SGOTM 15 - One Short Straw

Mitchum said:

So what is our priority? Make our star a bit shinier or give our turtle a small push? Delhi is ready to go and a granary takes it to the next step by making the whip more efficient. A granary and a single pop growth in Zlatorog will take a lot of worker turns and time. Too late in my opinion. I say we send one worker back to Zlatorog after Delhi has it's granary.

Click to expand...

I'd have to take a look to be sure, which would be tomorrow evening at the earliest (and if I don't get time, you can play on), but aren't we moving a Worker from Marble 1NE toward the same Forest? Will said southern Worker be too late arriving to Chop said Forest at a reasonably efficient time?

I had been thinking about that too, Dhoom. Losing the city wouldn't kill us because all it's done is build workers. No infra. No pop. The problem is the worker would not be done in time if he beelined us on the next turn. We have to make it until T62. I think we miss by one turn. I wonder if we'd make it if we work the deer instead of the gold...

Dhoomstriker said:

There is, of course, the risk that the Barb Warrior would walk past our unit onto the Der, but if it did AND if we could complete the Worker before the Barb Warrior could reach our City, then we would have the choice to abandon the City and guard the chokepoint versus trying to defend the City and possibly losing our chock-point-blocker in addition to losing the City temporarily (basically until we get our first Axeman).

Click to expand...

I thought that barbs could not step on resources except to pillage. Since the deer is campless, I think the barb has to avoid it. Not sure though.

Dhoomstriker said:

I'd have to take a look to be sure, which would be tomorrow evening at the earliest (and if I don't get time, you can play on), but aren't we moving a Worker from Marble 1NE toward the same Forest? Will said southern Worker be too late arriving to Chop said Forest at a reasonably efficient time?

Click to expand...

I'm away from the game too. I think the worker could make it, but it would delay hooking up Candhi to Delhi and might result in worker turns lost to movement. Not the end of the world, but it's all a tradeoff.

Mitchum said:

For some reason, we get a 25% crossing river bonus (and a 50% forest bonus) in the test game.

Click to expand...

Not in the REAL SAVE. You can verify this by clicking on any of our warriors, hovering the mouse over the barb and alt-shift to see the odds. The barb gets no river-crossing defense bonus.

Dhoomstriker said:

aren't we moving a Worker from Marble 1NE toward the same Forest? Will said southern Worker be too late arriving to Chop said Forest at a reasonably efficient time?

Click to expand...

Yes, we are. The problem is that Delhi quickly grows to pop4 after T64, so we're in a hurry to complete its granary. Same with Marble and GEL, actually. We would have 1 wkr at GEL, 1 at Marble, and 2 at Delhi, including the one from Zlatorog. Sending the worker down from Zlatorog delays the camp 5t.

Okay, Big Picture. Our Kyoto units will come from GEL+Delhi+Marble. Here's the breakdown:

315h -- 9 axes (1 to guard Marble+Chandi from Toku's derelict archer)
300h -- 6 cats
50h -- rax in GEL
665h -- Total

hammers coming from:

120h -- 4 chops (other chops go into granaries, saving two at Delhi for the GLib)
360h -- 6*2whip (Delhi 2x, Marble 1x, GEL 3x, hopefully)
182h -- 13t*14hpt (Delhi+Marble+GEL+copper+gold)
662h -- Total

T64+13t = T77 plus 3t from GEL to Kyoto = T80 = 875BC

That won't be easy, but we surely don't want to go slower than that. Probably we can chop the Marble/Chandi defender in Chandi, etc.

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

So to do the above AND still grow Delhi rapidly up to pop4 so it can work 2 sci, we need to chop the Delhi+GEL+Marble granaries asap.

Maybe we could send two workers up, thus delaying the Marble granary. I haven't tried that.

Here's a T62 test save for testing various worker possibilities. I'm pretty sure the worker-turns are carved in stone up to this point.

