SGOTM 12 - One Short Straw

Attacking: win: 2, lose:3
Defending: win: 1, lose:2

Capture unit: 2
Let a unit be captured: 1
Capture city: 6
Hit by nuke: 3
Use a nuke: 12

call this battle WW.

We keep a running total, called team WW.
After each battle, the following is added to team WW:
battleWW * (100+TheirCulture)/(OurCulture+TheirCulture)
where TheirCulture and OurCulture are Plot Culture values. These will be high everywhere except the outer cultural ring of a city, or if the city hasn't popped its borders yet (in which case they will be so low that there seems to be a WW spike for attacking brand new cities.)

If neither side has any culture on the tile, then we can ignore the second factor.

Finally, to determine the war :mad: we see in cities, we add up the Team WWs vs. all the teams we are at war with (call it Active WW), and use the following formula:

ActiveWW*Population/500

For happiness, truncation doesn't happen until all of the anger causes have been summed, so it could happen that two unhappiness causes have non-integral anger values, and then WW could appear to be the ceiling of this formula, rather than the floor.

It can look like there is an era-dependant factor, because there are many more units later in the game, and the cities are bigger, but there isn't anything directly in the code (except for the nuke proviso). I wonder if we'll start with a little war weariness from nukes. (We'll be able to tell from the scoreboard right away when we load the save). That would let us know if we have some war success too.
 
Just speculation, but if the AI do have some advanced techs that we don't then espionage and espionage buildings might be significantly more valuable than a regular game.
 
So, the big points of discussion so far sound like they are (in my perceived order of importance):

1) Where to settle our capital
2) How soon do we go for alphabet
3) What (if any) are we going to seek from Oracle.

Obviously there's a lot of interplay between the three since it affects our initial build orders etc. But I think the general permutations are some combination of SIP, settling NW, NE, and NEN? Do we want to divide up the testing and have people test one permutation a few times? I can do some testing this weekend.
Still to be tested, imo, is SIP REX, and all sorts of NEN possibilities, including CS sling, Currency or CoL sling, REX, REX+GLH.

One more thing, Erkon suggested teams to build test games with fractal and some islands. That's one other carryover from SG10 that I'm assuming: some islands around our capital and elsewhere. That could mean:
* wb exploration may be vital
* GLH is better with 3 offshore cities = 2gpt minimim for each trade route CIV-wide.
* SIP may be safe, in that the clams, for example, may be accesible by an island.

So a round of testing with REXing to some islands (Sailing+galley) would be a good thing too.
 
Just speculation, but if the AI do have some advanced techs that we don't then espionage and espionage buildings might be significantly more valuable than a regular game.

Can we steal techs we can't research ourselves?
 
mdy said:
Can we steal techs we can't research ourselves?
I don't think so, we'd need to know the prereqs...

Spying idea sounds nice but last few pages were mostly about reading AI's demos and graphs, we can't really split EPs on everyone and still hope to steal more than a few techs. Also if we really want to go down that route we'd need the GW which collides with nearly every early goal we have - plus, we're not industrious and have no stone or marble. I think we should just forget about stealing.

RE settling... I'd really like a coastal capital (can't believe I'm saying this) but it will get our capital up so much faster as we don't need to clear any fallout on seafood... after we have fishing and our first worker we can whip out 1-2 WBs and then we can really start whipping out settlers/workers/military while teching towards some wonder. Also we get earlier trade routes and benefit from GLH, that bonus is really big if we get some island cities... I'm for settling NNE, as it's looking like the strongest site to me and leaves space for a second city.
 
I had no trouble loading the save which is up.

We have the same relations and WW with all civs, they differ in score however.

sg-12-start.jpg
 
Nice to see the save is up.
I'm playing around with NNE, going for a 3-city CS sling. It feels fast, but it's tough to get all of the worker actions and tile selections right. I'll let you know what date I can manage when I've done it.
After that I'll load the save and check the event log to see what Erkon has done.
I suspect he's suicided a bunch of our units, to cancel out our war success. OTOH, it doesn't look like we're up against a bunch of futuristic enemies.
 
ZPV (Silu), I was wondering if you could check the formula for the tech bonus we get from AIs knowing techs already. What I'm wondering is when it's round down to integers and when it's rounded down to hundredths (or whatever). Here's what I'm using:

ROUNDDOWN(

ROUNDDOWN((BaseResearch + 1)*(1+

ROUNDDOWN(0.3*KnownAIsWhoKnowTech/TotalPlayersAtStart,RoundDownTo???))

,RoundDownTo???)

*TechPrerequisiteBonus,0)
 
LC, the formula is
[1 + (Baseresearch*[100+30*KnownTeamsWTech/TeamsAlive+20*Prereqs]/100)]
where I've used [] to indicate the floor (rounding down to the nearest integer)
 
I don't have much so far. Just first set of guesses...

