SGOTM 17 - The Shawshank Redemption

[Jastrow]

The graph looks to me like 2 * 4K (successive turns), and not 1 times 8K.

-Upgrading archers to LB is an extra 3K, so I dont see that working...
-Spearman or Chariots are 4K each... Possible?

Or it could be a combination of small things...

 

[Mitchum]

If we're in agreement that we need to take out Mansa soonish (i.e. now or in 20 turns), then we're going to suffer the crippled economy effect that you mentioned, either now or later, but it will happen at some point, right? If we do it now, we'll need fewer units and the crippling affect will be a bit smaller. Either way we're going to have to suffer through it so let's just get on with it...

As we've both agreed, you'll likely never be in agreement with Dhoom and me. As it stands, we're at 2 vs. 1. Rather than debate this back and forth forever, I suggest that we move on with the HA rush plan which has the majority vote. This will allow us to focus on how to get IT done rather than debating what IT is, something that we'll never agree on. Focusing our discussion on how to get IT done is much more productive at this point as I think we've reached the end of an unwinnable debate (i.e. we'll never reach consensus). Just like when Dhoom and I lost the Monarchy vs. Math debate (which probably dragged on for too long no thanks to Dhoom and me), we had to bite the bullet on Monarchy and focus on making the Math approach work as well as we could. I hope that you can do the same on this HA thing even though it's not your favored approach.

Of course, if another team member jumps in on your side, then I think the discussion warrants a bit more time where we can lay out the pros and cons of both approaches as succinctly as possible and take a final vote. At this point though, it appears to be just the three of us and I hope that we can move forward together on the majority plan and not lose anyone along the way.

Worst case, we fail miserable with the HA rush and you can feel validated without having to say "I told you so!" Best case, the HA rush is a massive success and we've all learned something. If this issue really has you buggered, you could always come back to this point in our game once the competition is over and try a different approach to compare the differences. In the end, the fun in this game for me is working as a team and learning and I'd be more than happy to come back to this point after the game and explore alternate paths with anyone who is interested. Just like WOTM-50 and our space games, I enjoy comparing strategies, learning, trying new things and becoming better in the process. Debating things when we've agreed to disagree is counter productive.

 

[Jastrow]

If we're in agreement that we need to take out Mansa soonish (i.e. now or in 20 turns), then we're going to suffer the crippled economy effect that you mentioned, either now or later, but it will happen at some point, right?

Then I vote for waiting until much later when we can afford it better (i.e. after COL)

If we do it now, we'll need fewer units and the crippling affect will be a bit smaller.

You know better than that... a beaker now is NOT equal to a beaker later.

As we've both agreed, you'll likely never be in agreement with Dhoom and me. As it stands, we're at 2 vs. 1. Rather than debate this back and forth forever, I suggest that we move on with the HA rush plan which has the majority vote.

I said that about 20 posts ago...

This will allow us to focus on how to get IT done rather than debating what IT is, something that we'll never agree on.

Fine, but I cannot focus on it...You are asking me which horrible option I think is less horrible. I have no reasonable way to evaluate between them. Do whatever you think is best, without worrying about my vote on any details.

Focusing our discussion on how to get IT done is much more productive at this point as I think we've reached the end of an unwinnable debate (i.e. we'll never reach consensus). Just like when Dhoom and I lost the Monarchy vs. Math debate (which probably dragged on for too long no thanks to Dhoom and me), we had to bite the bullet on Monarchy and focus on making the Math approach work as well as we could. I hope that you can do the same on this HA thing even though it's not your favored approach.

Actually, I cant... I dont want this miss-understood. It is not that I dont want to... I am fine going on with the majority vote, and want to contribute as best as I can, but the fact is I have little to offer to the discussion, and this for two reasons.

1- Simple fact... I am by far the weakest war-monger of the 3 of us.

2- I remain of the opinion that this war is crippling our research, which is fundamentally a bad idea. On all detail, my instinct will therefore to be to act in a way to minimize this impact... Since this is counter to both your opinions, pretty much all details are also going to get voted on 2-1 against me (which may well be correct, given "1" above), so much discussion from my side is not likely to help move things along any faster... I will of course comment when I think I have something useful to contribute, but I dont expect that to be very often.