Edit: You have to MM in Hunting if you want to start camping this turn...

.

Okay, given that high-level breakdown of LC's, I'm mostly interested in looking at Granary timing versus squares worked. So, if Zlatorog can't contribute, it sucks, but oh well, so be it.

@ Mitchum:
Hmmm, we should probably figure out when the "do or die" switch from the Gold to the Deer would need to happen if it could give us the Worker on time should the Barb Warrior beeline our City now... if we still can't make it happen, then I'm still fine with waiting for the Barb Warrior to make his move for 1 more turn, but then we'll have to move 1E GRiv For and then 1E Zlatorog if the Barb Warrior can go on the Deer.


I hadn't heard that bit about Barbs human units avoiding Resources... I remember from last game we discussed how Barb Animals avoid Resources unless there is a unit to attack on said Resource... but, it shouldn't be too hard to test... load up the saved game with a ton of AI Cities, then world-build in some extra Barb Warriors and Warriors of our own, turn after turn, just to see if we ever have a Barb Warrior walking on the Deer.


If we're delaying Hunting, then it would be nice to conclusively have convinced you guys of the merits of Warrior luring... I have done it in BtS... for example, I used it extensively in the Amazing Race BtS games that Mitchum and I played (non-BUFFY BtS games), since then we might be able to squeeze some Warriors out of other Cities other than the old-planned Warrior-build in Chandi.


1 Axeman should be a Spearman (which is why I don't see it as a waste to put Hammers into a Warrior if we can prove that we will return to building said within 10 turns), but the Hammer cost will not change, since a Spearman's cost = an Axeman's cost.

Dhoom, if you have time for testing, I see a possible solution. On T62 the Chandi wkr goes to Delhi-1NW, then he chops the Delhi-2N forest alone. The worker at Marlbe-1NE stays put and chops. Then the question is, how to optimize the granary in GEL? Is it better to chop and 1pop it from pop3>pop2 or grow quickly to pop4 and 1pop it to pop3? THen there's also how to deal best with Chandi, but I'm thinking that the chop doesn't have to come before the 1pop necessarily.

Dhoomstriker said:

1 Axeman should be a Spearman (which is why I don't see it as a waste to put Hammers into a Warrior if we can prove that we will return to building said within 10 turns), but the Hammer cost will not change, since a Spearman's cost = an Axeman's cost.

Click to expand...

Toku has had AH for an eternity and still doesn't threaten with chariots. I suggested that Mitchum check his threats each turn, but until we see chariots, I don't see any need for a spear right now. That said, GEL will produce many nearby units, so we could easily make a spear at the tail-end of our mobilization, if the situation changes.

@LC There are going to be hammers in barracks in both GEL and Delhi. You're proposing that we only finish the barracks in GEL for the Toku attack. That's fine. I just wanted to point out that we'll have hammers in Delhi too.

To Dhoom's point, should we have one spear in our stack in case Toku hooks up horses and build chariots? It would be brutal to have Toku attack our stack of axes/cats with chariots. If this is the case, we could change one axe to a spear and build a warrior for one turn in GEL rather than a barracks. That spear could just as easily kill a red-lined archer as an axe could.

Should I play until T62 with the warrior moves in the NW as outlined above (i.e. stay in place this turn, 1N if barb moves N or SW or 1E if the barb heads toward Zlatorog, 1E into Zlatorog if the barb moves to the deer, etc.)?

Should I consider stop working the gold and working the deer to finish the worker in 3 turns? Rigtht now, I have 45/60 hammers in the worker at +4H/turn. If I work the deer, I'd get +5H/turn. I could work the deer for this turn only. If the barb moves N or SW, I'll go back to the gold next turn. If he keeps moving toward Zlatorog, I'll keep working the deer until the warrior is killed or turns back. If we keeps coming, we'll finish the worker... wait. Dang, the worker will be done but it will be the IBT when the warrior attacks Zlatorog, killing the worker that was just created. We'd be a turn too late. Unless the barb kept the worker, in which case we could steal him back...

