SGOTM 15 - Kakumeika

[bcool]

bcool, I'm fine with using our usual turn set protocol, especially since you have committed to update our test save to the end of your turn set. My secondary concern is my turn set is taking too long and we may run out of time in our end-game.

Let's try to hammer out at least t72-t73 so I can play those turns Monday evening. Maybe t74-t79 can be played Tuesday or Wednesday evening, if necessary.

Sun Tzu Wu

I understand about the length our recent turnsets have been, but I think these turnsets were critical to get right. And you guys have done a great job getting it right.

I expect (hope) that the later turnsets will go faster. Once we get the AP and caravels I think the turnsets will go faster.

we still have until May 21st. So we have 2 months. So even if we take 1 week per turnset that is 8 more turnsets. Taking us to ~T150. That is about when we expect to win.

I think we are okay timewise. I think in the future we do our best at 3-4 day turnsets and that will give us wiggle room if the game gets more complex.

 

[Sun Tzu Wu]

I am astounded at the amazing results you guys have achieved with so much testing

Most of the credit goes to DanF5771 for detecting a while back that Tokugawa was in War prep mode and also for pointing out the viability of a Scout gambit move (which ensured that we reached Cultural Border Radius 3 and Def 40 before the attack). bcool also deserves credit for preparing some Axeman builds and a couple Worker chops to speed up additional Axeman builds and earning 160W to upgrade two Warriors to Axemen.

DanF5771 also deserves credit for the test save that was essential for testing the military situation. I was able to modify it (adding an Axeman in Delhi, and a Scout plus subsequent 3 Axeman builds) to turn a potential disaster for us into a potential disaster for Tokugawa.

Sun Tzu Wu

 

[bcool]

T72 T73 and beyond suggestions

I think Delhi should do something like...

T72 Forge (Corn, FP, GrassFarm, Copper, Gmine)
T73 barracks (same tiles)
T74 3 pop whip forge for max OF
T75 (use OF for axe) (could use OF for barracks if threat level is low)
T76 axe build partial
T77 2 pop whip barracks
T78 axe with overflow

DeerGold
T72 DeerGold should whip its granary this turn (works gold and deer after whip)
T73 OF from granary added to current axe build (gold and deer after whip)
???

Since PigsGems will be popping its borders next turn I think we should get Hoover and Karl (the workers 1N of Delhi) over there asap

So I suggest Karl and Hoover move 1W and put a road 1N1W of Delhi (T72)
This sets them up to move to the forest south of PigsGems and prechop T73 together
Then T74 they can move to the pigs and start the pasture (T74)

The forest in PigsGems can be chopped to finish the axe (so I think PigsGems should work the axe until the forest chop would finish it) I would rather chop the forest than stop working either the pigs or the gems by 1 pop whipping the axe.

Fritz can build continue the road by moving 1S1E and roading that hill. With the roads I suggest with Karl and Hoover that links Delhi and PigsGems and starts a decent path from both cities up to DeerGold for possible reinforcements.

With the prechop in PigsGems I think we can afford to delay the road up to DeerGold by a few turns.

Eiffel (the worker 3W of SheepTown)
Instead of mining near CowTown, I think Eiffel should prechop the forest 1S1E of Delhi. To get more axes out if we need them asap up north.

The mine STW suggests for Eiffel won't be used for some time to come since CowTown would have to grow back to 3 to use it.

Research question
Should we research archery? archers are significantly better against bowman than axes and they are cheaper and faster to build.

 

[Sun Tzu Wu]

Here is the test game updated to T70. I tried to simulate the war but failed (I'm not very familiar with AI unit scripts )

The hammers and food should be accurate (as well as unit placement and fortification) (we might have to give ourselves a bit more gold and research is off by a bit)

I uploaded the end of T68 and end of T67 as well if someone wants to try and simulate the war better.

A big thank you for updating these so quickly!

I'd like to suggest the prefix TEST-GAME- for all test games that more or less accurately model the real game so that they can be easily distinguished from special purpose test games that model at most one very specific aspect of the real game. The former type of test games are especially useful for preparing 10t and even longer PPP whereas the latter are not quite so useful in this regard in a generic sense (they are after all special purpose test games).

Sun Tzu Wu

 

[Sun Tzu Wu]

I think Delhi should do something like...

T72 Forge (Corn, FP, GrassFarm, Copper, Gmine)
T73 barracks (same tiles)
T74 3 pop whip forge for max OF
T75 (use OF for axe) (could use OF for barracks if threat level is low)
T76 axe build partial
T77 2 pop whip barracks
T78 axe with overflow

Great idea! For t75, I'd say overflow into Barracks as long as we can still 2P whip it. On second thought, we already have 2 unhappiness for whipping and 1 unhappiness for no garrison. CowTown is building the garrison for Delhi. I actually don't like whipping Delhi again and especially not twice. That would limit its working Population to 4.

