SGOTM 15 - One Short Straw

I'm good with it as well

ZPV said:

Preparation time = 10 turns. If he still hasn't decided he's ready after an additional 10 turns, he drops WHEOOHRN.
He decides he is ready, if the landmass gets AREAAI_OFFENSIVE.

Click to expand...

I seem to recall (and it would make sense) that if we are at war with another AI (i.e. Toku), then an AI (such as Hammy) will have greater incentive to declare war on us.

I forget why, but perhaps it is because the Power calculation will take into account both AIs' Power, meaning that if we are hoping to avoid being declared upon by Hammy via increasing our Power, we will probably have to delay the declaration of war on Toku until after Hammy has given up on attacking us.


Another random thought: if we go out and meet the other AIs, then they can be bribed into war against us or can choose to dog-pile us in a war. However, if we have yet to meet an AI, as much as they dislike us for having declared war on their friends, they will be unable to actually declare war on us until we meet them. So, going out and exploring the world (if such an opportunity presents itself) should probably only be done when we feel confident that our remaining military force can survive further war declarations.

LowtherCastle said:

Note: Hatty completed IW on T65, but doesn't threaten with anything but War Chariots so far. Hammy does not have IW yet.

Click to expand...

If Hatty used Iron to build a Spearman, Axeman, or Swordsman, would any of those units appear in place of the War Chariot bragging, either replacing the War Chariot message or else alternating between the War Chariot message and its own message?

LowtherCastle said:

We could finish the chop this turn into a scout or an axe, either completed next turn. An axe shouldn't have any problem surviving a barb and is useful as a defender.

Click to expand...

I'm fine with putting a Chop into an Axeman and delaying the Granary by building 1 or more additional Axemen. What I would suggest, though, is that:
1. We keep our Warrior defending the Forest chokepoint for now
THEN
2. We move our Axeman to the Forest chokepoint
THEN
3. As soon as the Axeman arrives there, send the Warrior out toward the NW to do a bit of advance scouting and possibly even army delaying... if Hammy has sent a stack, then a lone Warrior could potentially lure some of the stack back north temporarily, buying us extra turns to survive and build more of an army

I have "toyed" with AI stacks before... sometimes you can split the stack, sometimes you can lure the entire stack, and other times, if you are lucky, you can get the entire stack to stop in place and wait while it heals up all of the units in the stack before moving. That last possibility is unlikely with Hammy's Axeman attacking our Warrior, but we could end up with a reasonable luring situation, although I'm not sure if we'd need Hammy to have line-of-sight on our unit in order to be lured by our Warrior or if we'd even need to move our Warrior adjacent to his stack in order to properly lure part or all of it back to the north.


One concern that I do have is that if Zlatorog's Cultural Borders are nowhere near expansion (which they probably aren't), then Hammy is unlikely to declare war until he "enters" our Cultural Borders. In that sense, the "chokepoint" isn't going to work at all, since he'll first move his stack onto the chokepoint while being at Peace and then will declare war on the following turn when he "enters" our Cultural Borders on a square (likely a Forest, if there is one, which could include the Deer) that is adjacent to Zlatorog.

So, it might be worth testing to see what would happen if Hammurabi moved his units from the GRiv square to the NW of the GRiv For chokepoint onto the chokepoint and then we declared war on him... would his units get teleported back to the NW or would they get teleported to the S, or would our unit get teleported off of the chokepoint?

An alternative approach (and one that wouldn't require setting up a potentially complicated test) would be to just plan to pre-emptively declare war on Hammurabi once his units appear on the GRiv to the NW of the GRiv For chokepoint.

ZPV said:

@Dhoom: a fort won't really provide much benefit against a CR1 axe.

Click to expand...

True, although it will help more against Combat I Axes, and the AIs don't tend to build stacks with purely City Raider I promotions. Still, it's a moot point as we don't really have the Worker turns available at the present moment.

Dhoom:
If we are at war, there is a chance for AIs to spontaneously dogpile, without even being bribed.

