KGB Update and Discussion

[Aginor]

Most recent edit: 3680 BC

*Important note: Many of the production conclusions depend on Banana having BUILT their Warrior rather than having popped it. It is possible that some, if not many, of these statements will need to be revised in 3600 BC if Banana does NOT grow and the Mercs DO.*

Rival food, based on in-game allocation of worked city tiles, calculations from rival averages, and logical deduction:
3960: 5,4,4,0,0,0 - Vox is 5, Banana and Mercs 4.
3920: 5,5,5,5,4,0 - Banana is 4, Horde not founded, all others 5.
3880-3760: 5,5,5,5,5,4. Banana is 4, all others 5.
3720: 6,5,5,5,5,4. Vox is 6. Banana is 4. All others 5.
3680: 5,5,5,5,5,4. Demonstrated that Banana is the 4. All other rivals at 5.

Rival hammers:
3960: 2,2,1,0,0,0 - Vox is at 1. Banana/Mercs 2.
3920: 2,2,2,1,1,0 - Vox is at 1. Banana is at 2. Others unknown. Horde not founded.
3880-3760: 2,2,2,2,1,1. Vox is at 1. Horde is 2. Banana is 2. Others unknown.
3720: 2,2,2,2,2,1. Vox is now 2. Horde is 2. Banana is 2. Others unknown.
3680: 3,2,2,2,2,1. Vox is now 3. Horde is 2. Banana is 2. Others still unknown.

Rival Commerce:
3960: 10,10,9,0,0,0 - Vox is at 10. Banana/Mercs not clear.
3920: 12,11,10,10,10,0 - Vox is at 10. Banana is at 10. Others not clear. Horde not founded.
3880-3760: 12,11,10,10,10,10. Vox, Horde, Banana at 10. Others unclear.
3720: 12,11,11,10,10,10. Vox is at 11. Horde and Banana 10. Others still unclear.
3680: 13,12,11,10,10,10. Vox is at 13. Horde and Banana at 10. Others still unclear.

Summary by team, 3680 BC:
Mercs - 5-x-x.
Vox - 5-3-13.
Banana - 4-2-10.
Horde - 5-2-10, Plains Hill.
AC - 5-x-x.
Sara - 5-x-x.

Hammers to be accounted for: 2,2,1.
Commerce to be accounted for: 12,11,10.
2 Plains Hills remain unaccounted for.

16 tiles within rival city radii as of 3680 are water. One tile is next to the Horde capital, but the Horde controls no other water, so it's probably a freshwater lake. Another water tile is next to the capital of one of [Banana/Vox/Mercs]. Either Sara or AC has three tiles within its capital's radius, and 5 more appear to belong to that particular civ. That leaves 6 more unaccounted for. The Sara/AC that doesn't have 8 water tiles picked up 2 when it expanded, indicating that it has some, if not all, of the remaining mystery tiles. To put it another way, Sara managed to win the Fishing land lotto and is coastal. AC is coastal, and someone else (Banana/Vox/Mercs) is probably at least marginally coastal.

Vox is at pop 2. All others at pop 1. Should begin to change over the next 2 turns.

Tech: Vox and Sara have one civvy tech. Banana has researched Mining. AC popped either Hunting or AH. AC appears to be the only team that popped anything. Most probably, Vox and Sara both researched Myst. It is a certainty that Vox researched Myst, given their production totals and when their civvy tech completed. Mercs is researching something pricy - probably BW. AC is also going pricy - probably BW, but possibly the Wheel (we'll know that one next turn). Horde's research is still a big question mark.

Military: Banana built/popped a Warrior. No other Warriors built/popped; the other 1000 shift last turn is from Vox growing to size 2. No other actual military units built yet; new Scouts may be out there. Mercs are still totally undefended; they started with a Scout and apparently went Worker-first? Gutsy.

Neighbors: We have a trade route with Vox, so they're on this river SOMEWHERE. Passive religion spread could be a casus belli here for us or them, depending on who gets what religion-wise.

THREE of our rivals founded on Plains Hills and are working 3/0/x tiles. One is the Horde; the other two come from a pool of: Mercs, Sana, AC.

Unfounded speculation department:

Mercs probably is the unaccounted for tile-switching team post-border expansion. Can't prove it, but most likely they were the 9 commerce team and moved to a 10-11 commerce tile. (in other words: were working 4-2-9 on a Grassland Forest and moved to an Oasis/3-0-1.)

