SGOTM 13 - One Short Straw

[Dhoomstriker]

Oh, I looked at the save and I cancel my suggestion to trade Code of Laws to Cathy for her Gold. She is the Worst Enemy of 3 AIs, so trading to her would be a mistake. Those 3 AIs are Vicky, Joao, and Isabella.

Part of the problem right now is that we have a big situation to deal with in terms of Ragnar, but we also need to have a clear plan of action for all of the other little details that go on (such as, in this case, NOT trading Code of Laws to Cathy for her Gold).

That's another why I'd like to see very clearly-stated turn-by-turn plans of everything, so that we can be clear on which ideas are going to be executed and which ones are not going to be used.

 

[Dhoomstriker]

I agree that we can afford to delay trading for calendar. Trading for calendar with COL will set us back 125 gold, and we don't have 125 spare gold. If we traded currency to Cathy at the last moment, and we made her our first target she would not have much time to make use of it.

I'd like to get the idea of trading away Currency out of your head.

The AIs have Open Borders with each other and they're soon going to be loving each other to pieces with 2 AIs having a Favourite Civic of Hereditary Rule.

Trading away Currency = +3 Commerce per AI City

+3 Commerce per AI City = Massive AI REXing

Massive AI REXing = A really tough situation to deal with in a non-Vassal Conquest game

If we don't trade for Ragnar'g gold where else could we get it? He may be more likely to research feudalism once we gift it to him, but given his relatively slow tech rate I don't think we would have too much to worry about.

Think bigger picture here.
50 Gold is hardly anything to us. It's roughly 1 to 1.5 turns at a 0% Science Rate.

To Ragnar, Mysticism is likely to get him back into the AI-AI trading game. That means 5 AIs able to trade around techs to each other instead of 4 AIs.

No one really wants to trade him anything now, both because he is the dumbest AI and because there are no longer techs on the table that are roughly equivalent to the value of Mysticism. Ragnar getting Mysticism opens up a possible Meditation + Polytheism for a more expensive tech trading deal, and with him knowing as many or close to as many techs as Isabella, AIs will trade with him again. I don't know how to say it any better than: "IT'S A TRAP!"

 

[Dhoomstriker]

1) Accept that we are never going to get theology on a useful timescale.

Correct, let's forget about Theology.

Besides, without building 2 Taoist Missionaries or 2 Confucian Missionaries, we can only have 4 Cities with the same Religion as each other.

We've also decided against Forges, so let's forget about Organized Religion.

We also want to whip up an army, so Pacifism isn't a good choice.

That leaves us with Paganism, which is a good choice, as it is a Low Upkeep Civic, and while costlier Civic-wise than Pacifism, it won't cost us additional Unit Support Costs like Pacifism will.

However, I believe that the actual Civic switches should be part of the next turnset anyway. That said, it would be appreciated if you left at least one unit unfortified whenever uploading the saved game, so that we don't accidentally advance the turn. This approach becomes even more important on the last turn of your turnset, so that we don't accidentally advance the turn to a point past our Golden Age without having switched our Civics.


In the next turnset, we could potentially choose to switch to No State Religion. However, +1 Happiness from having a State Religion is still probably worth staying in Confucianism for.

The question is: Would access to theocracy make up for delaying the start of our wars by 5T?

For Theocracy in a maximum of 4 Cities? No.

 

[Dhoomstriker]

on T+1 Trieme 2 moves to... 2S 3W of marble. Trieme 1 moves 1N 2NE.

While these positions are not going to block Ragnar, in that we need to also immediately send Trireme 3 to the west and then keep sending Trireme 3 and our Galley to the west in order to even have a PRAYER of blocking Ragnar anywhere, I think that you're going to have to make these moves anyway.

For example, that movement path for Trireme 1 is the fastest way back home. For Trireme 2, assuming that Ragnar is going towards Paris, that position is the only place that Trireme 2 can still "stay ahead" of Ragnar.


The only alternative that I see is to stop Trireme 2 1W of there, just inside of our Cultural Borders. The biggest problem with doing so means that we'd almost for certain lose Gold City (and thus our last Great Person, since accumulated Great People Points disappear when a City is captured).