Of course, if another team member jumps in on your side, then I think the discussion warrants a bit more time where we can lay out the pros and cons of both approaches as succinctly as possible and take a final vote. At this point though, it appears to be just the three of us and I hope that we can move forward together on the majority plan and not lose anyone along the way.

We certainly should not be waiting any longer... We need to move on as soon as someone is ready to play.

Worst case, we fail miserable with the HA rush and you can feel validated without having to say "I told you so!"

I hope I am wrong, but yeah, that is the result I expect...

Best case, the HA rush is a massive success and we've all learned something.

This is not about HA rush itself being a success... It is about the research blow we are taking.

If this issue really has you buggered, you could always come back to this point in our game once the competition is over and try a different approach to compare the differences. In the end, the fun in this game for me is working as a team and learning and I'd be more than happy to come back to this point after the game and explore alternate paths with anyone who is interested.

I may to that if I have the time.....

Just like WOTM-50 and our space games, I enjoy comparing strategies, learning, trying new things and becoming better in the process. Debating things when we've agreed to disagree is counter productive.

Yes, I said this 48 hours ago (post 1356)...

 

[Dhoomstriker]

@Mitchum
So, your suggestion is to first hit Djenne and then hit Timbuktu with our surviving forces, rather than going straight for Timbuktu right away at the same time as going after Djenne?

If so, we should get some testing in (I'll be unable to access Civ 4 for roughly 56 hours... two days + 8 hours) to see how emergency AI whipping works. I saw Monte whip 1 Jaguar (which then had 0 Fortification bonus) when attacking Timbuktu on Turn 1, but I haven't tried to see what would happen when attacking on roughly Turn 3, as you are proposing.

One minor concern is that it will be harder to trap Workers by not going for both Cities at once, since the idea was to catch Workers out in the open on the Road network between Djenne and Timbuktu when Timbuktu's Cultural Borders fall... but, there won't be Workers in that area if we first take down Djenne and then head from Djenne toward Timbuktu.

So, we could "save" 120 Hammers by building 3 less Horse Archers and thereby end up capturing 120 Hammers less of 2 Workers that weren't trapped by our massive-Cultural-Border-removing-all-at-once strike. Is that really a savings when we'd much rather have those Workers right around this time in the game?

Now, that's an assumption that we'll be able to capture the Workers thus, but it's a pretty fair one, since the vast majority of Mansa's Cultural Borders will disappear by taking down those 2 Cities.


Yes, I think that 3 Horse Archers per defender plus 1 extra is a good amount of units to rely on. How much would it suck to go in with 5 Horse Archers, lose 2, have 2 retreat, and there still be 2 City Defenders when you have your 5th and final Horse Archer ready to attack? That would suck a lot... so, you would kill 1 of the defenders and have 3 wounded Horse Archers. On the next turn, you would face 1 more whipped defender plus 1 healed-via-promotion defender... and you'd feel compelled to attack with your wounded Horse Archers on the next round... Civ 4's random number generator could make it a bloody mess... and even if you retreat, you're down 2 Horse Archers, Mansa will have a promoted unit, and he'll be able whip even more defenders.


Before declaring war, we can have a Horse Archer in place next to each of the two target Cities, each of which can move 1 movement point along a Road back toward the stack (joining the stack at Djenne and getting almost certainly into the right location for teleporting for Timbuktu) without losing the ability to move 2 full squares on that same turn. That way, we won't be going in blindly and, for example, if Mansa has 3 defenders at Djenne, we can alter the plan, as required.



"Surgical" happens if we attack Djenne on Turn 0 and Timbuktu on Turn 1... Mansa won't have time to reinforce Timbuktu with defenders from other Cities and only Chariots that were in range can run into his capital. He can be counted on to whip 1 more unfortified defender.

