SGOTM 12 - One Short Straw

[Mitchum]

Good job, LC. I'm on a business trip this week and next, so my days and nights are pretty full. I'll try to download the save later tonight after dinner and comment on the settling locations.

 

[babybluepants]

Edit: thought about it, and I like LC's proposed settling pattern.

 

[Mitchum]

I vote for settling Silver 1S next and CS beeline. At this point, it makes no sense to stop the CS beeline now that we've already skipped Sailing and Masonry.

I also agree with bbp that we should try to stick with binary research to avoid losing beakers/gold to rounding down. At 90%, we lose one beaker and 1 gold per turn due to rounding. At 100% science, we only lose one beaker (at our current commerce level). It's not a major deal, but it could be the difference between learning CoL one turn earlier...

How long to we give mdy to "get" it before moving on to the next player?

 

[bcool]

De Gaulle +16 increase in score on T56 has been explained? Was it possibly him researching a tech? If so, do we know which one?

Any time De Gaulle or Churchill learns a tech it is a reason to stop and do our best to figure out what it is. The risk of getting beaten to the Oracle is very real.
I assume the plan is to get 1 turn away from building the Oracle and then switching

When do we anticipate De Gaulle's borders to pop and make the 2SW1S site unavailable?

Settling near Silver makes sense for the CS beeline, and barring other developments seems like the best way to go.

 

[babybluepants]

De Gaulle +16 increase in score on T56 has been explained?

I think he built a little something.

I'd prefer settling Fish-2SW1S ahead of silver. Don't think silver can contribute to COL research, and it's strategically stronger to settle towards DG, imho.

 

[LowtherCastle]

When do we anticipate De Gaulle's borders to pop and make the 2SW1S site unavailable?

T73 (1080BC) is my current calculation. We can beat him to it if we settle 2SW1S next. It's hard to know if there's a better galley-building site near Paris before we defog the north.

-------------------

T57 test save. INcluded the southern area to test roading, etc. Didn't bother with the northwestern area.

 

[LowtherCastle]

This is mysty's most recent post:

I'm still around but as bbp pointed out I have some RL issues, and as he pointed out as well, I'll be back

Moving from school to university isn't that easy after a 1-year break caused by the Swiss Army and I'm leaving my partent's house which means a lot more work as well. I'm always skimming the posts but I can't possibly read them thoroughly and post useful comments right now as I just don't have time (or, have other priorities). It's not what the posts are about but about the sheer amount of information to process when reading those many posts. I'll try to get back ASAP while wishing you all the luck in the meantime!

@mdy, I think in the interests of moving the game along, it would be better for us to assume that mysty, shyuhe, and ZPV will grab a turnset when they reappear. That would put you next on the roster. Got time for a turnset?

 

[babybluepants]

I'm getting Turn 70 for his third border pop, but I may be wrong (quickly calculated in my head). Don't know if our warrior will manage to reveal much in the north that would influence our decision in time?

I think the main question is build order. I prefer a fourth worker before the next settler. What do you guys think? There's also the question of whether we bring Oracle to within 1t of completion now, or just grow to 5 and then immediately build workers/settlers. Assuming we can tell when an AI gets PH, how much of a window would we have to complete Oracle for Currency?

 

[babybluepants]

I think I'll give mdy 'till tmr morning (my time) to reply. We need some more testing here anyway. If not, it's either Mitchum or me, depending on his availability (he stated that he's away for a week or so?)

 

[mdy]

Apologies for the delay in replying, real life has been keeping me extremely busy for the last 2 days. I am free now so I can take a turn set any time in the next few days.

When I ran a few tests (before LC played his last 3 turns) my best dates for the Oracle were 1000B.C if we grew to size 5 before building settlers/workers, and 975 B.C. if we delayed growing. Does anyone have better dates than this? If not I would favour growing to size 5 first in order to slightly increase the odds of getting CS.

 

[babybluepants]

We definitely wanna grow to 5, so we can work the 5 good tiles. For comparison, my date was 975 BC, with worker-settler-worker(partial) in-between (I get the second worker done 1t after Oracle). You were running a scientist (or two?) in Moscow. That would presumably speed up COL, but slow down production?

 

[mdy]

Yes, I was running a scientist to get the 1000B.C date.

 

[Mitchum]

I just ran a quick test. I finished the Oracle IBT T76 (975 BC) and T77. My build order was Oracle (grow to 5) -> worker4 -> worker5 -> settler3 -> warrior (to delay Oracle by another turn) -> Oracle. The workers had plenty to do to keep up with tile improvement to support the growth of Fish, to clear a path for Silver and to pre-clear/road the first deer resource.

I settled silver 1S on T75 (could have been a turn earlier, but the deer wasn't quite cleared yet). I think I read that others were able to complete the CS sling at 1000 BC. I'm not sure how since research was the gating factor. The only way I can see to speed it up is to work the gold in the capital right away to get the library bonus on the coins from the second gold mine. However, I was still about 20ish beakers short of finishing CoL on T75...

