SGOTM 11 - Shaka, not stirred

[Abegweit]

I would like the movement 3 and the blitz. And we could build a heroic epic to encourage RNG to do some compensation for what it did to us last game. I never had a horses' army however
I don't have any leaders that early...

The problem is that the unit is too weak. It can't take down cities by itself and it is vulnerable to attack. Consequently it has to stay with your main army, which makes the 3 movement useless. The blitz isn't wonderful either as it may well die on the second attack and almost certainly will on the counter. Sword armies mix well with horsies. They make out the strongest defender and the others mop up. MDI, with their 6 attack, are even better. Horses only have an attack of 3 and swords are 4. I occasionally build archer or horse armies when I don't have iron but they can lose 3/4 of their HP attacking a pike. A couple of trebs is nearly as good.

Unless we have a town like Entremont in SGotm10 - the town itself will be productive but even the closest neighbours are completely corrupt - right?

No! This is precisely the point. We could have a dozen productive towns around it. Rank from the capital. Distance from the FP. The key is to keep rank down.

 

[Abegweit]

Seriously, my intuition tells me Delhi too far away for significant corruption bonus on towns surrouding.

Your intuition is wrong (mine told me the same thing until I studied the issue). The distance makes no difference. The key is to keep rank down, and towns with rank below 30 or so are decently productive.

 

[Paul#42]

No! This is precisely the point. We could have a dozen productive towns around it. Rank from the capital. Distance from the FP. The key is to keep rank down.

Okay, if we do not found any more than ~3-6 towns on our (part of the) continent, that would work.

I would like to have a screeny from Delhi and [Egypt Capital ] soon. Maybe Ivan establishes the embassies soon, posts some screens? That would fuel our discussion and fantasy...

 

[Andronicus]

Your intuition is wrong (mine told me the same thing until I studied the issue). The distance makes no difference. The key is to keep rank down, and towns with rank below 30 or so are decently productive.

Which is what I indicated later in my post (although I said might instead of would ), though not sure that up to rank 30 will be productive

 

[Paul#42]

No matter how nice a FP in Delhi would be - we don't want to rely on a MGL and will start a prebuild in I soon, won't we?

 

[Abegweit]

I would like to have a screeny from Delhi and [Egypt Capital ] soon. Maybe Ivan establishes the embassies soon, posts some screens? That would fuel our discussion and fantasy...

I'd like India and Turkey, since they are the ones we have leverage with. Thebes would give us an more info about which bloc to dow on.

Which is what I indicated later in my post, though not sure that up to rank 30 will be productive

In re-reading, I see you did. And, yes, 30 is probably a bit high. I propose 11 towns in this area (one more to the southeast, southwest and west) plus whatever would fit reasonably around the FP.

All this is predicated on getting a leader, of course. Hand-building that far away would be very difficult, I think. Plugging reasonable numbers into Alexman's formula indicates that a prospective FP town would be more than 60% corrupt even with a court. Only 10% afterwards though.

 

[Abegweit]

No matter how nice a FP in Delhi would be - we don't want to rely on a MGL and will start a prebuild in I soon, won't we?

As I study this, I come to think that a local FP would be pretty close to useless. Two hundred shields to reduce the corruption in one city by 20% is crazy. The distance effect is negligble, since everything is close to the capital. So the only effect is to increase the number of cities where the corruption is 60-80% instead of 90%.

 

[Paul#42]

As I study this, I come to think that a local FP would be pretty close to useless. Two hundred shields to reduce the corruption in one city by 20% is crazy. The distance effect is negligble, since everything is close to the capital. So the only effect is to increase the number of cities where the corruption is 60-80% instead of 90%.

How about building the FP close to our capital (best in U, but also apropriate in I) and moving our capital to an AI capital site? Would that let our old core productive until we have established our new core with 10+ towns? This would decrease flip probability and shorten our supply lines. Given a nicer area somewhere else which I don't doubt.

 

[Abegweit]

Real examples of corruption.

Here are the formulas. I have removed factors which don't apply to our case (communal corruption, commercial trait, etc).

Nopt = 20 * (.8 + 3/8*Nwe + 0.25*Ni )

Cd = (0.5^Ni * Gd) / 50
Cr = R / (2 * Nopt)
C = Min( Cd + Cr, .9 )

Where:
Gd = distance from nearest corruption-reducing building
Nwe = 1 if FP, 0 otherwise
Ni = 1 if court, 0 otherwise
Cd = Corruption from distance
Cr = Corruption from rank
C = Total corruption

With FP, no court, Nopt = 20 * (.8 + .375 ) = 23
W/O FP, no court, Nopt = 20 *.8 = 16
With FP, court, Nopt = 20 * (.8 + .375 + .25 ) = 28
W/O FP, court, Nopt = 20 *(.8 + .25 ) = 21

Assume a city with a rank of 16 and a distance of 5 from the FP and 25 from the capital. Then:

No FP
Cd = 25 / 50
Cr = 16 / 32
C = 90% (cap)

Remote FP
Cd = 25 / 50
Cr = 16 / 46
C = 85%

Local FP
Cd = 5 /50
Cr = 16 / 46
C = 45%

No FP, court
Cd = 12 / 50
Cr = 16 / 21
C = 90% (cap)

Remote FP, court
Cd = 12 / 50
Cr = 16 / 56
C = 41%

Local FP, court
Cd = 2 /50
Cr = 16 / 56
C = 32%

As can be seen, a remote FP has almost no effect on corruption unless you add a court, which would have to be cash-rushed.