DeerGold
T72 DeerGold should whip its granary this turn (works gold and deer after whip)
T73 OF from granary added to current axe build (gold and deer after whip)
???

Since PigsGems will be popping its borders next turn I think we should get Hoover and Karl (the workers 1N of Delhi) over there asap

Agreed.

<several suggestions for ... chopping ... Axemen.>
<suggestions regarding Road from Delhi to Deergold City.>

I'd prefer that we complete the chops after we have Mathematics. Before Mathematics, I'd prefer whipping Axemen/Archers, since the only penalty is lost of Population, and 1 Happiness for 10t. The -50% yield on pre-Mathematics chops really hurts. We also need some Forests to build The Apostolic Palace too, right?

I'm fine with pre-chopping any Forests when there's nothing else better to do.

Research question
Should we research archery? archers are significantly better against bowman than axes and they are cheaper and faster to build.

Will Hummuragawa even use a significant # of Bowmen in his SoD? Even so, I wonder how effective they can be attacking fortified Axemen. The odds should still strongly favor the Axeman, even if it is the Axeman that attacks a Bowman on a non-forested, non-river, flat plot. I suspect that the AI uses Bowmen primarily for defensive roles anyway.

We should defend against Hammuragawa's SoD 1-SW of the sign "choke point" on the Forest there for +50% defense plus +25% fortification. Note that this Forest plot is also a choke point that Hammuragawa's SoD can't bypass without using Galleys. Of course Archers would defend better in Deergold City when reaches Cultural Radius 3 = Def 40. However, I don't think Archery is a worthwhile diversion; Axemen will defend well on a Forest, especially when fully fortified.

Sun Tzu Wu

 

[bcool]

I didn't suggest chopping forests. I was just pointing out that they could be chopped if we need to (prechopping so they are ready to go immediately). I would rather chop a forest before mathmatics than whip away a good tile like pigs or gems.

Not working Pigs loses us 3F1H per turn. not working gems loses 1H7C? per turn. Whipping down to 1 pop without a granary loses one of those tiles for about 4 turns at least.
Chopping before math loses 10 raw hammers. I would rather chop than whip down to 1 pop.

 

[Manco Capac]

We should defend against Hammuragawa's SoD 1-SW of the sign "choke point" on the Forest there for +50% defense plus +25% fortification. Note that this Forest plot is also a choke point that Hammuragawa's SoD can't bypass without using Galleys. Of course Archers would defend better in Deergold City when reaches Cultural Radius 3 = Def 40. However, I don't think Archery is a worthwhile diversion; Axemen will defend well on a Forest, especially when fully fortified.
Sun Tzu Wu

Sounds like a sound idea to take that chokepoint to defend as its defense is superior in short term than the city. Unless we invest in a wall (which is rarely attractive), that 50% is superior to even 40% second border pop. The downside: less time for fortification as the non-upgraded warrior axe has two turns of moving before fortification process and if the healing process will take a little time more but that is the weak side of downside points because we don't expect many successive SoD's.
Also, being nearer to the coming SoD, less time to prepare ourselves. In regards to being a chokepoint, yes it is but not in the same sense as roaming barbs (LEMMING?), the SoD WILL pass through the chokepoint, enemy guarding or not. Thus we shall not expect any galleys used instead of a landish SoD.

I favor that forest as defensive point in short term. If there was a bent river, then it would be a no brainer. Yes a bent river is near the capital along forests, but my guesses the SoD will pass by the defenders outside the city.

 

[Manco Capac]

Research question
Should we research archery? archers are significantly better against bowman than axes and they are cheaper and faster to build.

I vote no. Already the wars slowed us down and going for a dead end tech like this (unless going for a quick xbowmen war).

Looking the odds (I simulated the fortification with cover promo for the AI axe) and they are really good. Thus I strictly refuse teching Archery.

First case where a bowman with first strike level three attacking an unpromoted axe with max defense (forest+fortification): Really good odds.
Second case where the same bowman attacks an axeman without fortification. Still good to me.
Third case and worst possible with a bowman with C2 promo attacking an axe w/o fortification. 17,5% percent of winning is still low. So, archers aren't that necessary and the AI won't produce lots of them in a stack where melee units are accessible and main duty is to conquer.

BTW, I think cover promo is working the same as 25% fortification; I know combat mechanics are not that simple to calculate. For example C1 is not worth as diminishing of 20% from a CG1 archer. It is not accurate on how it works.