I don't see it being all that surgical if we have to wait a few turns to go after Timbuktu, as we'll likely see 5 defenders there by waiting that long.


Sure, gifting Dual Banana City is a good option and may actually be cheaper for us in the short term. Hopefully, Mansa won't attempt to Chop the Forests there before we're ready to declare war on him.

Again, we have to be careful with the timing, as, when an AI gets more Cities, it starts to build more Military Units as its Unit Supply numbers go up. So, it might make sense to build Wealth in Dual Banana City and gift it at around 3 turns from when we're ready to declare war, just to give a bit of time for units to walk toward there... or, it might be better not to gift it at all if Mansa has other Cities settled by then where his Workers will already be busy on tasks and thus won't be lured to come toward Dual Banana City.


Prague is definitely more centralized in terms of a Palace location for Distance Maintenance, which also needs to be a consideration when chosing where to place your Palace. It could be an illusion to get +10 raw Commerce (even if it gets increased by Bureaucracy, Oxford, and an Academy) at the cost of, say, -25 City Maintenance. Yes, Rathauses give us more flexibility, but I wouldn't rule out Prague as a targeted location for our super science City. It also has the settled Great Scientist going for it.

 

[Dhoomstriker]

Another point about Dual Banana City... lets say that it is one way or even the only way to make a southern Cultural Bridge. If that City becomes Mansa's capital, he'll meet the AI to the south, which could seriously mess up tech-trading options.

So, it might be best to only gift Dual Banana City if it gets settled on top of the G Riv Jungle square, where a Cultural Bridge would not be possible.

Note that until Dual Banana City regrew to Size 2, we wouldn't be able to recapture it without auto-razing it (worth testing for gifted Cities, but it's probably true even for gifted Cities). That fact could mean a delay of taking Peace for techs or else leaving that City in Mansa's hands for a while.


So, right now, I'm leaning away from gifting that City unless we can show that we can recapture a gifted City that only ever reached Size 1 without it being auto-razed.

 

[Jastrow]

Another point about Dual Banana City... lets say that it is one way or even the only way to make a southern Cultural Bridge.

Again, Huh???

How in the world are we using that city for a culture bridge???

 

[Mitchum]

Here is the lay of the land around Mansa's empire.

Spoiler :

How do you propose that we attack Timbuktu on T1? We could stage HAs somewhere east of Timbuktu or somewhere just south. I'm opposed to these options at first glance for two reasons. First, Mansa will most likely settle one if not both of these locations in the next 3 to 5 turns. Second, it will take forever and a day to get our stacks into position through all of that jungle and they will be eating up supply cost the whole time. This will cost quite a bit of gold AND delay when we can declare due to the time it would take to get all of our units into position. one... tile... at... a... time!

My thinking is that we'd have our entire stack except for one or two scouting units sitting at Djenne-SW+SW on the pigs resource, safely within out cultural borders. On T0, we'd capture Djenne by moving one HA at a time. That way, if it takes 7 HAs to take the city, we have them there. If it only takes 5, those extra 2 that we had there for good measure are free to do other things. Once we've caputured the city, our un-moved HAs can use roads to march to Djenne-NW+NW on T0. On T1 we move to the pig resource at Timbuktu-SE and then capture Timbuktu on T2 (not T3 as I forgot that we could move quite far with our HAs once Djenne's culture is gone).

As Duckweed showed, the AI should not whip any units on T0 when our stack is at Djenne-NW+NW as he doesn't feel threatened. In fact, he may well move some units out of Djenne toward other cities so that we could easily pick them off in the open. So we really should only face +1 whipped, unpromoted, unfortified defender on T2 when we actually attack his capital.

I think this approach will have us declare on him sooner (our units will be getting into position on roads as opposed to through jugle) at a lower cost (little to no supply costs). It will also likely free up the two or three safty HAs you wanted in the Djenne stack should it only take 2 HAs per defender.

