[SPOILERS] Bismarck of Mali (NobleHelium and Lewwyn)

[Lewwyn]

Yeah Jaken couldn't defog our capital since he couldn't get close enough without open borders. So no worries/gains/losses there.

[NobleHelium]

Eh, he should be able to see it on the diagonal, no? I'm pretty sure that would work. Two-tile vision over coast tiles.

[NobleHelium]

So, Lewwyn and I were talking about using that third ring chop to speed up the second settler. I did some simming for that, it definitely speeds up the settler by a turn. I actually like this now, because I was still going by the really old numbers where I was whipping the work boat at 1 turn ETA. Whipping it at a 2 turn ETA means we have less food in the box and there's actually an interim turn after whipping the galley where we need one more turn of growing back to 4. So we'd probably build a second warrior before starting the third worker. All right that was probably really confusing, so I'll turn this into a list.

With Chop:
1. We build the third worker sooner, after the settler. We research Hunting before Sailing, because if we research Sailing first we either don't start the galley until after both techs are done, or we put the third worker on hold which seems stupid.
2. We get the galley out a few turns later because of that, and then build a warrior after it (2 turns for this one due to low overflow).

Without Chop:
1. We build a warrior after the settler, then build the galley, then another warrior.
2. Then we finally build the third worker after that. After the third worker we grow to 5.

That probably doesn't show the entire picture either, but the key point is that with the chop with the build order that I'm doing, we reach size 5 in the capital on turn 59, whereas without the chop we reach it on turn 61! Now it's quite possible my no-chop sim can be optimized further, but it's definitely looking like the chop is better in basically every way except the mainland scouting department. So the question is, do we want two warriors scouting the mainland, starting a few turns earlier, rather than only one warrior scouting a few turns (probably about five) later? We can also do a warrior and a scout for the former.

[NobleHelium]

So earlier today I messaged T-hawk about the third ring chop target mechanic, and in the exchange he said this.

I just took a brief look at the code, the function is CvMap::findCity which finds the nearest city to a given plot coordinate. (It's also used for stuff like AI calculations and random events, not just chops.)

The distance metric used is equivalent to that of city border radius, so anything in 3rd ring counts as equally distant (3N1E is not farther than 3W.) Beyond that, the only tiebreaker I can see ends up being city acquisition order. The code loops through all cities, and picks a later city only if its distance is *less* than the previously found nearest city, so an earlier city will win a tiebreaker.

That didn't seem to coincide with the observed results in the sandbox even though it fit with the live game, until I noticed...

I checked the F1 screen in the sandbox and apparently the second city is above the capital. That probably happened because the capital isn't actually the original capital, I placed the original capital in another location which I then deleted via Worldbuilder. What probably happened is that when I deleted the original capital its city slot opened up and the second city got placed into that slot for some reason. That still seems weird, but that must be how they coded it.

In the earlier sandboxes with world wraps, I didn't do the funky city deletion thing, and the chop was going to the capital. So it all fits together now.

[NobleHelium]

Reminders for next turn: research Priesthood, start cottage. That's about it.

[SevenSpirits]

I checked the F1 screen in the sandbox and apparently the second city is above the capital. That probably happened because the capital isn't actually the original capital, I placed the original capital in another location which I then deleted via Worldbuilder. What probably happened is that when I deleted the original capital its city slot opened up and the second city got placed into that slot for some reason. That still seems weird, but that must be how they coded it.

Ohhhhhh. That's good info.

[NobleHelium]

So, turn 43 was an interesting one. Not because there was a lot of interesting stuff to do, but because we gained some information and now have some strategic decisions to make or not to make.

First we got this, of course.



Then, the interesting stuff happened. I checked our research rate on Priesthood and on something else that Jkaen could not possibly have researched (like Metal Casting), and our GNP went up by 1 on Priesthood. Well, obviously that means Jkaen has researched it. Now if we look at the cities and the demographics...



Ellimist gained 6 points from his growth to 5. Kuro lost 13 points from his capital whip back to 2. Now Jkaen's score has gone up by a lot, 19 points to be exact. I puzzled over this for a bit until I realized that he got his third ring on the same turn as Kuro, turn 24, due to the Buddhist holy city. And actually I thought this wasn't possible either, because the summation of the culture values in the spreadsheet only adds up to 98. But then I realized research completes before production, and cultural production is still part of production; so he got the 5 culture on the turn that he finished Meditation, giving him 103 and pushing him over the 100 threshold. Previously we got 10 points for each of the first two rings, which were 4 tiles each. The third ring is 5 tiles, and is thus 12-13 points. His growth to 5 gives him 6-7 points, so that adds up to 19.

Of course now the question is, why didn't Kuro's population loss get offset by his third ring points? The answer is that Jkaen actually got his third ring on turn 23 (he plays after us), and as previously stated, land points are awarded on full turn rollover rather than individually. Kuro got his land points on 24, even though they were both visible on the same turn to us. So next turn Kuro should gain 12-13 points from land.

Okay, but that's not all the interesting stuff this turn! The more interesting part is whether we want to change around our Oracle build due to the knowledge that Jkaen has Priesthood. There's really no other reason to research Priesthood this early since it only leads to Writing (which you can get from Pottery, a mandatory tech) or Monarchy (not useful this early unless you have wines). He researched it on turn 39, since we know all of his prior techs. So he has 16 turns, from turn 40 through the end of turn 55, to beat us to Oracle. 16 turns is definitely enough to chop out the Oracle. So, we could go with the EOT 52 build, which would cut his window to 12 turns, and he only built his second city last turn. He could have pre-chopped though of course. The EOT 52 build is quite a gambit and we might be very behind on growth compared to the others if we lose it. We won't be able to chop that third ring forest and we'll be working two mines and eating into granary food for a few turns at the end. Oracle is 100 base hammers for 620 beakers, though...

Either way, we should probably delay the granary whip by one turn. Whipping it one turn later doesn't cost us anything I believe, and we can then overflow into Oracle if we so choose.

[NobleHelium]

Need to sleep so full report tomorrow. Short version is, Lewwyn convinced me to send a note asking Jkioan whether they're building Oracle. I thought about doing that but figured it might alert them to speed it up. Well we'll see.

Reminders for next turn: whip granary, decide which Oracle build we want to go with and chop next forest accordingly. We probably want to discuss or at least think thoroughly before playing the next turn. Hopefully Jkaen will have given us a response by tomorrow night.

[NobleHelium]

So yeah, no response from Jkaen. What did I say, Lewwyn?

[NobleHelium]

Let's hold the turn for a day. I'm tired as hell anyway.