Sending out the warrior to scout rather than the axe is risky, because there might be barbs trapped between us and the Great-Walled Hammy.

If we DoW, our units get teleported. If he DoWs, his units get teleported towards his nearest city.

@ZPV -- Urgent

Dhoom triggers some interesting and perhaps critical code-diving questions.

1. Will Hammy DoW if he encounters our axe out in the field or wil he wait till his SoD arrives at our borders?
2. If borders, does he DoW on the turn he's able to enter our borders?
3. If DoWing the turn entering, will he do that from a tile we occupy? (Because the act of DoWing will teleport his SoD 1NW, closer to his nearest city. Or does he prefer to move 1S to a vacant tile, then DoW the ensuing turn?
4. If he only DoWs next to our border froma vacant tile, can we essentially prevent him from ever DoWing us, by occupying all three western tiles, adjacent to our borders?

LowtherCastle said:

@ZPV -- Urgent

Dhoom triggers some interesting and perhaps critical code-diving questions.

1. Will Hammy DoW if he encounters our axe out in the field or wil he wait till his SoD arrives at our borders?
2. If borders, does he DoW on the turn he's able to enter our borders?
3. If DoWing before entering, will he do that from a tile we occupy? (Because the act of DoWing will teleport his SoD 1NW, closer to his nearest city. Or does he prefer to move 1S to a vacant tile, then DoW the ensuing turn?
4. If he only DoWs next to our border froma vacant tile, can we essentially prevent him from ever DoWing us, by occupying all three western tiles, adjacent to our borders?

Click to expand...

1. He'll do it at our borders. He won't pick off our axe in the field.
2. The turn borders get in the way of his path to his target city. So probably in our case.
3. I'll see if I can tell. Ditto #4

ZPV said:

Found it. He needs a SoD and a target city. If he has those, he can even switch war plan early.
He'll be coming once the stack is together.

Click to expand...

DanF5771 said:

After a fixed number of turns it checks whether it has accomplished to spam enough offensive units to attack (I have also described it in this post to understand what was going on in Sisiutil's ALC#24). If yes the battle plans change to AREAAI_OFFENSIVE / AREAAI_ASSAULT and the war plans change to WARPLAN_LIMITED or WARPLAN_TOTAL which is followed by the actual DoW.

Click to expand...

ZPV, just to double-check my understanding, your code-diving shows that DanF is wrong that this occurs only AFTER a fixed number of turns?

LowtherCastle said:

ZPV, just to double-check my understanding, your code-diving shows that DanF is wrong that this occurs only AFTER a fixed number of turns?

Click to expand...

The way he describes only happens after the 10 (normal speed) turns.
There is another way to change the AI's warplan, triggered by an assembled city_attack stack, and that doesn't have the time constraint.
edit: I'll double-check the AreaAI test, to make sure it is every turn and not only after 10 turns.
edit2: I've done the double-check. He can definitely hit AREAAI_OFFENSIVE on any turn.
edit3: the stack won't march until the warplan changes. This is very convoluted, so I can't be too sure of my findings.

Okay. Some more questions, in case we can't prevent a DoW with our wall of western units. This time to figure out how soon he'll assemble his SoD.

DanF5771 said:

As part of his (Saladin) personality he also favors to build units with (the wrong) UNITAI_COUNTER (XML-value UnitAIWeightModifier) during times of peace. Together with his protective trait this turns him more into a strong defender than a real threatening offensive aggressor (compare to Monty!).

Click to expand...

Toku is also protective and the XML has his preferred unit as CITY_COUNTER.
1. Can he assign CITY_COUNTER to axe and spear builds?
2. Is WHEOOHRN still a "time of peace" or does he stop favoring CITY_COUNTER under WHEOOHRN?
3. When does he make the CITY_COUNTER (or other) assignment, when he first starts the build or when it is completed? Because he completed 2 units the turn after he went WHEOOHRN.

ZPV said:

Dhoom:
If we are at war, there is a chance for AIs to spontaneously dogpile, without even being bribed.

Click to expand...

If we don't meet an AI, it cannot dogpile us.