What event(s) should trigger a stop and upload?

1. Oracle is built in a distant land
2. Barb attacks and kills our archer in the NW either on a forest or in Zlatorog
3. Others?

EDIT: X-Post

I just ran a few tests:

Test 1: I put a stack of crossbows at Zlatorog-W (i.e. deer-S) and a stack of barb warriors at deer-W. All 14 barb warriors attacked my stack of crossbows at 0% odds.

Test 2: Same as above but I put a camp on the deer resource. All barbs suicided against my stack until there were two barbs left. Those two barbs went to the deer camp.

Test 3: Same as above but I moved my stack to the deer camp rather than at deer-S. All barbs suicided against my stack until there were two barbs left. Those two barbs went to the open tile at deer-S.

So, based on this quick test, it appears that human barbs will not step on unimproved resources but will instead attack at impossible odds. If there is an improvement, they will consider not attacking but pillaging instead.

Strange. When LC mentioned that Toku wanted Mining I didn't believe him because Mining was not listed as a tech that Toku wanted. However, when I went back, I see that Mining is listed as a tech that Toku can research. Why is Mining not listed as a tech the Toku wants since we know it and could trade it to him? He would have "Nothing to gain" if we were to give him Mining. Crazy AI...

@Mitchum
1. I prefer barracks at GEL because it will produce more units now, grow faster, etc.
2. The Zlat warrior can actually go 1E to the forest this turn. (I think Dhoom was suggesting this.) Next turn the barb's either on our chokepoint or it's safe to move to the chokepoint. That potentially gives us 5% fortified on that tile.
3. I'm currently testing the possibility of camping the deer starting T62. It looks promising, but I have no idea if it's optimal. So if you want to play to T62, then research Hunting in the process, maybe on T61 in 1t, if that will work out.
4. I would keep working the Zlat gold tile.
5. Maybe Toku just traded for Mining is is about to finish it or something.

Yes, that means Toku is currently researching Mining, and is over 50% done. He really values the privilege of saying "no, we won't trade you this, but it's out of choice"

Barbarian_leader has a large AttackOddsChangeRand, meaning on some turns he'll suicide every unit at 0.0001% odds, but on others he won't attack below about 30% odds. It's determined globally every turn.

Sounds good. I'll play until T61. I'll research Hunting rather than Aesthetics on T59 at 0%. If we decide to start camping on T62, we can up the slider to 100% on T61 and learn Hunting on T62 (at least in the test game we have enough beakers).

So we'll have some hammers in a barracks in Delhi that will get completed some time after we build up our inital Toku attack force.

I think we need to plan to have a few extra axes in our initial Toku stack so that we can take out Osaka on the first turn (T0) of the war. If we do this, we can actually get to Kyoto one turn sooner since we can march our stack through the captured Osaka and get within 2 tiles of Kyoto on the first turn of the war. That gets us next to Kyoto on T+1 with an attack on T+2.

If we bypass Osaka, we'll be next to Kyoto on T+2 for an attack on T+3

FYI. I plan to play in about an hour to T61. Speak now...

Play on!

ZPV, is the formula to determine how much tech value we can get from Toku for a DoP, something like this:

Silu said:

Begging mechanics for non-vassal deals:

Maximum amount that is accepted is

(TURNS_KNOWN + 50) * 2 * (OUR_POWER / THEIR_POWER), further multiplied by 3 if that AI is our land target.

Note that this remembers earlier gifts granted, so if this formula gives a value of 1000, and we already have begged 700 during the game, the max gift value granted is only 300. Again, the value of gold is GOLD*2, or GOLD*3 if the AI is in financial trouble. Beakers are approx BEAKERS*1.5.

