SGOTM 12 - One Short Straw

[babybluepants]

Your two problems seem mutually exclusive, since inserting a chariot into the build order delays the need for PH. At the end of this sequence we're going at 15hpt in Moscow, which translates into a 2t Oracle delay with a chariot added.

I just ran one test, where I built a chariot after the worker and teched Alpha instead of Math. Currency in 1360 BC, and peace with DG for PH (with chariot 1 tile from Paris). Not sure we even want peace, but it's an option worth discussing. Without a cheap Gandhi DoP via unit-threat, it will take quite a lot to get Alpha in trade.

I mentioned before: we have until at least T43 to decide whether we want an early chariot. Let's play 2-3t and confirm the land connection.

 

[LowtherCastle]

I mentioned before: we have until at least T43 to decide whether we want an early chariot. Let's play 2-3t and confirm the land connection.

My only problem with that is that these next three turns are the only ones that make sense to store some cash, if we want. That's why I first noticed the PHood discrepancy. I was trying to figure out how much cash we could store and not delay Phood. It seems we can't count on a T52 PHood. That may mean we can store some cash for a T53. But if we're going for a T54, maybe even more cash. The more cash we can store, the more beakers we get for free.

 

[bcool]

Quick update on my availability

I'm going to work, so I won't be able to play or test for the next 10 or so hours.

With so many AI having so few land tiles in there initial cultural borders, Paris could easily block the land beyond. So tough call on saving cash and the early chariot.

We could move one space safely with the warrior I think, but that probably wouldn't reveal too much.

 

[babybluepants]

If the land connection is 1SE of Paris, as you guys deduced, it should be revealed next turn, I think. Not sure, though.

My only problem with that is that these next three turns are the only ones that make sense to store some cash, if we want. That's why I first noticed the PHood discrepancy. I was trying to figure out how much cash we could store and not delay Phood. It seems we can't count on a T52 PHood. That may mean we can store some cash for a T53. But if we're going for a T54, maybe even more cash. The more cash we can store, the more beakers we get for free.

This is tricky. I've been storing 20g cash for 1t after PH, having run 90% for a couple of turns. This gets me to Math @ 100% + 2t @ 90%. The reason is that I'm initially rushing to Oracle, and pre-gold-mine we don't get much cash per turn. If we really can't get PH on T52, the cash-storage issue is irrelevant - we can do it whenever. We'll need to invest hammers in something else for 1t anyway, so it's a matter of chariot vs. 1t quicker Oracle, I guess.

Bigger issue is whether we even need the chariot this early. We'd have to be ridiculously lucky to worker steal, so the only purpose would be potential copper mine denial. Defensively, I doubt we need it.

 

[babybluepants]

Quick update on my availability

I'm going to work, so I won't be able to play or test for the next 10 or so hours.

imho, we're really not in a rush with this TS.

 

[Mitchum]

I'm trying to figure out how many AI know Sailing and I'm getting some funky results:

Base Research  GNP   Final Research
  10           24          16
  11           24          16
  12           25          17
  13           26          18

According to the spreadsheet:

Based on 16 beakers @ 10 base, more than 6 AI know the tech???
Based on 16 beakers @ 11 base, 4 or 5 AI know the tech.
Based on 17 beakers @ 12 base, 3 or 4 AI know the tech.
Based on 18 beakers @ 13 base, 3 AI know the tech.

Am I doing something wrong?

Also, this seems to indicate that we can run at 10 base beakers and get the same final beaker output as when we run at 11 base beakers. If we can work the slider correctly to get 10 base beakers, this would actually save us 1 g.p.t. over running it at 11 base beakers, right?

 

[babybluepants]

Don't know what's up with the Sailing thing and can't check now. It would be good to know, I guess. EDIT: If we want GLH, we just go for it. There's not all that much we can do to speed it up or slow it down, except choosing to build an explorer WB at some point. I'd say it's a longish shot, but still possible.

10-11: Clever, but I don't think you can get 10b+1g from 11c due to rounddown, if that's what you're suggesting.

 

[LowtherCastle]

I'm trying to figure out how many AI know Sailing and I'm getting some funky results:

Base Research  GNP   Final Research
  10           24          16
  11           24          16
  12           25          17
  13           26          18

According to the spreadsheet:

Based on 16 beakers @ 10 base, more than 6 AI know the tech???
Based on 16 beakers @ 11 base, 4 or 5 AI know the tech.
Based on 17 beakers @ 12 base, 3 or 4 AI know the tech.
Based on 18 beakers @ 13 base, 3 AI know the tech.