We're going to have to assume that Ragnar has 3 full Galleys (the Catapult evidence supports this point). If Ragnar declared, he wouldn't want to move onto our Trireme there, but then he would also be blocked from accessing Paris. You might think that this situation is good, but you have to plan ahead for what Ragnar will do when his path to Paris is blocked. What will he do? Find a new land target. At that point, only Gold City will be a valid land target. Because our Galley and Trireme 3 are too far away, and because our Triremes will be "behind" Ragnar's boats, the most that we could do is hope to get lucky with the random number generator and sink 2 of his 3 boats... leaving him free to capture Gold City.

Therefore, engaging Ragnar by moving Trireme 2 two squares east won't work. We'll have to move Trireme 2 three squares east like you are suggesting and then "buy some time" with future well-planned-out moves while we get our Galley and Trireme 3 to the west. Without both of those 2 boats coming to the west, I think that we will be guaranteed to lose at least 1 City.


The silver lining in this grey cloud is that we might as well allow Ragnar to declare war on us. Given our Triremes' current positions, we will gain nothing by declaring war on him.

 

[Dhoomstriker]

Given that we barely have a month left I'm not sure we have time to do this. Blocking Ragnar with 2 triemes is not that difficult as long as you are carefull and don't leave triemes on sentry.

These turns are too critical to play the "lack of time" argument.

As Duckweed said in the maintenance thread, 50 turns can be played out within a day--when you have the upper hand on the AIs and aren't about to lose one or more of your core Cities and when the fighting decisions are just about "how many Catapults/Triremes do we attack an AI City with before attacking with our Maces."

Currently, assuming that Ragnar is coming for us (which we have to assume), I cannot see any possible way for you to stop Ragnar from capturing at least one of our Cities with just those 2 Triremes.

So, I say it is a State of Emergency and I'd like to see the action-by-action visibility into the process.


I can accept mistakes, since they happen to everybody, and we can deal with mistakes, learn from them, and move on.


What I cannot accept is that when the situation is critical enough for us to lose a core City that we can't have visibility into the Active Player's exact plans.


I will be away for the next few hours, but I can be here to support you... if you will let me.

 

[babybluepants]

I don't see how the suggested block deals with every conceivable scenario. He could be going for Gold via Marble-1S5W, get blocked by our T+2 Trireme1 move and then switch targets to Marble? I'm not sure how an AI deals with a blocked target city, since it's not a very common situation. In play-testing some Warlords this morning I saw them attack two cities simultaneously in an odd map situation, and even switch specific unit targets on a turn-by-turn basis.

That said, I also fail to see what else we can do here. It's a shame our second trireme is half-turn behind... The only more extreme action would be to insta-whip another trireme, taking the hammer penalty. We can't really deal with a landing in either Marble or Gold on a 2t schedule. I also don't particularly like committing all ships west, as they won't make it there in time and we could get attacked by barb galley next 2t, as well.

 

[shyuhe]

Does anyone remember how high of a modifier you have to reach with Ragnar for him to be pleased? If we can rack up +4 fair trade this turn, it'll bring him to +7. We can then beg gold from him to take a 10 turn peace while we get some more boats in place. This will give us enough time to get triremes in place to fend him off.

If that plan doesn't work, I don't see how we can get by with our two triremes -- Ragnar's galleys can out run one of them.

 

[Mitchum]

Worse than losing a city would be losing a GS. I assume that they are scattered around our empire in the cities that spawned them. We need a GS protection plan too...

 

[Dhoomstriker]

Does anyone remember how high of a modifier you have to reach with Ragnar for him to be pleased?

He is currently Cautious at +1. From my old saved games, I see that he was Annoyed at -2 and -1. So, there is a chance to get him up to a Pleased status with gifts of techs and/or Gold.

If we can rack up +4 fair trade this turn, it'll bring him to +7.

