SGOTM 14 - Kakumeika

[bcool]

hunting has no prerequisite so it is one of the few techs that do not get the 1.2 modifier for having the prerequisite. That explains why hunting is so low.
mysticism should be the same has hunting since both teams must know mysticism one founded buddhism the other founded Judaism

 

[bcool]

prereq mod	GNP 0 know	GNP 2 know	GNP 4 know	base research
1.2	26	27	28	17
1	23	24	25	17
1.2	31	32	33	21
1	27	28	29	21

prereq modifier / GNP expected if 0 know / GNP expected if 2 know / GNP expected if 4 know / base research

this is the best I could come up with. I assumed that expenses are not subtracted from GNP and I used the formula from http://www.civfanatics.com/civ4/strategy/tech_research.php, but there seems to be something wrong since the numbers are not predictive. They don't exactly match what we expect from the game.

So all we can say is more AI know masonry than sailing.

prereq mod	GNP 0 know	GNP 2 know	GNP 4 know	base research
1.2	25	26	28	18
1	22	23	24	18
1.2	30	31	32	22
1	26	27	28	22

If I add 1 to the base research and subtract expenses it looks a bit better (hunting matches), sailing is consistent with 2 AI knowing the tech, but masonry isn't consistent with all 4 knowing it. But I think it is safe to assume that is the case.

 

[bcool]

I'm in favor of keeping the espionage on the east witches because they are ahead in score and we have a good chance of getting a culture bridge, a blessing for spys and invasions.
I'm still in favor of sailing first for earlier Great Lighthouse too.

The fact that we haven't met any industrial AI and we know all of them do not yet know sailing is promising I think for the advantage that teching pottery gives. Stone city is much better off with a granary first. Plus there would be no risk of a great merchant if we build the GLH T75.

I'm fine with espionage on the eastern witches.

 

[Kaitzilla]

I'm fine with espionage on the eastern witches.

Ok, shall I play until masonry finishes researching?

 

[bcool]

sure.

 

[shulec]

I did a few tests on our test game. All AI had hunting, I didn't use this for comparison. On T46, we met our second opponent.

Both teams had masonry and sailing.
Our GNP was 33 for both techs.

I removed masonry for one team.
Our GNP for masonry was 31 and sailing was 33.


I removed sailing for one team.
Our GNP for masonry was 33 and sailing was 31.

From this one test, nothing could be inferred by the different GNP's other than both AI teams had sailing or masonry, and one of them didn't have both.

 

[Kaitzilla]

Played until T48 and stopped. Micro is going according to plan and Burke is healing up. No barbs spotted. I'm waiting on our pottery vs. sailing decision before playing to T51.

Stonehenge was founded T47 by an unknown civ.


The north witches seem be be splitting espionage points against us with someone else, so I think they might have contact with an unmet AI. They still haven't met the eastern witches.

 

[shulec]

I am starting to think that 68% land domination without vassals and conquest on this wheel map will take a long time.

If all AI are connected to the wheel, cultural and diplomacy victories are looking a whole lot better.

If we decide to consider culture, this can change our whole approach. In other words, I may not be upset if we lose the Great Wall.

 

[bcool]

Okay we need a new vote on sailing first or pottery first in the light of new tests indicating we have a choice between
1) sailing and a T73 GLH in a stone city with 2 pop and no granary
2) pottery and a T75 GLH in a stone city with 3 pop and a granary (or T74 GLH w/o granary)
Please vote within 18 hours so Kaitzilla can continue his turn tomorrow. (we can extend this deadline if people object)

I vote for pottery first since we haven't met any industrial AI, and we know that at least 1 pair of AI do not know sailing yet.
and a T75 GLH beats all but 1 of WW's 8 tests with 4 industrial AI.
I believe the risk of losing the GLH is quite low and the advantage of a early granary in stone city is significant ~40 additional food, 60 hammers for granary + ~30 or more hammers more it will produce, ~50 additional commerce from working the gold mine earlier even considering the loss of ~16 commerce from additional trades from the GLH on T73

 

[Kaitzilla]

If you guys are willing to gamble a little bit, I suppose T75 TGL isn't that bad with all those benefits. We will have 100% chance of great spy too. I'll change my vote to pottery first unless someone comes up with a good counter argument.

Sailing just seemed to work better before wet rice was discovered now that I look back at my reasons.


I'll be online again tomorrow night, but it might be later than I started at today.

 

[frogdude]

The recent test games have been compelling. I'm happy to vote pottery. I'm all out of devil advocate arguments.
If someone beats us to the GLH in those 2 intervening turns, hopefully it'll be the nearby witches and we can go punish them for it

 

[Kaitzilla]

I am starting to think that 68% land domination without vassals and conquest on this wheel map will take a long time.

If all AI are connected to the wheel, cultural and diplomacy victories are looking a whole lot better.

If we decide to consider culture, this can change our whole approach. In other words, I may not be upset if we lose the Great Wall.

It is true we can attempt culture thanks to marble, and diplo thanks to such a nice starting area. I'm not convinced it is faster than beating up the north and east witches, and then settling the terrible middle of the map. We do have horses within reach now, so it isn't impossible against emperor AI.

If someone wants to take the time, they can take my theory map from post #602 and color in the parts we'd need to have to reach 68% dom limit. This could be done by opening a future age wheel map in Civ 4 and counting tiles to get a rough idea, then copying and editing the picture.