The thing with giving away Dual Banana ASAP is that it should almost immediately draw out 1 or 2 units from Timbuktu plus a worker or three. When we do DoW, if Mansa hasn't moved units out of Timbuktu yet, he may be prompted to do so to protect this city. Finally, if Mansa doesn't have spare workers to send to Dual Banana, maybe he'll build another 1 or 2 of them instead of building more units... I'm fine either way but I think this option does deserve some consideration.

 

[Jastrow]

Obviously, of the two war plans above, I prefer Mitchum's (because it has lower supply costs and needs a couple of units less, for lower unit costs as well.)

 

[Dhoomstriker]

Turn -2 Our Timbuktu stack has gathered on the G Pig that is 2SW of Djenne
Turn -1 Our Timbutu stack moves out of our Cultural Borders going 1NE G Copper, 1NW G Jungle (remember, we BUILT that Road--good forethought, haha), 1NW GH Rice Farm, 1W G Jungle
Turn 0 Declare War
Turn 0 Horse Archers march 1NW Plains, 1N P Cottage Road
Turn 1 Horse Archers move 1NW G Road OR 1N P Horse OR 1NE G Pig, then attack Timbuktu

We have these turns of Unit Supply costs:
Turn -1, Turn 0, and any survivors on Turn 1 who don't move into Timbuktu to guard the City (that's another reason why you don't want to skimp out on units--you don't want to defend a captured City with a wounded unit, as the unit that you "saved" by not building it will be equal to the unit that died in the counter-attack... i.e. by building +1 unit, you end up keeping +2 units... that's what is meant by having sufficient forces in Civ 4).

That's the same amount of Unit Supply costs as Mitchum's plan, but by attacking sooner, we remove the opportunity for Skirms from Kumbi Saleh and reduce the time for units from any of Mansa's roaming stacks (such as from part of a Settler Party that has just moved out since the suspected-to-be-already-built Settler finally got its Settler Party guards) from moving to reinforce Timbuktu before we attack it.


As long as Kumbi Saleh has not expanded its Cultural Borders again, being on the P Cottage Road that is 2S of Timbuktu gives us multiple chances for stealing Workers that were near Timbuktu or that were near Djenne but ended up being just out of reach of the Djenne Horse Archers on Turn 0.

If it only takes 5, those extra 2 that we had there for good measure are free to do other things.

Those units will still be equally free with the approach that I am suggesting. The difference is that you are hoping that they'll be available, while I'm planning on using any excess Horse Archers to ensure that we can capture a nearby Workers (i.e. we will be too busy to be marching on Timbuktu). Ideally, I'll have an Archer in our Djenne stack so that I can stuff it into Djenne on Turn 0.


As for a Cultural Bridge, any of Silver Bridge, Fish City, or Dual Banana being settled on the P Riv square have the potential to form a bridge of Culture to the south as long as there is a Coast square adjacent to one of the Ocean squares that our Culture will cover. We only need there to be one Coast square SOMEWHERE that is close enough, and given the size of the map, as I outlined earlier, you'd have to steal a lot of land from the southern AI to prevent there from being any sort of a Cultural Bridge.

The thing with giving away Dual Banana ASAP is that it should almost immediately draw out 1 or 2 units from Timbuktu plus a worker or three.

The Workers will first have to be finished a previous task and then have incentive to improve Dual Banana. Given that Mansa probably doesn't know Calendar, he'd likely only send 1 Worker to a Flood Plains square at best.

The cost is that with his increasing Unit Supply, he WILL build more Military Units, and it could be that the timing of completing said Military Unit could mean that it gets finished at the end of the turn before we're about to attack Timbuktu, giving Mansa yet another defender there.

if Mansa doesn't have spare workers to send to Dual Banana, maybe he'll build another 1 or 2 of them instead of building more units...

From my experience, AIs build City Defenders before they build Workers. Do you have evidence that shows otherwise?

When will we declare war on Mansa again to recapture Dual Banana City? Or, are you willing to delay getting Iron Working sooner by delaying Peace just to wait for this City to grow to Size 2 under Mansa's control?