ZPV said:

Sending out the warrior to scout rather than the axe is risky, because there might be barbs trapped between us and the Great-Walled Hammy.

Click to expand...

Then maybe we shouldn't explore to the north-west of our empire with either a Warrior or an Axeman. Having a Barb unit between us and Hammurabi might slow down Hammurabi a bit when he comes, by luring his forces away or to split up.

Better to have him encounter the Barb units than to have our unit clear a path for his full stack, right?

Another way of looking at the situation: let's say that Hammy sends, say, 4 Axemen and 1 Spearman. It could be that by having 2 defending Axes on our chokepoint, we could lose both of our units in those battle. But, if we have 2 defending Axeman and a defending Warrior, our Warrior will likely die but will come up as a defender ahead of a badly-wounded Axemen of ours, saving an Axeman and even maybe winning a fight versus an attacking Spearman. So, perhaps having our Warrior stay on the chokepoint along with the Axeman is going to be the best approach, followed by building yet another Axeman for that chokepoint. Doing so would ruin the "lure Hammy's stack" idea, but if by not going toward the NW, we keep at least 1 Barb unit alive, then this luring might already end up taking place, thereby delaying Hammurabi's war.

Even if a Barb unit attacks Hammurabi's stack and wounds a unit in it, the time that Hammurabi will spend (1 or 2 turns) healing with his stack could buy us some valuable time.

Knowing when Hammurabi will come won't change the timing of when he does come, so perhaps we should just assume that Hammurabi is "already on his way" and make appropriate preparations with that assumption in mind.

ZPV said:

If we DoW, our units get teleported. If he DoWs, his units get teleported towards his nearest city.

Click to expand...

But if he DoWs by "entering" our territory, while his units are standing on top of the chokepoint that we are standing on top of, assuming that the chokepoint is neutral territory, what will happen to his units? Will they just bypass the chokepoint? Will his first unit bypass the chokepoint and the rest be teleported to the NW of the chokepoint?

A silly question, but what turn are we on? The SGOTM server says that the latest saved game is 1200 BC, but is that the latest? Basically, what's the test saved game that corresponds to the current turn or is at least the test saved game that comes just before the current turn, if we don't have an up-to-date test saved game?

klarius said:

There will be no large SoD build. That's only happening while preparing a MAX_WAR. ATTACK or ATTACK_CITY units will collect up to 2 units to form a mini stack.
ATTACK and COUNTER units will prefer to stack with an ATTACK_CITY group head, if not available then ATTACK. ATTACK_CITY will not stack at all to other units.

As soon as the AI has enough CITY_DEFENSE, ATTACK and CITY_ATTACK (based on number of cities and population), it wants all the other stuff also (with the RNG playing a big role).

Click to expand...

klarius said:

What's the birthmark of the sword . That's a random number set when the unit was born.

OK, the sword (assuming attack city) can only leave via target city, if nothing to attack is in range. As the AREAAI is defensive for Fred (WARPLAN_ATTACKED_RECENT), this is only possible if the stack size is greater or equal stackOfDoomExtra. This number is 2 + birthmark % 7, so a random number between 2 and 8. Still the city is not allowed to get under defended by them leaving.

Click to expand...

klarius said:

Another thing: Counter and reserve units group with the stackOfDoom boolean set.
That means the stack size they still add to, is not only 2 (as it looks at the first glance ), but 4-10 depending on the birthmark of the group head.

And another fact that might be a problem. An attack city or attack group head and his stack will wait for units that want to stack with them, if located inside own territory.

Click to expand...

Here's some stuff from SG7, which was Vanilla, btw, but maybe hasn't changed significantly.

We know that Hammy must be preparing for Total_War (what klarius is calling MAX_WAR I think) because we are not a land target of his, assuming he went WHEOOHRN because of us. So my takeaway from this is that he will want to create SoDs as large as the birthmark of the group head allows, 2, 4, 6, 8, or 10. Or, as large as is required by the current TotPop+#ofCities, which would be 6, if 5 is not possible for boolean reasons.