Life Expectancy:
Rival Worst = 57 = 8/(8+6) --> Mao: +2 (bonus) +2 (fresh water) +2 (EXP) +2 (forest) -1 (Pop) -5 (fallout - 10 tiles)
Rival Best = 77 = 7/(7+2) --> +2 (bonus) +2 (fresh water) +3 (forest) -1 (Pop) -1 (fallout - 2 tiles)
Rival Avg = 68 = approx. 6/(6+3) --> +2 (bonus) +2 (fresh water) +2 (forest) -1 (Pop) -2 (fallout - 4 tiles)

Those numbers are a bit weird. I think there may be quite a bit of variation. No one is unhealthy at size 2, though.

Land Area:
One AI with 8/9 tiles, one with 6/9 tiles, four with 5/9 tiles. The borders will pop in 5t, giving us more info.

Crop+MFG:
It's too low. Typically, you would start the game with +6 total (2f1h from CC + another 3f / 2f1h / 1f2h / 0f3h tile). Doesn't appear that anyone has total production of 6. Not sure what to make of it...

GNP:
Standard. It's 10-12bpt, which translates into 8 from Palace, 1 from CC and 1 more (plus 1.2x bonus in 3-4 cases).

Approval:
Standard. We have 2 CHA leaders with 85, the rest with 83.

Soldiers:
Scary. :eek: 230k is massive. It's obviously not just tech. Stalin doesn't have enough score to get more than 10-12k in tech, which gives him the equivalent of 5 ICBM's OR 7 Tactical Nukes. We need to test this for even getting a peace treaty.
The others are higher than expected, as well. This is what it should be, without further techs or units:

Churchill (Fish+Mini) 8000
Mao (Agri+Mini) 8000
Gandhi (Mini+Myst) 8000
De Gaulle (Agri+Wheel) 12000
Roosevelt (Agri+Fish) 6000

It adds up to 42k, which leaves 12k missing.

Edit:
SCORE:
Haven't had a chance to really study it. Gandhi is at +9 for Alpha and 1 pop. We have the equivalent of 8 techs for 47, which puts a tech at just under 6, as usual. That means 1 pop = 3 or 4 --> Stalin, Mao & Church have +24 in tech, DG & Roos +18.
 
Still to be tested, imo, is SIP REX, and all sorts of NEN possibilities, including CS sling, Currency or CoL sling, REX, REX+GLH.
I'll do some of that this weekend, since I've been testing inland so far.
COL or Currency sling sounds better now that Gandhi has Alpha. We could maybe do something like: peace with Wtg >> Gift Math for OB >> Currency for Alpha >> gift to others?
 
Nice work bbp :)

bbp said:
Soldiers:
Scary. 230k is massive. It's obviously not just tech. Stalin doesn't have enough score to get more than 10-12k in tech, which gives him the equivalent of 5 ICBM's OR 7 Tactical Nukes. We need to test this for even getting a peace treaty.
The others are higher than expected, as well. This is what it should be, without further techs or units:

Churchill (Fish+Mini) 8000
Mao (Agri+Mini) 8000
Gandhi (Mini+Myst) 8000
De Gaulle (Agri+Wheel) 12000
Roosevelt (Agri+Fish) 6000

It adds up to 42k, which leaves 12k missing.

Could the missing power come from other starting techs, you didn't mention them.

Well the only interesting thing for now are crop yield, mfg, power and gnp while most are too early to read properly (like, we can't determine how many are coastal right now as they didn't put any fishing boats out (well, maybe somebodydid for them...)), only stat that could help us find out more about the AIs is power + score.
 
Yeah, the extra power is from techs. Impossible to say which, though.

The low land areas do suggest coastal for all but maybe one AI. Seems unlikely you could have less then 7-8 land tiles if you're inland. If they were indeed given WBs, we'd see a Crop Yield spike a.s.a. next turn...
 
Well, my main question was, did you take into account all the starting techs an emperor AI gets? If you did, I don't understand what this list of civ-specific starting techs is about.
 
After taking a cursory look at the scores, here is the situation:

(actual game - test game)

Stalin 74 - 74
China 74 - 80
English 74 - 80
American 68 - 80
French 68 - 80
Indian 56 - 80
OSS 47 - 47

The only things that affect score on T0 are pops (1 for everyone), Wonders (everyone has a palace) and techs. Everyone's score for land will be 0 until T10 IIRC.

If I take Ecology away from Ghandi and give him Alphabet, his score drops to 56 to match the real save. The French and Americans likely don't know Ecology either since their scores are too low. China and England either know Ecology and not some other basic starting tech or they don't know Ecology but instead know a tech from the era before Ecology.

In a nutshell, I think this indicates that many of the AI don't know Ecology because the don't need it... i.e. the don't have any fallout.

I'll try to look deeper at the scores and demographics later today.
 
Take a look at the event and combat logs.
We've exploded about 50 nukes (that's how the map got this way). There is location info hidden in the event logs too, but I don't know how to get it.
and we've suicided 70ish warriors against enemy Modern Armour.
I'd guess no dice for immediate peace, but hopefully we'll be able to trade with Gandhi soonish.

Espionage should go to Gandhi, so we don't overlap research.
 
Back
Top Bottom