Mercs is 5-1-11 working an Oasis, and is researching BW.

Sara is 5-2-12 working an Oasis, is Plains Hill, finished Myst and is going for a religion.

AC is 5-2-10, is working a 3-0-1 food special on a river, and is researching BW.


Additional info to be gained on these turns:

3640 BC: Sara and AC should have grown when we get this turn. Otherwise, they went Worker-first.

3600 BC: Banana and Horde should have grown. Also, Sara or AC may tile switch this turn from what the governor automatically provided. This turn may be messy to figure out.

3560 BC: Possible Banana/Horde tile switch from governor production.

[Mudhut]

Re: KGB Update

Originally posted by Aginor

Rival land tiles: went from 12,333 to 13,666. Thus the Horde controls 8 land tiles and one water tile. Grog may want tank, but it looks like Khan want boat.

7 tiles within rival city radii as of 3920 are water. At least one of those tiles belongs to civs that founded in 4000, and the one tile that we know about is one space away from that civ's capital.

So the second part of this quote needs to be updated, yes? My proposal:

8 tiles within rival city radii as of 3920 are water. One tile is next to the Horde capital. Another water tile is next to the capital of one of [Banana/Vox/Mercs]. The other six tiles were picked up in border expansions - we have no idea (yet) who they are spread across.

These calculations are dependent upon this:

Originally posted by vulture
Going back to the land demographics, and sticking with the 1 tile = 1000 sq km assumption, the first two turns our competitors were working [9,9,8,0,0,0] tiles. One of them is on the coast, or next to a lake.

[Blake]

So it would be reasonable to assume that the first turn, the Horde moved their settler. That move revealed a nearby Plains Hill, and they figured it'd be a good place to settle, so they moved to the PH. The 3rd turn, they settled.

[Mudhut]

Actually, when I went to check the demo screen, the minimum land figure (at 3880) is now reading as 6,000 sq m.

This means that one of either Sana or AC have a 6-land tile, 3-water tile cultural border (ie they settled on the coast).

It is also possible that BOTH Sana and AC have water in their inner rim of tiles (when I say inner rim, I mean the 9 squares you can work before your first cultural pop).

So the complete discussion of rival tile sets:

The Horde have eight land tiles and one water tile in their inner rim.

Bana/Vox/Mercs are indistinguishable at the moment. Oneof them has an inner-rim tile that is water (ie, one of the eight tiles directly adjacent to their capital). All three have had their first cultural expansion. There may be up to three outer-rim water tiles distributed amongst them (outer rim = additional 12 squares you get to work at your first cultural borders pop).

Sana/AC are indistinguishable at the moment. One of them has three water tiles in their inner radius. The other has 0,1,2 or 3 water tiles in their inner radius -> obviously the water tiles they don't have are with Bana/Vox/Mercs in their outer rim.

So, does this help us at all?

[Aginor]

Originally posted by Mudhut
Actually, when I went to check the demo screen, the minimum land figure (at 3880) is now reading as 6,000 sq m.

So, does this help us at all?

Probably not immediately, but it's going to be very nice to know when the 3760 border expansion hits, as we can figure out a lot more about what Sana and AC actually have having accounted for 3 of the mystery 6 water tiles and assigned them specifically to one of the two civs.

Good catch on the 6000 - completely missed that.

Looks like 3840 is a pretty boring turn for the KGB - we expect to see exactly the same demographics screen on resources, and I don't think that anyone can finish Hunting/Fishing until 3800. Still, someone might pop something out of a hut, and that's something to keep an eye on.

[Mudhut]

Also, we can get something (not much) from the health line. The average/min/max for our rivals are all reading as 85. We are the *only* ones to deviate from this, at 70 years (15 years behind!).

Now, I'm not sure of the exact conversion from healthy/unhealthy to life expectancy on the demographics screen. But I suspect that this means we're the only ones with health issues -> and therefore the only flood plains start?

So, does anyone know the conversion from healthy/unhealthy into life expectancy? I'm at work or I'd figure it out right now.

[Mudhut]

Now, this isn't an exhaustive study, but it still might help. I took the sample start that NYE put together, and played with our terrain a bit. Here's the table:



Note that we're at the top, with -1 from Pop, -2 from flood plains, and +2 from forest, water, and difficultly level. Our life expectancy is 70.

The algorithm isn't linear, and I think we need to investigate a little more to nail it down BUT it *seems* that there is only one way to get 85 years of life expectancy. This is higlighted in yellow.