Also it's very important to note that the beg is refused outright if they haven't forgot all the former begs (successful and unsuccessful). AIs have a 5% chance of forgetting one beg each turn. So, the optimal begging rotation regarding this is a somewhat complicated optimization problem. Once upon a time I did simulations to find out the best beg interval regarding just this forgetting mechanic, results are here.


Surprisingly, demands (for Civs below Pleased) work exactly the same, except they're refused outright if their power is more than 1.33x ours. And of course they give a diplo penalty.

Click to expand...

It would be interesting to konw if there's any way we could get Alpha from Toku. I think our power should be more than 1.3 his (Agr AI modifier?) after we mobilize.

Mitchum said:

Dang, the worker will be done but it will be the IBT when the warrior attacks Zlatorog, killing the worker that was just created. We'd be a turn too late. Unless the barb kept the worker, in which case we could steal him back...

Click to expand...

I suppose that if you have time, you could test whether a Barb unit would capture a Worker inside of a City and if the Barb would actually keep the Worker, we probably should work the Deer for this turn, just to keep our options open.

If, however the Worker would get auto-killed, then just continue working the Gold Mine and we'll be forced to try and defend Zlatorog should the Barb Warrior come after us.

Mitchum said:

So, based on this quick test, it appears that human barbs will not step on unimproved resources but will instead attack at impossible odds. If there is an improvement, they will consider not attacking but pillaging instead.

Click to expand...

Wow, that was cool.

The implication to me is that:
If the Barb Warrior comes at us, we definitely want to be on the GFor Riv that is 1W of Zlatorog.

Now, if we are willing to say that:
Next turn, unless the Barb Warrior moves to the chokepoint, we will move to the chokepoint, then on the current turn, we can already move 1E GRiv For.

Since we are willing to blindly move into the chokepoint if the Barb Warrior is not there, we don't really care about seeing over the Lake, so we might as well move 1E to the GRiv For so that we'll get an extra 5% Fortification bonus if the Barb Warrior does move to the chokepoint.

Mitchum said:

Strange. When LC mentioned that Toku wanted Mining I didn't believe him because Mining was not listed as a tech that Toku wanted. However, when I went back, I see that Mining is listed as a tech that Toku can research. Why is Mining not listed as a tech the Toku wants since we know it and could trade it to him? He would have "Nothing to gain" if we were to give him Mining. Crazy AI...

Click to expand...

My guess? He just learned Mining in a trade.

However, on the turn that you learn a tech, the corresponding F4 screen looks a bit weird... like, if you get Code of Laws in trade, Philosophy might appear in the "Won't Trade" or "Can't Trade" column.

So, Toku very well may already know Mining thanks to having immediately traded for it upon learning Alphabet.

That's just a guess, but it may explain the discrepancy.

LowtherCastle said:

1. I prefer barracks at GEL because it will produce more units now, grow faster, etc.

Click to expand...

Is that Barracks in place of buiding a Barracks in Delhi, or have we already invested the Hammers in a Delhi Barracks?

LowtherCastle said:

2. The Zlat warrior can actually go 1E to the forest this turn. (I think Dhoom was suggesting this.)

Click to expand...

Hmmm, it seems that you read my mind.

LowtherCastle said:

4. I would keep working the Zlat gold tile.

Click to expand...

Given the evidence presented by Mitchum's quick testing, I would agree.

LowtherCastle said:

5. Maybe Toku just traded for Mining is is about to finish it or something.

Click to expand...

Hmmm, it seems that LC are on the same wavelength about a lot of things... just perhaps not about building a Warrior versus a Barracks for a partial build item.

LowtherCastle said:

It would be interesting to konw if there's any way we could get Alpha from Toku. I think our power should be more than 1.3 his (Agr AI modifier?) after we mobilize.

Click to expand...

In my test, I captured both Kyoto (captial) and Osaka (his capital at the time I took it since I captured Kyoto first in this test), killing 8 to 10 archers. Toku still had 2 cities left and we wouldn't give me crap for peace. I think he would give Archery and nothing else. I did not check our power levels at the time, however.