Not quite. But gifting him techs and Gold could bring him up to +5 towards us. What I don't know is whether that's enough for Pleased status. I THINK that "Pleased status = the lowest Cautious status +5." If that's true and if Ragnar is Cautious with us at "0," then gifting can get him up to Pleased status. But, I don't know what his status is at "0." mdy, do you have any screenshots or saved games that you can check to see if Ragnar was ever "0" with us and if he was, was he Annoyed or Cautious with us?

Cautious at 0 = it should be possible
Annoyed at 0 = it probably is not possible, but again, I'd need someone to confirm my "lowest Cautious status + 5 = Pleased status" equation, because I could be wrong and it might be possible if "the lowest Cautious +4 = Pleased"


But, what would we be able to gift?

We have 5 possibilities:
Alphabet
Mysticism
Currency
Metal Casting
Gold


Of them:
Alphabet = Fine, we can give away Alphabet.

Mysticism = that's like gifting him up to Monarchy and enabling a ton more AI-AI trades. Ragnar would not be the only AI that would rapidly increase their tech level (and yet we still would not see Construction getting traded-around, so this situation would NOT help us). I'd really rather not give him Mysticism at any cost, even more because:Hereditary Rule is his Favourite Civic

So, then we'd have a bunch of AIs that would get even faster to being Friendly-Friendly with each other and with no dumb kid on the blick, AI-AI trading could quickly get out of hand. All because of one little decision.

See, the developers weren't all that unfair. But, people complain about the fast tech pace of the AIs. Well, that's often because the human player gets in there and messes up the scheme of things by making the "dumbest player" no longer be the "dumbest player." It is very often a mis-management of a human player that creates an insane tech pace... the exception to this rule is the beloved Mansa Musa, who basically acts like a poorly-playing human tech trader would play. Don't take my word for it, but tech trading is one of my strengths in Civ 4 and this fact is due to observing and sticking to simple guidelines, one such being: "don't bring the dumbest AI up to a level that is relatively equal to or higher than the second dumbest AI."


Currency = that's like gifting him +12 Commerce per turn and the very real possibility for that tech to get traded-around pretty quickly, since it has a very low Monopoly threshhold.

Metal Casting = we had better not, given that it enables Triremes.

Gifts of Gold = very feasible, and a really good option, as we want to keep Ragnar from getting 2 techs from us if at all possible


The next steps
1. So, first we'd like to find out if he was Annoyed or Cautious at an overall Attitude total value of "0." (That includes positive and negative Diplo modifiers summed together).

2. Next, if he was Cautious, then we want to confirm or deny that "the lowest Cautious value +5 = Pleased."

If the above two look like they will work, then I suggest the following:
3. Gift Alphabet

4. If we got +4 from Fair and Forthright Trading, then we are done with gifting, regardles of whether or not he is Pleased

5. If we got +2 or +3 from Fair and Forthright Trading but are not Pleased (I don't see how we'd be Pleased but it's worth checking), then we start gifting him increments of 10 Gold at a time until we get one more +1 from Fair and Forthright Trading with him

6. If we were at +2 and made it to +3, then we will know if we have enough Gold remaining to be able to arrive at +4.

7. If we have enough Gold to get to +4, keep gifting Gold in 10 Gold increments until we reach +4 Fair and Forthright Trading and then stop gifting.

8. If we don't have enough Gold to get to +4, then we'll gift Mysticism and then 10 Gold increments until we reach +4 Fair and Forthright Trading and then stop gifting.


I really, really, really don't like the idea of gifting/trading away any of Mysticism, Currency, or Metal Casting to Ragnar. If you feel like the AIs are expanding rapidly now, just try giving/trading Ragnar one of those techs and mark my words: you will watch the AIs just run away with the game.

We can then beg gold from him to take a 10 turn peace while we get some more boats in place.

Actually, no, we wouldn't ask for 10 Gold. We'd Demand as much Gold as we can safely get. There is absolutely no difference in the code between asking for a minimal amount and an amount that an AI will give to you up to the maximum Demandable amount. All that we'd do is throw away potential Gold... it would be like selling a tech for 10 Gold when an AI has 100 Gold and will pay you all 100 Gold... you'd simply be missing out on 90 Gold (plus having a minor difference in Fair and Forthright Trading if you weren't already at +4).