Am I doing something wrong?

Also, this seems to indicate that we can run at 10 base beakers and get the same final beaker output as when we run at 11 base beakers. If we can work the slider correctly to get 10 base beakers, this would actually save us 1 g.p.t. over running it at 11 base beakers, right?

Not sure how you're doing that. What I get aligns to 3 AIs having Sailing:

Base Research  GNP   Final Research
   9           21          13
  10           22          14
  11           23          15
  12           25          17
  13           26          18

What surprises me, though, is that the 12base-beakers ==> 25/17 is different than my table, which gives 16 instead of 17. That's the first case where the table has not matched up with the game. Weird. Perhaps my formula is not quite as precise as the code in the game, though I did use 100s instead of decimals. Maybe I shouldn't have.

 

[Mitchum]

What surprises me, though, is that the 12base-beakers ==> 25/17 is different than my table, which gives 16 instead of 17. That's the first case where the table has not matched up with the game. Weird. Perhaps my formula is not quite as precise as the code in the game, though I did use 100s instead of decimals. Maybe I shouldn't have.

I must be doing something wrong then. Regarding you using 100s, I noticed that but didn't think much of it at the time. It could be messing with the ROUNDDOWN function in some instances.

 

[LowtherCastle]

Mitchum, did you by any chance accidentally look at the max AI GNP for the first two stats?

 

[Mitchum]

@LC. I know what I did. I was adjusting the slider to get the various base commerce numbers but I was not subtracting our g.p.t. from our GNP value...

An easier way to do it would be to simply adjust the tiles worked so that the gold per turn is 0 and the slider says at 100%. This may not be possible later in the game when our base commerce is much higher, so we'll have to adjust the slider and remember to subtract the g.p.t. (and culture and EPP) from the GNP number.

 

[LowtherCastle]

Here's my take on scouting Paris.

We have more or less safe paths. And shorter or longer. Personally I prefer Path #1, even though I hate that it's the longest. It exposes that one fogged tile that might have who-knows-what-?, it exposes the inaccesible Paris-NE grass tile where Erkon the Schmuck surely put copper, and it reveals the tile we are assuming connects Paris to us. And it's safe.

From there we could risk going SW to the hill, thus exposing Paris for future production monitoring, etc, or, assuming we've already learned enough, we could just forget about the rest of the Paris tile for now and go south.

Anyway, this shot gives us soemthing to discuss, I think.

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

If I'm not mistaken, here are the only two scenarios in which Paris blocks us from tiles beyond:

Spoiler :

Paris has 5 land tiles in the inner cross, 6 land tiles in the outer cross. W is for water tiles.

V1    -------------------
      !     !     !     !
  W   !  W  !  W  !  6  !  W
      !     !     !     !
-------------------------------
!     !     !     !     !     !
!  7  !  W  !  W  !  1  !  W  !  L
!     !     !     !     !     !     !
------------------------------- - - -
!     !     !     !     !     !
!  8  !  2  !  3  !  W  !  W  !  L    Hil
!     !     !     !     !     !     !     ! 
------------------------------- - - - - - - 
!     !     !     !     !     !
!  9  !  4  !  [COLOR="Blue"]W[/COLOR]  !  5  ! Hil ! Hil    L    War
!     !     !     !     !     !     !     !     !
------------------------------- - - - - - - - - -
      !     !     !     !     
      ! 10  !  [COLOR="blue"]W[/COLOR]  ! 11  ! Hil   Hil   Hil     
      !     !     !     !     !     !     !    
- - - ------------------- - - - - - - - - -

   W     W     W     W   
      !     !     !     !
- - - - - - - - - - - - -



V2    -------------------
      !     !     !     !
  W   !  W  !  W  !  6  !  W
      !     !     !     !
-------------------------------
!     !     !     !     !     !
!  7  !  W  !  W  !  1  !  W  !  L
!     !     !     !     !     !     !
------------------------------- - - -
!     !     !     !     !     !
!  [COLOR="blue"]W[/COLOR]  !  [COLOR="blue"]W[/COLOR]  !  2  !  W  !  W  !  L    Hil
!     !     !     !     !     !     !     ! 
------------------------------- - - - - - - 
!     !     !     !     !     !
!  8  !  3  !  4  !  5  ! Hil ! Hil    L    War
!     !     !     !     !     !     !     !     !
------------------------------- - - - - - - - - -
      !     !     !     !     
   W  !  9  ! 10  ! 11  ! Hil   Hil   Hil     
      !     !     !     !     !     !     !    
- - - ------------------- - - - - - - - - -