I don't actually see Mitchum's plan requiring less units because we do not have a Super Medic. In fact, there is a great risk that we'll need more units to take down Timbuktu due to giving Mansa an extra turn to dump nearby units into Timbuktu. Further, we give Mansa 1 more turn of production, which may not seem like a lot, but AIs can often get 2 cheap units out of one whipping action because of their cheaper unit cost and 2 turns' worth of working a GH Riv Mine and a P Horse Pasture, in addition to the Hammers from the whip, any partial Hammers already dumped into a Military Unit that was in progress (say, because Mansa started to build a Skirm to send to Dual Banana City) and the cheaper cost of units for an AI.


Essentially, we're setting ourselves up for needing more Military Units or else being turned back at the gates of Timbuktu with either defeat or excess losses from having to attack with wounded Horse Archers on the second turn of fighting for control of Timbuktu (which really should just be 1 turn of fighting there).

 

[Dhoomstriker]

Even if Mansa doesn't whip right away in Timbuktu, you can COUNT on him switching to building Military Units at the end of Turn 0. That could easily mean +1 Skirm in Timbuktu, especially if the timing of overflow Hammers works in his favour.

It could also easily mean +1 Skirm that got sent from Kumbi Saleh defending Timbuktu, perhaps because Kumbi Saleh had a lot of units and AIs do like to have extra defenders in their capitals.

Mansa could also easily pick Timbuktu as his "rally point" for any "AI_ATTACK" Skirms to form his first stack of doom, and we'd be giving them 1 more turn to arrive there.


Now, if we wanted a "lean" army that had sieging Cats where we wanted to slowly Bombard at each City and thus didn't care how many defenders got stuffed by Mansa into a City and if we had time on our hands (there was no race to get to Alphabet), sure, we could just "reuse" our Catapult stack from Djenne on Timbuktu.

But, for this type of war, where our primary advantage is speed of attacking before a City can get a lot of defenders, a LOT of bad things can happen in the span of 1 turn to make us suffer a lot more losses in taking down Timbuktu.

Just the very fact of one extra Skirm arriving in Timbuktu makes it require 2 to 3 additional Horse Archers to kill it, due to Cultural Defences and Skirms having intrinsic City Defence ability could find us falling short of taking Timbuktu with an attempt to have "too lean of an army."


So, I really don't see how "less is better" and it really looks like "less risks requiring more," as the war drags on and we have to attack Timbuktu in a second battle, either on the turn after the first battle at Timbuktu or else later when the Djenne Horse Archers have healed.


Even if we capture both Cities with leaner armies but just have wounded units, we're just asking to lose another unit in a counter-attack. Every wounded unit that dies because a full-health unit did not exist to save it cost us more in resources than simply building the extra unit in the first place, given that Mansa will not be the last player in the game to feel the taste of our Military Units. Sure, in a true Space game, you might even be happy to see some wounded units die off here. But, we're not playing a game where we "just barely need to kill off Cathy's Riflemen, no matter how many Cavs it takes, even if we lose almost all of our army in the bargain," so a different line of thinking to our military approach needs to be applied, and I'm recommending an approach where we know that we'll have sufficient units to reduce Military Unit losses to a minimum.

 

[Jastrow]

As for a Cultural Bridge, any of Silver Bridge, Fish City, or Dual Banana being settled on the P Riv square have the potential to form a bridge of Culture to the south as long as there is a Coast square adjacent to one of the Ocean squares that our Culture will cover. We only need there to be one Coast square SOMEWHERE that is close enough, and given the size of the map, as I outlined earlier, you'd have to steal a lot of land from the southern AI to prevent there from being any sort of a Cultural Bridge.

OK, I was under the impression that double-banana had already gotten its border pop, as well as Fish-city. Has that not happened yet?? Can I can a screenshot of the southern see area then? I must have something completly confused in my mind... NEvermind... I just had a look at the save myself. We had discussed founding double banana before even copper, and I was thining we had... I also see Fish is still two turns away. I get it now... I also found something interesting however... See next post.