I will perform a calculation shortly that assumes that Ragnar has the highest Power level listed and will then give a very, very, very safe value that we can feasibly Demand (i.e. no chance of failing). Already, that's being overly conservative, since I believe that the last time that I checked, Willem actually had the highest Power level, but we'll pretend that Ragnar has it just to be ultimately safe with our Demand amount.

If that plan doesn't work, I don't see how we can get by with our two triremes -- Ragnar's galleys can out run one of them.

If he's going for Paris, then there is a way to force him to "lose" 1 movement point in the most direct route to Paris. In so doing, we could have 3 Triremes attack him on the same turn (since Trireme 1 just needs that one extra movement point to be able to reach the Galley stack), while also having the Galley as a "clean-up" attacker for any Trireme that failed in its attack. In addition, attacking on the same turn with all 4 of our boats prevents a single loss from allowing one of Ragnar's Galleys to get Combat II and mostly-full health from promotions.

Of course, if 2 Triremes lost their fights, then some of Ragnar's troops would land. At that point, we'd simply hook up our Iron, pay to upgrade a Warrior to an Axeman/Spearman, and then pray.


Here's the Demand calculation, again being very conservative and assigning Ragnar more Power than he likely really has:
{[(166 - 107) * 2] * 100000} / 162000
= {[59 * 2] * 100000} / 162000
= {118 * 100000} / 162000
= 11800000 / 162000
= {11800000} / 162000
= 72.839

Throwing in an EXTREMELY safe fudge factor would give us a value of 50 Gold.


So, we should not be Demanding 10 Gold but instead be Demanding 50 Gold. Technically, we could risk asking for a higher amount, but 50 Gold is a VERY safe amount to ask for... but only if he's Pleased with us.

At our low Power level, if we cannot get him up to Pleased, then forget about making a Demand. You never know... if we don't make a Demand and we wait a turn, we may get additional positive Diplo modifiers, such as for Years Supplying a Resource. Also, if we wait a turn, we could even consider doing something like switching to No State Religion on the following turn when in our Golden Age, which would also boost our relationship with Ragnar (although at the cost of several GPP). Basically, if he's not Pleased with us, don't make a Demand, and we might still have options. But, if we make a Demand when he's Cautious, we will throw away that possibility.


As it is, I will still ask that you send Trireme 3 three squares west. Moving it there won't hurt us and may end up being necessary if this "Pleasing Ragnar" approach is not effective.

 

[Dhoomstriker]

As for the Golden Age timing... there IS one idea worth considering here, but I think that it really should be tested:
IF we can get our final Great Person with TWO TURNS remaining in the Golden Age, then:
1. With 2 turns to go in the Golden Age, we can Lightbulb our way up to Astronomy
2. We then set Research to Civil Service
3. With 1 turn left to go in the Golden Age, we'll be able to switch Civics, including into Bureaucracy


What needs to be tested is:
Would we need to delay our Golden Age for 1 turn or for 2 turns in order for this plan to work (i.e. in order for us to generate our last Great Scientist with 2 turns remaining in the Golden Age)?


Also, obviously, if Ragnar can't get up to Pleased as per the algorithm used in the last message (gifting Alphabet up to +4 and if not there, gifting Gold up to +4, etc), then we may need to launch the Golden Age "on time" in order to be able to switch out of our State Religion to No State Religion in order to try and get Ragnar up to Pleased.


I don't see switching into Theocracy as a possibility at all, but Bureaucracy certainly is... and it could be worth a 1 or 2 turn delay (which needs to be tested to see if it really is 1 or 2 turns) if it means getting ourselves into Bureaucracy without 1 turn of Anarchy, right?

 

[Dhoomstriker]

By the way, I noticed that Vicky is Pleased with us, but only has a +3 positive Diplo modifier for Fair and Forthright Trading with us.

If you are desperate for Gold, you could consider Demanding all of her Gold (20 Gold) immediately.