 

[Manco Capac]

I understand the point of view against retaliating against Toku without a tech edge as inefficient, but let's take the situation we are ready to make peace and perhaps to try to sneak a small company of 5 axes towards a city with 2 protective archers. If the next turn, Toku hasn't whipped something intelligent or wasn't capable to do so, we attack, snatch that city, then make peace. Losing at most 2 axes aren't worth a city of 100 (settler) or even with a free granary which worths 60 ? I think spending 70 for that is fair.

 

[Sun Tzu Wu]

Recently the test game has 1 Wpt more city maintenance (13 Wpt) than the real game (12 Wpt). I added 6W by working WB'ed 2F1H7C instead of 2F1H in Delhi for 1t and F51H2C (forested sheep pasture) instead of 5F1H in Deergold City to be just short of correcting it. I may make the latter change permanent to keep the net wealth rate in sync at 29 Wpt on t70.

The test game also has a few beakers more research towards Writing, but that's not related to the above issue and can be independently corrected later.

The test game omits Tukugawa's stack and war, since I don't know how to set such things up in World Builder.

I plan to update this test save to t71 and t72 and then continue work on the PPP for my final 8t.

Sun Tzu Wu

 

[mabraham]

Sounds like a sound idea to take that chokepoint to defend as its defense is superior in short term than the city. Unless we invest in a wall (which is rarely attractive), that 50% is superior to even 40% second border pop. The downside: less time for fortification as the non-upgraded warrior axe has two turns of moving before fortification process and if the healing process will take a little time more but that is the weak side of downside points because we don't expect many successive SoD's.
Also, being nearer to the coming SoD, less time to prepare ourselves. In regards to being a chokepoint, yes it is but not in the same sense as roaming barbs (LEMMING?), the SoD WILL pass through the chokepoint, enemy guarding or not. Thus we shall not expect any galleys used instead of a landish SoD.

I favor that forest as defensive point in short term. If there was a bent river, then it would be a no brainer. Yes a bent river is near the capital along forests, but my guesses the SoD will pass by the defenders outside the city.

Once we have Maths we will have the option of putting a fort on that forest for the extra defence bonus. Now Hammu's melee units can use city-attack upgrades, and we can use city-defence upgrades if we have such units. I think we also get city healing bonuses?

However I'm not so sure we'll be able to afford to maintain such a forward position. We have only a few tiles of vision, so we need to have our army already in position to start to fortify. Reinforcements will have more time to travel a shorter distance to defend in DeerGold. Assuming we haven't got peace with Toku by the time we have to defend Hammu, surely it is more efficient to have the bulk of our force central so that we can react to either front than to maintain a split force at either front. You need more units to do the latter, even though you will get better benefit from fortification. I'd like to get away with as small an army as needed for defence, so that we can pursue our VC goals with as much of our resources as possible.

 

[mabraham]

I vote no. Already the wars slowed us down and going for a dead end tech like this (unless going for a quick xbowmen war).

Looking the odds (I simulated the fortification with cover promo for the AI axe) and they are really good. Thus I strictly refuse teching Archery.

First case where a bowman with first strike level three attacking an unpromoted axe with max defense (forest+fortification): Really good odds.
Second case where the same bowman attacks an axeman without fortification. Still good to me.
Third case and worst possible with a bowman with C2 promo attacking an axe w/o fortification. 17,5% percent of winning is still low. So, archers aren't that necessary and the AI won't produce lots of them in a stack where melee units are accessible and main duty is to conquer.

BTW, I think cover promo is working the same as 25% fortification; I know combat mechanics are not that simple to calculate. For example C1 is not worth as diminishing of 20% from a CG1 archer. It is not accurate on how it works.

I agree we don't want to plan to tech archery in the short term.

 

[Manco Capac]

Once we have Maths we will have the option of putting a fort on that forest for the extra defence bonus. Now Hammu's melee units can use city-attack upgrades, and we can use city-defence upgrades if we have such units. I think we also get city healing bonuses?

Brilliant! (No sarcasm intended, but a to you to point it out.) I haven't thought of it! In addition to a superior forest defense, CR promos are nullified letting axemen attackers less powerful. Remember that Toku attackers were promoted in goal to attack a city with CR1, not shock promo. I like even more the idea of fortifiying in the forest instead of the city as the SoD has to pass through the small company of defenders to get to the city. I remember DanF said popping barbs can't go through enemy units blocking areas, but normal AI's can and will make their path regardless the units blocking the path.

I agree we don't want to plan to tech archery in the short term.

Which means we won't have defensive promos for awhile exclusively for cities. Anyways, forts are quite away to my opinion and the war can come short way before than expected.