Also, it shouldn't be a problem that TOku likes to create CITY_COUNTER, because they happily attach to group heads, so he only needs enough group heads (1, 2 or 3) to collect 6 units total. Those group heads need to be ATTACK_CITY or ATTACK and he needs at least one of them.

Does that sound about right, ZPV?

Dhoomstriker said:

Then maybe we shouldn't explore to the north-west of our empire with either a Warrior or an Axeman. Having a Barb unit between us and Hammurabi might slow down Hammurabi a bit when he comes, by luring his forces away or to split up.

Better to have him encounter the Barb units than to have our unit clear a path for his full stack, right?

Another way of looking at the situation: let's say that Hammy sends, say, 4 Axemen and 1 Spearman. It could be that by having 2 defending Axes on our chokepoint, we could lose both of our units in those battle. But, if we have 2 defending Axeman and a defending Warrior, our Warrior will likely die but will come up as a defender ahead of a badly-wounded Axemen of ours, saving an Axeman and even maybe winning a fight versus an attacking Spearman. So, perhaps having our Warrior stay on the chokepoint along with the Axeman is going to be the best approach, followed by building yet another Axeman for that chokepoint. Doing so would ruin the "lure Hammy's stack" idea, but if by not going toward the NW, we keep at least 1 Barb unit alive, then this luring might already end up taking place, thereby delaying Hammurabi's war.

Even if a Barb unit attacks Hammurabi's stack and wounds a unit in it, the time that Hammurabi will spend (1 or 2 turns) healing with his stack could buy us some valuable time.

Knowing when Hammurabi will come won't change the timing of when he does come, so perhaps we should just assume that Hammurabi is "already on his way" and make appropriate preparations with that assumption in mind.

Click to expand...

The possibilities are:
we can build an axe now and scout, so we know exactly how many units we need and when
or we can build lots of extra units, delaying our plans to invade Toku (which we have invested a lot, including an Oracle slingshot, in).

But if he DoWs by "entering" our territory, while his units are standing on top of the chokepoint that we are standing on top of, assuming that the chokepoint is neutral territory, what will happen to his units? Will they just bypass the chokepoint? Will his first unit bypass the chokepoint and the rest be teleported to the NW of the chokepoint?

Click to expand...

You don't DoW by entering territory. You DoW and then your unit moves. So all his units get teleported.

We're on T70 1200BC, as uploaded.

Dhoomstriker said:

Knowing when Hammurabi will come won't change the timing of when he does come, so perhaps we should just assume that Hammurabi is "already on his way" and make appropriate preparations with that assumption in mind.

Click to expand...

We want to leverage each war for tech "trading." We're going to need Iron Working before we can bulb Machinery+Engineering, so we'd like to get it with a DoP. To do that with Hammy, we'll need to be able to threaten one of his cities, in addition to some war success. So we need to know where his nearest cities are and have a unit on location to threaten it. Of course, Hammy will also have to research IW for us, but we could take other tech(s) instead, and get IW from a DoP with Hatty. (And we need to get Alpha from Toku before any other tech trade.)

Or an alternative. If we capture a worker or scout with no losses, we'll be able to CF/DoP after 5 turns, if we threaten a Hammy city. That means we'd have to do that with his SoD farther than 5 turns + 2 tiles (threaten distance) from Zlatorog. That would be pretty cool. I think if we DoW him and then CF, that resets his WHEOOHRN.

In any case, I think it's in our interest to send an axe out scouting, so we can know how far away he is and whether we can build the Zlatorog granary or units.

LowtherCastle said:

Here's some stuff from SG7, which was Vanilla, btw, but maybe hasn't changed significantly.

We know that Hammy must be preparing for Total_War (what klarius is calling MAX_WAR I think) because we are not a land target of his, assuming he went WHEOOHRN because of us. So my takeaway from this is that he will want to create SoDs as large as the birthmark of the group head allows, 2, 4, 6, 8, or 10. Or, as large as is required by the current TotPop+#ofCities, which would be 6, if 5 is not possible for boolean reasons.