   W     W     W     W   
      !     !     !     !
- - - - - - - - - - - - -

 

[babybluepants]

I was thinking Path 1, followed by 1SW, and then exploring further south. I don't think that 1SW is dangerous at all. At this stage he shouldn't have more than the 1-2 garrison archers. We might screw with his worker a bit, too.

 

[bcool]

So warrior moves NW, then W, then SW?

And we don't have to worry about saving gold? (if we save gold for 1 or 2 turns we later get the research bonus from the library)

So when I get home I should test to see how much if any gold we should save to get Phood by turn 52? or 54? (with chariot build)

 

[LowtherCastle]

It's complicated. You were also getting overflow from the worker into the Oracle on T52, which might push the Oracle another turn later. I don't think we can count on getting T52, so I was thinking see how much gold we can save in the next turn or two and still get Phood on T53. Then I think we can just go with that, because what we actually decide to do after the settler may depend on what the warrior sees in the next couple of turns.

Instead of a chariot, maybe we'll just build a scout, or put hammers into the barracks. I figured out that CI-promoted chariots require about 1 less per archer, so that saves hammers and maintenance.

 

[babybluepants]

We settle the second city at 2000 BC, and are aiming for Oracle at 1440 BC (which is probably impossible if we can't get PH on T52 anyway). That's 28g of savings needed for -2gpt city maintenance. We can about finish Math by then with minimal wasting of beakers, though we may be a few beakers short. If we're already talking delay to 1400 BC, then it's irrelevant as we can finish Math easily with that extra turn.

 

[bcool]

with LC's fixed tech test save I was able to get the oracle at the in between turns 1440 BC and 1400 BC (would have been recorded as finishing at 1440 BC if De Gaulle didn't beat me to it).

(2 AI had meditation, but I'm not sure if De Gaulle picked up priesthood before I did)

I went with alphabet instead of mathematics (since bbp's comment about the difficulty with trading with Gandhi without the opportunity for cheap peace with him)

I saved 10 gold, and put 1 turn into a barracks after the worker was finished.

With this tentative plan for the future, should I go ahead and play 3 turns to reveal De Gaulle's lands and save 10 gold?
Moving the warrior NW, W, SW is okay too? (obviously not moving next to a revealed archer)

I'll wait 24 hours unless you guys don't think I need to before I play.

 

[Mitchum]

I hate to self-research Alphabet when we should be able to get it in a trade from Gandhi for Math + some other cheap tech.

What is the current thinking here? Currency for peace unless Gandhi researches Writing and then we try for Math for peace?

Since Gandhi doesn't have a religion yet, will he be on a CoL beeline or could be be researching Writing now?

 

[babybluepants]

Before LC's discovery of the warrior-threat trick and moving Gandhi close by, I was having a lot of trouble opening up an Alphabet trade. We need:
(1) Tech gift for DoP
(2) Tech gift to get him to Cautious for signing OB and temporarily opening up trade
(3) Tech to trade for Alphabet in the >=Aesthetics range
By the time we Oracle Currency, Math won't be enough for (1), and might not even be enough for (2). By the time we research another 500b tech, Math + cheap techs won't be enough for (2).

This obviously has a cumulative effect. The sooner we have Alpha for DoP-ing with others, the cheaper it is to gift them into OB (mainly Roos, but maybe also others). We need this before we choose a conflicting religious side for further diplo, which we need asap so we can accumulate the plusmod quicker...

That's in my experience anyway. Maybe I was doing something wrong...

 

[babybluepants]

With this tentative plan for the future, should I go ahead and play 3 turns to reveal De Gaulle's lands and save 10 gold?

I don't really see why we have to save gold right away, esp. if we're not getting to PH on T52. Ok, it doesn't really make a difference...

I'm cool with playing 3t. That's when the workers complete the gold mine, right? Don't wanna commit to moving them from there yet.