If you aren't desperate, then she'll probably eventually go down to +2 Fair and Forthright Trading, at which point she would not be Pleased with us, so, some time over the next few turns (i.e. some time before she stops being Pleased with us) you should aim to make a Demand from her (but make sure that she is still Pleased with us before Demanding). A Pleased Demand should have different text anyway... such as "Can you Spare this for a Good Friend?" instead of harsher-sounding language that sounds more like a threat for the case when you the AI is Cautious or lower towards you.


Of course, the sooner that we make a Demand out of her, the sooner that we'll be able to ask a second time in the future, so that's another factor favouring an immediate Demand.


At worst case, after a few turns, she'll get a ton of Gold somehow (a trade, a Failed Wonder, whatever), at which point you might regret the Demand... except that at that time, we could also just trade Monotheism to her for her excess Gold (but only if Ragnar does not get Mysticism--if Ragnar gets Mysticism, then we really don't want to "cheapen" the value of Monotheism any further by giving it to Vicky--so, Don't Give Mysticism to Ragnar (TradeMark) ).

 

[Dhoomstriker]

Our Worker that is going to stay on Iron Island (now that it looks like the Roads are pretty much done there) should probably get positioned to the north side of the island... either just inside of our Cultural borders as far north as it can go or else on top of the Silk Resource. Probably within our borders whenever we have sufficient units outside of our Cultural Borders in order to avoid paying +1 Gold in Unit Supply Costs, but feasibly on the Silk if we don't have too many units outside of our Cultural Borders.

I think that right now--without checking--we probably have too many units outside of our Cultural Borders for the Worker to be placed on the Silk immediately, but it can at least go as far north as possible within our Cultural Borders.


That way, it will offer slightly more spawn-busting potential.

 

[babybluepants]

Did we not already beg 1gpt from him earlier in mdy's TS? I somehow thought I read that in a report.

Edit: I really don't see why he'd be going for Paris. They usually attack the closest reasonable city, IMHO. Either Gold or Marble seem much more viable to me.

 

[shyuhe]

I think we asked for gold from Willem. We haven't been pleased with Ragnar so I doubt we begged gold from him? I may be wrong though.

To be honest, I'm pretty sure we'll have to give Ragnar multiple techs to get him to pleased at this point. How close are we to getting astronomy? If we are < 10 turns away, I'd gift Ragnar alphabet + MC since we will get galleons to destroy his little galleys. That should almost certainly push Ragnar to pleased.

 

[Dhoomstriker]

Did we not already beg 1gpt from him earlier in mdy's TS? I somehow thought I read that in a report.

Really? That would have been weird, since we most certainly did not have Ragnar at Pleased and would have thus failed had we done so.

We also would have received a -1 negative Diplo modifier, which I did not see in the game, so I doubt that we did so.

EDIT: Cross-post with shyuhe.

I really don't see why he'd be going for Paris. They usually attack the closest reasonable city, IMHO. Either Gold or Marble seem much more viable to me.

I admit that it's an assumption, but is as good of a guess as any: go for the jugular by taking out a player's capital and a nice shiny Wonder at the same time. That what I would do if I were in Ragnar's position, and I imagine (since AIs know our Cities' defenses) with Warriors everywhere making defenses equal from City to City, that this target would be the highest-valued target.

When you saw an AI "shifting targets" were you per chance changing the land-based defenders in the relevant Cities?

If he's going for Marble City, there's not much that we can do to stop him if we can't get him to Pleased. At least we'll have just recently generated the Great Person there for our Golden Age, so losing the City (while it would suck) would probably be the "best" out of our Cities for him to capture.


He might simply even go for any City that doesn't have a defender in it, which would mean that we might be able to "toy" with him by keeping only our eastern Cities empty of a defender. I don't have the game open now, but I THINK that Stone City did not have a defender, which might actually make it a higher-priority target.


There is the possibility of leaving Iron City's Warrior outside of the City, but then if Joao sneak attacks, he'd probably waltz right into the City, thus I think that leaving Stone City would be a safer option, assuming that doing so can actually influence Ragnar.

 

[Dhoomstriker]

To be honest, I'm pretty sure we'll have to give Ragnar multiple techs to get him to pleased at this point. How close are we to getting astronomy? If we are < 10 turns away, I'd gift Ragnar alphabet + MC since we will get galleons to destroy his little galleys. That should almost certainly push Ragnar to pleased.