 

[mabraham]

A balanced slightly more risky (how much risk is T75 vs. T73?) approach gives a significant boost I think. Easily +40 food +50 commerce +30 hammers in stone city alone. Mabraham's spreadsheet might be able to give more exact numbers on what a T73 GLH would cost us.

OK I re-did my T75 granary+GLH with the long-range chop to allow more food into the bin during GLH. I compared that to my T73 no-granary GLH out to T88 building some hypothetical large wonder, using my spreadsheet. I worked rice, crabs, oasis, gold, Gmine, stone quarry, coasts in that order as pop increased.

• The T73 no-granary GLH whipped its granary ASAP, then grew to size 6 with 18 food in the box. It produced 99 hammers and 131 commerce after GLH. The GLH two turns earlier returned 6 more trade routes in the 3 coastal cities, for 12 more commerce. It also gave 2*2*3 earlier and one turn earlier on the presumed GSpy. The tech path also gave a handful of turns of earlier trade routes to the gems city - lets be generous and call that 10 commerce.
  
• The T75 granary+GLH grew to size 8 with 24 food in the box. It produced 114 hammers and 212 commerce after GLH (the gold starts working T76 and stays working throughout), and is 60 hammers further ahead because it hasn't needed to build a granary after GLH.

One can argue about the order of working the tiles during growth, and whether working coasts is fair to the latter strategy, but even so the latter is ahead by something like 40 post-granary food, 75 hammers and 50 commerce. There's also some unknown and small advantage from earlier granaries elsewhere - but it is not clear whether we'd start building them early enough to matter for this decision. bc seems to be keener on the capital spamming settlers than getting its granary. The gems city can start building a workboat for a southern city if we don't have pottery at the time.

Were we to lose GLH T73-4, we'd be picking up about 120 fail gold, which is not the end of the world either - faster Maths and Construction.

 

[bcool]

I might be wrong about spamming settlers in the capital before the granary. With a granary we can consider whipping settlers and overflowing into the pyramids. And whipping settlers out might delay the first settler or 2 but probably with a granary the end result would be more hammers produced. We need to give the capital some of the same testing loving that we gave stone city.

thanks for doing that analysis mabraham. I'm glad my out of thin air numbers weren't too far off. Wouldn't the earlier GLH just give 2*2 ?

Plus with a size 8 city earlier we might increase our trade route value. What is the population that 1 trade routes become 2 trade routes?

 

[shulec]

I am undecided based on the discussion. I abstain.

 

[Walter_Wolf]

The pottery - sailing path gain looks very appealing. It might make a diference in the long run.
How unfortunate one must be to lose TGL for a 2 turn delay. Also, T75 sounds pretty early from my tests (if anyone thinks they are worth anything).

Still, SH T47 IS early.

 

[frogdude]

The pottery - sailing path gain looks very appealing. It might make a diference in the long run.
How unfortunate one must be to lose TGL for a 2 turn delay. Also, T75 sounds pretty early from my tests (if anyone thinks they are worth anything).

Still, SH T47 IS early.

I tried at some point to see if there was any correlation between the time of SH, and the time of TGL. I did this mostly by staring at your stats for awhile so not particularly scientifically.
I couldn't spot any strong trend but there was, if anything, a mild tendency for an early SH to mean a later TGL

 

[Sun Tzu Wu]

STW did suggest that, but I see no value in it. Anything that would change our minds is already a stopping criterion (or should be)

The point I meant to make is do not chop the forests until computing on paper that they will provide sufficient Hammers to complete The Great Wall in the current turn. Meaning do not commit our forests to building a great wonder until we are sure to complete it.

The other reason is our turn set guidelines state that turn sets end at the beginning of turns, before anything non-reversible is done, like moving units. We should try to comply with our own guidelines, even when we don't agree. Remember that I was in favor of ending turn sets at the end of the turn. I definitely would not object to the Workers completing their forest chops as an explicit exception of our guidelines, but we need to make it clear that we are allowing an exception to our guidelines rather than modifying them.

Sun Tzu Wu

 

[mabraham]

The point I meant to make is do not chop the forests until computing on paper that they will provide sufficient Hammers to complete The Great Wall in the current turn. Meaning do not commit our forests to building a great wonder until we are sure to complete it.

That is in the PPP. All three chops are simultaneous on the final turn of the GW build. Hence why I saw no value in not chopping

The other reason is our turn set guidelines state that turn sets end at the beginning of turns, before anything non-reversible is done, like moving units. We should try to comply with our own guidelines, even when we don't agree. Remember that I was in favor of ending turn sets at the end of the turn. I definitely would not object to the Workers completing their forest chops as an explicit exception of our guidelines, but we need to make it clear that we are allowing an exception to our guidelines rather than modifying them.

Sun Tzu Wu

Sure. I'm happy to go to start of T52 to follow our principle, since the only other thing that is going on are warrior and settler moves - and Burke will still be healing up.

 

[Sun Tzu Wu]

You get a free beaker at 0% science. You also avoid losses from rounding errors from the slider. Very occasionally useful. You also get the benefit of banking gold for a turn before committing beakers. In our case, a mega-early AI GLH would allow us not to commit as many beakers to Sailing - wildly unlikely, however.

Very well stated! I often forget about free beaker per turn at 0% Research, unless I'm doing it for more than one turn at a time.

In the worst case at 10-90 % Research, I believe its even possible to lose one beaker of Research and one unit of Wealth. It doesn't happen very often, but when these things should divide evenly, sometimes one loses one of each. Can anyone else verify this? I always use binary research, unless my Wealth reserve falls below 0 and I don't notice it.

Sun Tzu Wu