 

[Jastrow]

As I said, I just looked at the save... It looks very much like the water is exactly 3 tiles wide, and then there is land... It contains at least two mountains, and tons of bananas... Screenshots are attached, but resolution is not optimum, so you may need to look for yourselves in the real game. Settings are maximum zoom, and a field of view angle of about 20.

First picture shows two mountains... A broad one 2S of the visible fish (which is fish-city) fish, and a pointy one on tile to the west of it.

Second picture shows the same pointy peak, and the top of a banana tree to its west.

Third picture shows two more banana trees

4th are the same too peaks with a better field of view setting. They are the only two mountains I saw, but there are somehting like 5 or 6 total banana trees visible.

5th shows that the square south of the mountain at silve bridge SHOULD definilty be coast, as you see the waves, but I also see waves on the square the map maker made into ocean, so unfortuntaly, that is not proof of anything.

Despite the water being only 3 wide, to me, it looks like the soutern edge of the second tile is too dark blue to be adjacent to coast all the way along the path (except at the Silver-bridge end), but I am not the worlds best fog-gazer, so would appreciate a second opinion on that.

 

[Dhoomstriker]

Ugh, the forum ate my post... new plan while I have a few seconds to type it... 18 Horse Archers + 2 Spears + 0 Archers... attack Timbuktu -> Kumbi Saleh -> Djenne

Gift Dual Banana City that gets settled on the G Riv Jungle on Turn 0 of the war

1 Spear guards Prague, 1 Spear and 1 Horse Archer pillage Djenne's Copper, 1 Horse Archer defends from Mansa's roaming units, the other Horse Archers are for Timbuktu plus defending the captured Timbuktu, plus capturing Workers around Timbuktu.

 

[Mitchum]

@ Jastrow

Good find. Based on what you saw, what do you think we should do? As you said, the resolution isn't great so it was hard for me to see what you were talking about. Do you think that there is a wall of peaks all the way from the east side of the map to just south of Silver Bridge with the only possible opening south of Silver Bridge as shown in picture #5? Should we think about getting a settler over there ASAP? Should we use our Dual Banana settler? I think it will be a different land mass if we can reach it which would provide several +2 TRs in our core cities. Before we use our current settler to found Dual Bananas, we should see if it would be possible to get a city set up over there.

@ Dhoom

I see that you have updated war plans. I'd like to hear more when you have a chance. I'm not sure why we wouldn't attack Djenne on T0. In any event, the next PPP covers the buildup of units, not necessarily the actual attack itself. Also, I guess the plans could change based on any advanced scouting units we have in Mansa's lands before the DoW.

 

[Mitchum]

I just looked at the save and it does look like there are peaks south of Fish with bananas all along the coast. I'd like to get a unit on our galley that is just SW of Rice. Maybe we should load the Rice archer on the galley and move the warrior in Fish back to Rice (if we feel we need an MP unit there. I'm really hopeful that there is a reachable AI down there.

 

[Jastrow]

I am not sure what we should do... As I said, the water on the soutern edge of the second tile looks to dark to me, so I am afraid that the map make put more of his "coastal ocean" there, such that it is impossible to cross, even if it is only 3 tiles wide... IF that is right, it might argue for an Astro beeline, but we will have a much better idea if that is true or not once Fish pops its borders, which happens in two turns, I beleive.

 

[Dhoomstriker]

I see that you have updated war plans. I'd like to hear more when you have a chance. I'm not sure why we wouldn't attack Djenne on T0.

Do you want to go in with less Horse Archers or not? That's what I'm offering with just going after Timbuktu and leaving Djenne around for later.


I listed several reasons why taking Timbuktu on Turn 1 greatly increases our chances of capturing that City with minimal losses.


Monte isn't Mansa, but he did slowly collect units in Djenne in a test run... but it turned out that they were his gathering attack force and he eventually sent it out toward our City of Prague.

The same thing could easily happen in Timbuktu, instead, if we delay going after Timbuktu. The difference is that Djenne is in a vulnerable position, as we can camp our within our Cultural Borders without committing to an attack until the timing is right. Timbuktu is deep within Mansa's Culture, so we don't have that flexibility.