So preliminary analysis is that nobody else has a flood plain start (at least, not enough flood plains to get a health penalty from them) and that everybody has enough forests to get +2 health from them (5-6 forests, yes?). Not hugely revolutionary info, but it helps us narrow it down a bit more.

To make this more thorough, we should investigate:
-Do any techs change life expectancy?
-Do any health resources change life expectancy (in case someone settled on one....not that you would!)?
-Larger case set, to get more data points.
-Somebody check it, in case I've missed something important!

[vulture]

To put Mudhut's data in a slightly different format, I've made a table with net health modifier across the top, and negative health modifier (number of FP in our case) down the side

    +2  +3  +4  +5  +6 
 0  --  --  83  85  87
-1  --  71  75  77  --
-2  62  66  70  --  --

The jumps are more regular here. Going from +2 to +3 or +3 to +4 yields +4 years life expectancy. After that it's +2 years per additional health point. Going down from the 1st to 2nd rows is -8 year, 2nd to 3rd is -5 health.

Obviously this isn't enough data for the pattern to be conclusive. And it's possible that food surplus is factored in as well (and obviously some techs may well have an effect).

But it's a basis to start a systematic investigation for those who have more time than me...

[Aginor]

Originally posted by Mudhut
BUT it *seems* that there is only one way to get 85 years of life expectancy. This is higlighted in yellow.

So preliminary analysis is that nobody else has a flood plain start (at least, not enough flood plains to get a health penalty from them) and that everybody has enough forests to get +2 health from them (5-6 forests, yes?). Not hugely revolutionary info, but it helps us narrow it down a bit more.

Now THAT's some interesting speculation, as it leads to the immediate follow-up: if not flood plains, where in the hell are all these 3/0/1 tiles coming from??? I can conclusively prove the existence of two, and unless you choose to believe that the Horde is working a 3/0/0 and is on a 2/1/2 city tile, the third seems pretty probable.

I just have a hard time buying 7 civs with a river that's not in the desert, especially given that I *know* one of the initial 3 founders had a water tile in the starting radius (and there's a good chance it was an isolated freshwater lake as their source of fresh water). It's possible, I suppose, that the civ with that tile is working an Oasis, but that's the only way I see it.

Blake, feel like another trip into the game code to resolve this one?

[Blake]

Hmkay. I'll see what I can hunt down...
Grains/cow/sugar on a river is pretty common.

[Blake]

Here's the code.

if (pPlayer.calculateTotalCityHealthiness() > 0):
  iHealth = int((1.0 * pPlayer.calculateTotalCityHealthiness()) / (pPlayer.calculateTotalCityHealthiness() + \
    pPlayer.calculateTotalCityUnhealthiness()) * 100)
else:
  iHealth = 30

What it does in the case of a single city is simple:
goodhealth / (goodhealth+badhealth)


edit: got it wrong the first time.

[Blake]

The basic table of life expectency for good vs bad health.

[Blake]

Note that the approval rating is EXACTLY the same except with happy/unhappy. In other words, the same table will work for approval rating.

We can use approval rating as yet another measure of growth. It also indicates when happy resources are hooked up. Right now it's equal across the board, this means a civ with mining has NOT founded on gold/gems/silver, a civ with hunting has NOT founded on furs or ivory.
If there is a sudden jump in approval rating it means that a civ has founded on one of these resources, and then researched the appropriate tech.

[Blake]

I find it extremely interesting that no-one else has a floodplain starts - or at least not enough floodplains to cause any negative health.

I believe this gives us a decisive advantage in medium run commerce, very few things beat floodplains for good growth and high commerce. We can get out cottages up faster with less workers - there's no need to clear river banks of forest and such.
In short, we can grow fast AND research like crazy, to CoL or whatever we please.

It'll be a surprise if we don't pull into #1 Commerce once our cottages start coming online.

edit: **** I pentuple posted.

[Mudhut]

Blake, you are a champion. To my mind, this is the best news of the game so far. Exclusive flood plain access = very very strong position.

The only possible downside is that IF the other teams do the same analysis, they will know that there is one (and only one) team with a floodplains start. Their rival health averages would show up as ave=82, min = 70, max = 85 and there is a pretty clear resolution of that. Assuming they tracked the changes at the start correctly (and depending on where they sit in the turn order) they should be able to track it to us/AC/Sana. They may decide to try and take out the cottage-floodplain-giant early, simply because it will be so strong.

Our skirmishers must be ready