Actually, it might be easy enough to check.

Assuming that Ragnar is not already self-teching Alphabet, I can simply advance the test game that I used for the Construction-trading testing to the same turn as the real game. In this particular test game, I had us "meet" the AIs at identical times (including on "in between turn times" for Cathy and Vicky--i.e. those girls "met us").

Just about everything else in that game is wrong--it's not really island-based... I had to World-Build-out a lot of the land just to make us be isolated until post-Astronomy so that I could meet the AIs on the right turns... but this aspect should be the same.

All that I need to do is put us on the same game turn, give Alphabet to the other AIs that have Alphabet in the real game, and then, assuming that Ragnar isn't researching Alphabet, we can see how much he'll give us.

Since we only met him 59 turns ago, I estimate at least +2 from Fair and Forthright Trading, but I'll see what the test game says. Again, if Ragnar is already researching Alphabet, then the value might not be correct, but that's about the only variable that I can't control in the equation.

 

[shyuhe]

Ah, found my answer I was looking for (an old Silu post!):

Is there a quick and dirty way to estimate how valuable a tech gift must be or how much gold is needed to get points of this bonus? Does it vary by leader? Map size, gamespeed, difficulty level? Also, if you miscalculate and only give them 99% of what they required to get +1 to this bonus is it all wasted, or do they have some kind of memory that adds the value of multiple gifts/advantageous trades together to reach the threshold?

The bonus is pretty simple. There's a permanent value stored (that never decays) that records your "fairness" towards an AI. This value is divided by the turns you have known the AI, multiplied by 10. Tech gifts give approximately 1.5x its beaker price (depends on how widely known it is), I think gold gives 1 per 1g. This should give you some idea on how much you need to gift. So if you've known someone for 200 turns, you'd need at least 8000 "fairness points" to get +4 fair trade. This is why you can get +4 for 20g on the turn you meet someone (also why it "decays" so fast then)

Pretty sure this never varies between levels (except the inherent difference of beaker costs) or leaders. On Marathon techs give triple value, but there are three times as many turns so it evens out. Gifting units for fair trade is less efficient on Marathon (* 2/3) but I've yet to play or see a game where the situation warranted gifting units for fair trade

And yeah this acts as a memory, so you can gift 10g a hundred times and it has the same effect as gifting 1000g.

We're at turn 166, and have known Ragnar for about 59 turns. So each +1 fair trade modifier costs 600 fairness points. Since we want +4, we need 2400 fairness points or about 1600 beakers

Alphabet is 450, currency 600, metal casting 625. We would have to give him all three of those techs to reach +4 fair trade

 

[babybluepants]

Here - I did reasonably good estimates last game. Don't really remember the math now, but if you wanna give it a shot, Dhoom, go for it.

Edit: xpost with shyuhe

Edit: Note that it should be Turns*5 rather than Turns*10 - explained in the next post by Silu/LC.

 

[Dhoomstriker]

Assuming that Ragnar is not already self-teching Alphabet, I can simply advance the test game that I used for the Construction-trading testing to the same turn as the real game. In this particular test game, I had us "meet" the AIs at identical times (including on "in between turn times" for Cathy and Vicky--i.e. those girls "met us").

Here's your answer: Alphabet = +3 Fair and Forthright Trading

So, what are the odds of Ragnar self-teching Alphabet in the real game messing up this value? Very, very low. He values this tech as a "0" in terms of his tech-selection process.

Thus, this idea has a lot of merit, with only Gold being required to "top up" Ragnar to +4 Fair and Forthright Trading.

What I still don't know is if that +4 will be enough to get him to Pleased.


Edit: Cross-posted with the both of you.

 

[Dhoomstriker]

We're at turn 166, and have known Ragnar for about 59 turns. So each +1 fair trade modifier costs 600 fairness points. Since we want +4, we need 2400 fairness points or about 1600 beakers

Are those numbers perhaps for Normal Game Speed? SGOTM 13 is using Epic Game Speed...