Skirms are a joke to deal with in the open... even if they camp out on the G Jungle square 1N of our City of Prague, Horse Archers and Protective Archers get good defensive odds against them. So, I'm not worried about a stack of units building up in Djenne that might later come after us. In fact, that situation would be ideal... let Mansa collect his units in Djenne, send them at us, then we take down his units in the field and an easily-capturable Djenne, in whichever order makes the most sense at the time.


The war plan directly affects the build-up since we need to decide how many units to build and the war plan will guide that decision.

Get your war plan in place then figure out how to get the resources you need to make it happen... don't get a few resources and then try to throw them at something in hopes that you'll just get lucky in battle.

I'd like to get a unit on our galley that is just SW of Rice. Maybe we should load the Rice archer on the galley and move the warrior in Fish back to Rice (if we feel we need an MP unit there.

Sure, that should be easy enough to do.

it might argue for an Astro beeline

Getting Alphabet to get Code of Laws to get Caste System to get a Great Artist beats getting multiple Great Scientists to Lightbulb Astronomy

Spoiler :

 

[Mitchum]

The war plan directly affects the build-up since we need to decide how many units to build and the war plan will guide that decision.

Agreed. But plans can change based on information on the ground, right? I'm fine if you build up the planned HA army. But, depending on how many units are in each city, we can and should adjust plans accordingly. I'm sure that you'll have an early HA scouting each of Mansa's cities as soon as you can. Depending on what we see, we may decide to add or subtract units or wait a few turns to DoW until a settler leaves Timbuktu. So I don't want to set in stone any war plans until the turn we pull the trigger.

BTW, you talk about units escorting settlers turning back and going to defend the city. If anything, I've seen the opposite. Even with a stack of seige outside a city, I've seen an AI send out a settler party the turn before I was planning to attack. It has to do with what script is assigned to the units.

Regardless of what we do, I'd like to get as many units as we need as fast as possible. If we're going in with HAs, the sooner the better before Mansa either builds up a ton more units or he researches Feudalism.

I know that you ran some odds calculations earlier with HAs. I plan to do the same. 17 HAs sounds like a lot but if Mansa has an odd spear around, it may not be enough...

EDIT: Regarding a Great Artist, I think Jastrow has shown from fog gazing that we don't need to bomb a GA to access the land across the channel to the south.

@ Dhoom When do you plan to have a PPP ready?

 

[Dhoomstriker]

BTW, you talk about units escorting settlers turning back and going to defend the city. If anything, I've seen the opposite. Even with a stack of seige outside a city, I've seen an AI send out a settler party the turn before I was planning to attack. It has to do with what script is assigned to the units.

Can you conclusively say that this information comes from BtS games? Settler Parties in BtS act significantly differently than from previous versions.

One noteworthy point is that when a Settler Party gets threatened, in BtS, if a Settler Party is near a City, the Settler and the other units can get split from each other. This fact can mean that a Settler can be left defenseless (but unreachable by your units unless you have a unit that "breaks" the unreachable definition of an AI, such as Woodsman II unit that is adjacent to a Forest) but it also means that the AI units can readily take on a new assignment immediately, such as guarding an AI City. This behaviour is quite different from Vanilla + Warlords where you were more apt to see a Settler Party appear to completely ignore your incoming army and just go about its business of setting up a new City.


I agree that plans can change and will need to be flexible, but having agreement on 18 Horse Archers will require a different approach than building only 12 Horse Archers and the timing of the arrival of our units will even make it different from an approach requiring 23 Horse Archers.


Yes, Feudalism could totally change things... for example, when Mansa learns Feudalism, we'll still have a small window where he might only have 1 Longbowman per City (although that situation tends to occur more on lower difficulty levels) and we'll minimally have a small window where those Longbowmen don't have their Fortification bonuses (since you lose said bonus when upgrading a unit)... depending upon what our scouting shows and how many units we have ready at that time, our plans may have to change to doing something like going after Djenne, then playing more of a cat-and-mouse game with Mansa where we try and bait his Longbowmen to come out into the open before going after Timbuktu.