EDIT: Cross post. I missed the last three posts...

BTW, I won't be able to play for another week or so.

 

[Mitchum]

Yes, I was running a scientist to get the 1000B.C date.

OK. Doing this accomplishes two things. It speeds up research and it slows down the growth of Fish, which should allow 4 workers to keep up with the needed tasks.

Do we need to agree now where to settle our third city before doing too much more testing?

 

[LowtherCastle]

One guess at AI research

      GNP  EPP  GOld c/ep bpt 1.24 1.28	1.44 1.48
deGa  35   20%  10%   21  10    8    -    6    -
Chur  31    0%  10%   11  19   15    -   13    -
Gand  27   10%  10%    8  16   12   12	 11   10 
Mao   25   10%  10%    6  16   12   12	 11   10 
Stal  23   20%  10%    6  12    9    -    8    - 
Roos  20    0%  10%    6  13   10    -    9    -

Before I had Churchill at 14basebpt, dG at 7bbpt, G at 11bbpt. My conclusions are that CHurchill is not reearching agri, dG not masonry and these are the likely finish dates of their next tech:

     Church  deGaulle  Gandhi
T58          
T59  PH                medit
T60  PH/pott           medit
T61  pott    PH        AH/sail/poly
T62  AH                AH/sail/poly/wtg
T63  Mono    pot(msn)  wtg
T64  Mono    pot(msn)
T65
T66          AH/sail
T67          AH/sail
T68
T69

This could be completely wrong of course, because the GNPs and gold% are just guesses. Time will tell.

 

[LowtherCastle]

AI Hammers

I think dG is doing 10(or now 12)basehpt. Can't get a read on Churchill, but my guess is 12 or 13hpt. That means dG would need as little as 4t (with big overflow, an accidental chop, and poprush) for the Oracle, earliest T65. Churchill as little as 7t, earliest T66. So we should probably have the Oracle down to 3t from completion by T61.

dG could have upped his research though, so we should monitor his espionage build prod stats next turn (couldn't do it this turn because he evidently finished an archer). Hopefully, he won't complete a build next turn. If he's still going at 10hpt, then his research is probably still ~10bpt.

 

[mdy]

OK. Doing this accomplishes two things. It speeds up research and it slows down the growth of Fish, which should allow 4 workers to keep up with the needed tasks.

Do we need to agree now where to settle our third city before doing too much more testing?

It looks like the current 3 workers are more than capable of keeping up in the near future. This suggests we should build another settler next.

The best 3 sites available at the moment look like the ivory/rice on the coast, the silver/2 deer site and the rice/pig/ivory/fish site to the S. I favour building the ivory/rice next as we can get it off to a quick start by improving the ivory before the city is founded, and it is closer to our core and allows galley access to De Gaulle.

I think I read that others were able to complete the CS sling at 1000 BC.

I get CS in 1000 B.C by:

T57
worker W of capital moves to 1E of it to clear fallout.
worker S of Novogrod clears fallout 2W of Novogrod
Other worker pastures cows.
Adopt binary research.

T59
Novogrod builds workboat, starts library

T60
Novogrod works fish.
St Pete grows to size 5, works 2 gold, 2 cows, and scientist.

T62
Cows pastured, worker scrubs fallout on other ivory

T64
Start settler in St Pete
road 1E of St Pete
Camp northern ivory

T65
Novogrod works lake

T66
Worker near capital moves to scrub fallout on the rice

T67
End binary research

T68
worker on northern ivory moves to scrub fallout 1NW, 1 W of the current location.

T69
camp other ivory

T70
Novogrod size 3, works cows

T71
Build settler in capital, start worker

T73
road ivory and rice.

T75
1000B.C. discover code of laws, Oracle CS, Novogrod now at size 4. Found 3rd city.

Novogrod would then finish her library on T77, giving us a GS for an academy on T94.

 

[LowtherCastle]

I'm getting Turn 70 for his third border pop, but I may be wrong (quickly calculated in my head).

Okay, if we decide to settle 2SW1S, then this is important. Let's make sure we've got it right. Can you find any errors in this?

              total  turn#     #ofturns    culture
              -----  -------   --------    -------
palace  2cpt   2cpt  T00-T02       2            4
revolt  0cpt         T03           1            0
palace  2cpt   2cpt  T04-T18      15           30
buddhi  5cpt   7cpt  T19-T55      37          259
StoneH  8cpt  15cpt
monume  1cpt  16cpt  T56-T73      18          208
                                  --          ---
                                  73          501 T73 TOTAL

This means:
1. We monitor Paris builds, in case he builds a monastery or the GW.
2. We need to settle 2SW1S by T72 at the latest.

 

[babybluepants]

16 x 18 = 288, not 208
Didn't he also revolt to Slavery?

I'm always slightly confused by AI turns. The log says that wonders complete one turn, but we get an in-game notice the next. T19 is log-time for Buddhism, and T55 for SH (not 56).

 

[Mitchum]

Also, didn't he revolt to Buddhsm?