Also, it shouldn't be a problem that TOku likes to create CITY_COUNTER, because they happily attach to group heads, so he only needs enough group heads (1, 2 or 3) to collect 6 units total. Those group heads need to be ATTACK_CITY or ATTACK and he needs at least one of them.

Does that sound about right, ZPV?

Click to expand...

The numbers have changed since vanilla, but the principle of units attaching themselves to a group head hasn't changed.

The spearman is probably CITY_COUNTER. The axemen can't be.
the CITY_COUNTER unit doesn't count towards the number needed for AREAAI_OFFENSIVE - he needs 6 of those.

AIStackOfDoomExtra is now ((AI_getBirthmark() % (1 + GET_PLAYER(getOwnerINLINE()).getCurrentEra())) + 4). i.e. It's Birthmark%2 + 4 most likely, so it's 4 or 5.

Okay, sure, send out the Axeman scouting and leave the Warrior on the chokepoint.

We may kill off Barb units but then at least (as long as we don't lose our Axeman to a Barb unit) we will have a unit "in the field" that can potentially lure Hammurabi's units or can perhaps threaten his City from a Forested square (assuming that war breaks out).

Okay, ZPV, the question is, should I continue now or not?

Do we need to test my idea of preventing the DoW? What's the easiest way to set up an AI to WHEOOHRN us? I suppose we could make ourselves a land target of Monty, but using mountain tiles for the common borders, so that we could set up a 3-tile window along our border and put units on the 3 tiles outside our borders and see if his SoD will DoW when it lands on an occupied tile.

I did a bit of playing around in an older test game, trying to get an AI to declare war on us, but I have yet to be successful. Anyway, I did have these random observations:

Hammurabi won't accept the Deer + Cow City unless he happens to still only be at 3 Cities when we attempt to gift it to him... this point may differ in the real game if Hammurabi's City isn't where we think it is, but we had better not count on this City-gifting possibility unless we are also willing to gift away Zlatorog (which I don't think is a great option, now that we have decided to grow Zlatorog).


If we obtain a City with Buddhism in it, we will have to be very careful about clicking the mouse at the end of a turn (really, during the in-between-turns phase) in the real game... an AI will eventually come our way and demand that we convert to Buddhism, and we will have to be certain to refuse said AI's request.


In a test saved game, the Apostolic Palace was built on Turn 123, 200 AD. On Turn 134, 475 AD, the Religious Victory vote came up. All AIs were still running Buddhism as their State Religion. It was basically the AI Team of Two versus another AI in the vote. We would have the option of gifting-away all of our Buddhist Cities to prevent that kind of a loss, but obviously doing so wouldn't be to our advantage, other than avoiding the possibility of an instant loss.

LowtherCastle said:

Okay, ZPV, the question is, should I continue now or not?

Do we need to test my idea of preventing the DoW? What's the easiest way to set up an AI to WHEOOHRN us? I suppose we could make ourselves a land target of Monty, but using mountain tiles for the common borders, so that we could set up a 3-tile window along our border and put units on the 3 tiles outside our borders and see if his SoD will DoW when it lands on an occupied tile.

Click to expand...

He will DoW. I've just tested it.

I really can't figure out where the latest test saved game is located.

I see one for T64, 1440 BC:
http://forums.civfanatics.com/showthread.php?p=11339389&postcount=1459

I then see some other saved game files for Turn 75, 1000 BC.

Do we simply not have an up-to-date test game? If so, that's fine--no need to rush out and make one on my account.

Dhoom: I haven't updated the test save from T64 yet.

ZPV said:

He will DoW. I've just tested it.

Click to expand...

Okay. So it goes.

I'll continue onward then and we'll see what happens. I'm assuming we'll want to go ahead and have a few axes at the chokepoint to repel his SoD. By my calcs, we'll be able to get IW in trade, so that's actually a nice solution and rather well timed, assuming he researches it for us. We won't be in any hurry to DoP most likely. If not for IW, we can DoP for Sailing or something, after we bulb Engineering.