boni. I furnished a test game, here's more or less theoritical values.

Spoiler:

Originally Posted by littlej

I downloaded this http://forums.civfanatics.com/downlo...o=file&id=6570 or this version http://forums.civfanatics.com/downlo...o=file&id=6571

So as long as I have that, plus this chart
Friendly >= +10
Pleased >= +3
Cautious
Annoyed <= -3
Furious <= -10

And roughly here are the general modifiers based on your actions together (this does not include rank or any other hidden modifiers)

MEMORY_DECLARED_WAR -3Pts
MEMORY_DECLARED_WAR_ON_FRIEND -1Pts
MEMORY_HIRED_WAR_ALLY -2Pts
MEMORY_NUKED_US -2Pts
MEMORY_NUKED_FRIEND -1Pts
MEMORY_RAZED_CITY -2.5Pts
MEMORY_RAZED_HOLY_CITY -2Pts
MEMORY_SPY_CAUGHT -1Pts
MEMORY_GIVE_HELP 1Pts
MEMORY_REFUSED_HELP -1Pts
MEMORY_ACCEPT_DEMAND 1Pts
MEMORY_REJECTED_DEMAND -1Pts
MEMORY_ACCEPTED_RELIGION 1Pts
MEMORY_DENIED_RELIGION -1Pts
MEMORY_ACCEPTED_CIVIC 1Pts
MEMORY_DENIED_CIVIC -1Pts
MEMORY_DENIED_JOIN_WAR -1Pts
MEMORY_ACCEPTED_STOP_TRADING 0.5Pts
MEMORY_DENIED_STOP_TRADING -1Pts
MEMORY_STOPPED_TRADING -1Pts
MEMORY_HIRED_TRADE_EMBARGO -1Pts
MEMORY_MADE_DEMAND -1Pts
MEMORY_TRADED_TECH_TO_US 0.1Pts
MEMORY_VOTED_AGAINST_US -2Pts
MEMORY_VOTED_FOR_US 2Pts
MEMORY_EVENT_GOOD_TO_US 1Pts
MEMORY_EVENT_BAD_TO_US -1Pts
MEMORY_LIBERATED_CITIES 1.5Pts


I should have a better grasp of how to handle my diplomatic relations. We'll find out tonight.


PS - Thanks MadScientist, you really did help me clear some of that up.

Stop trading is just bad.

Originally Posted by shulec

@ Tachy- yeah, I have been sleep deprived the last four nights, but I can do well on a few hours sleep. I have my first day off in about 20 days and my son has no school! Guess who's sleeping in after finishing the test game! If I fall asleep before finishing the test game, I will do it tomorrow morning in about 24 hours.

Looks like I am not the only one where the personal turnset had an impact on RL.
I'm sweating just to think my next turn-set arrives...soon...

[shulec]

Originally Posted by shulec

Isengard came out maybe 1-2 pop behind due to my build order mistake.

I looked at my final test game save and was happy to see that my mistake didn't hurt much as both games have us at pop 9 on T144. In the real game, we are three turns from growth whereas we should be at one turn to growth. We are actually ahead on hammers. I will now stop dwelling on this.

[shulec]

I am attaching the test game that is saved at the end of T 143. It keeps crashing on me. While making the game, I was getting crashes along the way. If anyone can fix it, I would be grateful. It will probably need a tweak or two to match the real game. I will attach the T 134 beginning of the save too.

I got it to work. There were several AI at war. I made them make peace and it worked.

For the previous turns, I was able to alter the map to get it to go to the next turn.

[mabraham]

Originally Posted by mabraham

See screenshots for a possible UN site that we will need to organize settling this turn set. We want a bunch of forests to chop to go with our GEng, assuming that there's enough forests there to finish the UN... should do some maths some time.

Actually, our GEng will give 500 + 20 * x where x is the population of the city in which we rush the UN (see http://forums.civfanatics.com/showthread.php?t=148467), which is 1000 hammers (see http://www.civfanatics.com/civ4/info/wonders). So a city like GPfarm of about size 16 will rush 80% of the UN, and finish it off in a dozen turns or so. Washington will probably be similar. Maxing growth in such a city during teching MM is probably best policy.

Chopping 10 forests in a tiny city (like I suggested above) only gives about 300 hammers and costs a pile of worker turns, so is probably not the right approach.

[Sun Tzu Wu]

Originally Posted by bcool

shulec!

Sun Tzu Wu... UP NOW
mabraham... on deck
Tachywaxon... in the hole
Walter_Wolf... waiting
Kaitzilla... waiting
frogdude... waiting
bcool... waiting
shulec... waiting

Great turn set shulec!

Please consider this a "Got It" post. I will review our current game state in detail with short term and long term goals in mind. I will provide comments on Grand Strategy, "Normal" Strategy and Tactical Considerations where the micromanagement will be more focused on critical and major goals (actions to achieve minor goals may or may not appear in the PPP).

I will provide a draft PPP ASAP, allow at least 24 hours for comment, finalize the PPP, allow 12-24 hours for further review depending on the extent of the changes, followed by execution of turn set. This will help us keep moving at 3-4 day per 10 turn set.

Sun Tzu Wu

[shulec]

Here is the test game:

Phants city has 6 extra hammers that can be adjusted for when it grows.

We have 25 extra beakers.

If anyone can fix these, please do.

Edit: test game moved.

[bcool]

Shulec, do you have the T143 test game save? I can fix the hammers if you have that save for Phants city and the extra beakers (which are far less important).

Thanks

I'm updating the gpp spreadsheet in the mean time.

Updated the gpp spreadsheet to reflect reality. Great person times predictions pushed back 1 turn but Gems and Silver still can produce great people ahead of GPFarm with some tweaks and a dip into Silver's granary food (possible starvation but I don't think it will be necessary)

Gems' 2nd great scientist at 1400 gp is optional and might not be possible unless we slow down GPFarm, although I still haven't accounted for when the Great Library goes obsolete and we lose 2 scientist specialists in GPFarm.

For the coming turnset the great people you run will become critical please consult the spreadsheet when deciding how many specialists to run and double check my food and pop calculations that led to my specialist numbers on the spreadsheet. It is important that both Gems and Silver max food with the population that is not being used as a specialist.

Please also note that Gems is +21 gp in the test game compared to the real game
Silver is +12 in test game
The others are accurate or are not important.

[Sun Tzu Wu]

I should be able to publish a preliminary PPP late Friday night or more likely early morning Saturday (server time zone).

Sun Tzu Wu

[shulec]

Originally Posted by bcool

Shulec, do you have the T143 test game save? I can fix the hammers if you have that save for Phants city and the extra beakers (which are far less important).

Please also note that Gems is +21 gp in the test game compared to the real game
Silver is +12 in test game
The others are accurate or are not important.

Here are two saves. One is the unaltered map save. The other is a map with 18 map alterations, all labeled, that need to be reset on T144. The latter is the one I used for the final test game save.

EDIT:
Note that Phants city and Culture bridge are 1 food short in the test game. Phants city had 7 extra hammers in the test game.

I don't know if you saw this, but some of the tiles were not producing what they should have. The grass farm north of silver was only producing 2F. The Plains form north of Gems was only producing 1F and 1H.