 

[mabraham]

I understand the point of view against retaliating against Toku without a tech edge as inefficient, but let's take the situation we are ready to make peace and perhaps to try to sneak a small company of 5 axes towards a city with 2 protective archers. If the next turn, Toku hasn't whipped something intelligent or wasn't capable to do so, we attack, snatch that city, then make peace. Losing at most 2 axes aren't worth a city of 100 (settler) or even with a free granary which worths 60 ? I think spending 70 for that is fair.

Spending 2-3 axes that will be fairly useless when we are at peace in order to capture a city is worthwhile, if we want to pay for having that city as well as the 6-7 mainlaind + possible island cities we already want. Unless we learn Osaka has iron, I don't want to keep it because it doesn't have a food source. Sheeptown is not good enough with only one of the food sources because it has to work the gold. I could see taking Osaka to speed up our culture capture of the corn, perhaps spreading Hinduism there and gifting it back at some point. Toku's third has no food sources that we know about, so any that exist will be under Kyoto's culture influence, so I can't see us wanting that.

Basically I'd rather judge our army size with a fairly slim margin of safety and put those hammers into expanding where it will be useful for us.

Speaking of culture-capturing the corn, we should be able to estimate when our monument plus oracle will overpower the library. We may want to invest in some more culture in Sheeptown too - which is another reason I like it building its granary sooner rather than later. The argument that "it grows fine" is never a valid one in itself - only if the thing being built is more valuable than the granary. A granary is normally the best building you can build anywhere. So I am skeptical about the axe in PigGems, for example, as it was not going to arrive in time to defend any front except Hammu's. It would have helped defend Delhi if Sheeptown fell, but I think the SGOTM is lost if that would occur, so don't care about the axe for that purpose.

 

[Manco Capac]

Spending 2-3 axes that will be fairly useless when we are at peace in order to capture a city is worthwhile, if we want to pay for having that city as well as the 6-7 mainlaind + possible island cities we already want.

Indeed, increased maintenance crossed my mind, but I had the idea of using present axes to attack once we are sure we can make peace with him. Of course, there is the annoying possible bad luck where we lose enough to bing Toku back to his "refuse to talk". Yes, it let SheepTown with fewer defenders, but if ever he brought another SoD, then we bring the war to an end at once.

Giving back the city isn't a bad idea for a diplo modifier, but Toku was our main enemy at first and Hamuragawa has to like him less him in long term. And making peace with Toku as fast as possible with him is crucial because "shared war modifiers" are starting to accumulate (+1 by 10 turns IIRC). Already friendly each other, we don't want a situation impossible to get Hammuragawa to like us more than his twin brother.

Okay, I divert, my main idea was to get a quick city before war end, not pushing far.

 

[Sun Tzu Wu]

The TEST-GAME now matches Treasury of 29W and 18 Wpt in the Real Game at the beginning of turn 71. Maintenance values now also match; I must compared the two earlier with one before a whip and the other after a whip though I was sure all cities had the same Populations and they must all have the same relative positions (they were WB'ed that way).

Sun Tzu Wu

 

[Sun Tzu Wu]

This is the current test save. A Forest that grew a few turns ago in the test save has finally been removed. The Scout was deleted to simulate the combat of Tokugawa's Stack attacking it. Hinduism was added to PigsGems (Bombay was renamed to PigsGems) to simulate natural spread that just occurred in the real game.

Tokugawa's six unit stack is now 1-SW of SheepTown's PH Gold Mine. The test game does not show this.

I won't make progress in the PPP yet tonight, but the test game is up to date to t72.

Any suggestions for the PPP are welcome.

Sun Tzu Wu

 

[bcool]

I updated the test game a little bit.

I fixed the Hindu holy city (which gets removed if you remove hinduism from a city in world builder--it is a known bug of world builder)

I put us at war with Mansa and Isabella (our neighbors in the game unfortunately Isabella is in the team in the test game)

I added units in our culture to represent Toku's stack, gave them the appropriate promos and gave them the script (attack city lemming)

they will move to the gold and attack the city with this script. It is impossible to simulate the real AI actions I believe but hopefully this does a decent approximation.

 

[WastinTime]

In my experience, the fort will allow CR promotions to work, not nullify them.

 

[DanF5771]

Very nice! Well done Sun Tzu Wu!

Re war and retaliation:
I quite agree with mabraham, taking Osaka is probably too costly now (has Barracks and can whip defenders quickly).
Although it would be nice for earlier capture of the Corn and/or the permanent diplo points for liberating it back to Toku (attention: our plot culture in inner ring gets cleared then).

A soonish Trireme from PigGems could scout Hammu's SoD and wreck havoc in his fishing grounds, distract him to build Galleys (perhaps kill a Galley for enough war success to end the war early?) and move further across the equator to meet northern AIs (if not also blocked by peaks). Not sure how quickly he can build up an invasion force.