 

[Abegweit]

How about building the FP close to our capital (best in U, but also apropriate in I) and moving our capital to an AI capital site? Would that let our old core productive until we have established our new core with 10+ towns? This would decrease flip probability and shorten our supply lines. Given a nicer area somewhere else which I don't doubt.

Not a bad idea at all. Plug the numbers into the formula above. T'would appear that our old core would be 40-50% corrupt.

 

[Andronicus]

Not a bad idea at all. Plug the numbers into the formula above. T'would appear that our old core would be 40-50% corrupt.

Am I reading this correct to suggest courts first would be beneficial to ensure corruption didnt drop too low in old core?

Of course this all assumes that ellusive leader

 

[Paul#42]

Not a bad idea at all. Plug the numbers into the formula above. T'would appear that our old core would be 40-50% corrupt.

~30% if we build courts in time before moving the palace...
Ever built a court in your capital?
(I remember a Sotd of an AI doing this)

Of course this all assumes that ellusive leader

Yes, indeed. He would have to wait some turns before rushing the Palace to let old core towns do at least some work on courts...

EDIT: After all it is the same leader we would use to rush a distant FP - but this time we do not rely on him to show up by building a local FP and then waiting.
If we do not get a leader again we can still abandon our capital and make a palace jump if the new area is really great. This leaves us more options than just waiting for a MGL to show up...

 

[Abegweit]

Am I reading this correct to suggest courts first would be beneficial to ensure corruption didnt drop too low in old core?

Of course this all assumes that ellusive leader

Doesn't require a leader to abandon a palace, vanilla-style. Courts would, of course, be beneficial but I don't think that spending 80 shields to get 1-2 spt gives a very good ROI. Especially in a short game, which is what this one is supposed to be.

In view of all this, should we start an FP in the core and decide what to do when/if we get a leader.

 

[Paul#42]

Remote FP
Cd = 25 / 50
Cr = 16 / 46
C = 85%

Remote FP, court
Cd = 12 / 50
Cr = 16 / 56
C = 41%

Doesn't require a leader to abandon a palace, vanilla-style. Courts would, of course, be beneficial but I don't think that spending 80 shields to get 1-2 spt gives a very good ROI. Especially in a short game, which is what this one is supposed to be.

In view of all this, should we start an FP in the core and decide what to do when/if we get a leader.

If your great analysis is correct, it would be quite a difference, with court we would get four times as many shields in our "example town" (59% vs. 15% non corrupt spt). That's quite a difference for towns that are used to having ~15-20% corruption only. It should definitely be more then 1-2 spt, rather 8-10 spt at size 12 or 4-6 spt at size 6.
In Hlobane (46% corrupt right now) for example the courthouse could benefit even before moving the Palace.
Just in case we got too much money and don't know what to rush...

 

[Abegweit]

If we build an FP in the core and then jump the capital, the FP is local to that core and thus the gain is from 45% to 32% or about a shield pre-rails.

Edit: According to CAII, Hlobane will be only 38% corrupt once we switch govts. An FP in Isa would reduce that to 23%

 

[Paul#42]

If we build an FP in the core and then jump the capital, the FP is local to that core and thus the gain is from 45% to 32% or about a shield pre-rails.

local? I understand local as being close to the capital and remote as being away from it...

So you mean, if Isl builds FP, Palace jumps to Delhi the FP is local to Delhi?

 

[Abegweit]

local? I understand local as being close to the capital and remote as being away from it...

So you mean, if Isl builds FP, Palace jumps to Delhi the FP is local to Delhi?

No. I meant local to the city for which we are calculating corruption. If the FP is built in Isa, then it is local to Hlobane and remote to Delhi. If the FP is built in Delhi, then it is local to Edo.

 

[Paul#42]

That's the way I understood it.
So what's wrong with this example: if I look at Ulundi, 5 tiles away from FP-city Isa and 25 tiles away from capital Delhi (in fact 18, just take 25 for ease), it will have 85% corruption without and 41% corruption with courthouse.

Where's my mistake?

 

[Abegweit]

Local FP
Cd = 5 /50
Cr = 16 / 46
C = 45%

Local FP, court
Cd = 2 /50
Cr = 16 / 56
C = 32%

Edit: the number 5 in the first example is the distance to the nearest corruption-reducing building. The number 2 in the second is half that distance (not sure if you are supposed to round down but that is typical Firaxis style and it doesn't make much difference anyway).

 

[Paul#42]

Local FP
Cd = 5 /50
Cr = 16 / 46
C = 45%

Local FP, court
Cd = 2 /50
Cr = 16 / 56
C = 32%

Ah, now I see...
Sorry, I'm not focused on this, my work keeps distracting me from the essential stuff