Timbuktu is the key City, since we'll need to capture it before we can get a suitable western Galley City, but if Timbuktu turns out to be impenetrable thanks to a stack of Longbowman, then capturing Gems City and slow-building a Galley will have to be our backup plan.

I know that you ran some odds calculations earlier with HAs. I plan to do the same. 17 HAs sounds like a lot but if Mansa has an odd spear around, it may not be enough...

Spears are yucky, but they, too, fail to perform well once they've been wounded, so it's more of a matter of getting in some initial hits on them and then they'll die like any other unit to repeated attacks. The two "Worker-stealing Horse Archers" are also insurance against being stonewalled by any unit, be it a Skirm, Spear, Sword, or whatever, where by "stonewalled" I mean having a couple more of our units than the odds call for going in, attacking, and dying without doing any damage.

EDIT: Regarding a Great Artist, I think Jastrow has shown from fog gazing that we don't need to bomb a GA to access the land across the channel to the south.

I agree, but I was more directly addressing Jastrow's comment about potentially needing Astronomy and showing how on a small-sized map like this one, we are very unlikely to need Astronomy and if we do, there's going to have to be a massively-wide Ocean somewhere to make us need Astronomy, and even if there is such a thing (say, with an AI's capital in the far south-west... we know at least 1 Creative AI's capital is in a corner, thanks to bbp), it should be sufficiently far off that getting to Astronomy at that stage of the game should be trivial... we know that we can at least get to the south-east (using a Great Artist, although fog-gazing indicates that shouldn't even be necessary) and north-west (we can already see enough Coast squares) without Astronomy.

@ Dhoom When do you plan to have a PPP ready?

I'll basically need some time to sit down with the test game and figure out the best whipping strategy.

With any luck, we'll meet an AI to the south who might not trade us Alphabet but could trade us Iron Working, which could throw off too long of a plan, so it might makes sense to only plan out 5 turns or so.

I suppose that it makes sense to position our Workers optimistically in hopes of getting Iron Working soon.

If Mansa baits us with, say, 2 Workers and we have an army of, say, 8 Horse Archers ready, we may even decide that with Iron Working in hand, it'll be worth it to grab Djenne and a couple of Workers at the cost of a drawn-out war (i.e. +2 Workers "sooner" being deemed more valuable since they'll let us improve more Pigs to be able to whip more Horse Archers), so planning out to a full 18 Horse Archers build-up probably doesn't make sense.

Either way, we'll need our 18 Horse Archers or more, even if we declare war prematurely, so Police State is definitely going to be used.

With any luck, getting Iron Working will also mean that we won't need the Happiness from Representation and with only 1 Great Scientist and 0 Scientist Specialists (sorry, they will be sent to the Farms or to ride horses), Representation won't be needed for a while to come.

 

[Mitchum]

Regarding Police State, we should compare what we'd lose with respect to F, C and H for one turn of anarchy vs. how many hammers we'd "gain" in PS. Then the decision becomes quite clear as to which way to go... or at least we'd know the tradeoff and could make an educated decision. We'll also have to factor in the added cost of running the PS civic.

Here's a first crack at what we'd give up losing one turn's production to anarchy based on current production and citizen assignments:

Aachen: 7F, 7H, 18C
Prague: 3F, 5H, 12C, 15B
Ivory: 2F, 9H, 7C
Rice: 2F, 7H, 9C
Island: 4F, 5H, 13C
GP Farm: 5F, 4H, 10C
Horse: 1F, 5H, 9C
NE: 3F, 1H, 4C, 7B
Fish: 2F, 2H, 10C
Silver: 2F, 1H, 8C

Note that cities that are building wealth had their hammers included in the H number and were subtraced from the C number.

Total: 31F, 46H, 100C, 22B less expenses (85C) = 31F, 46H, 15C and 22B.

At this time, Police State costs an extra 4gpt.

So, we'd need to compare the 25% hammer gain from PS against the loss of 31F, 46H, 15C, 22B and an extra 4gpt.

Does that sound right?