TT01: Beating Demigod

[vmxa]

Lurker:

I don't know the details of the article, but for sure you are going to get an unhappy AI if you plop a town next to their borders. How they will react, depends on the relative strengths and other factors.

It has been used to provoke the AI and that is why they disallow it in Civ4. You cannot plant a town next to their borders.

Was it written prior to C3C?

 

[ThinkTank]

I think he was suggesting something like this:

(1) A and B are consistent (i.e. do not imply a contradiction).
(2) B implies C
(3) A and C are not consistent

This does make some sense, but it is logically invalid. In fact the following argument schema is valid:
If A and C are not consistent, and B implies C, then A and B are not consistent.

Thinktank,

Let B=F, C=T. In such a case B->C. Since A and C are not consistent, A can't match C. So, since C=T, A=F. But, B=F also. So A and B are consistent, even though A and C are not consistent and B->C. This completes the counterexample to present schema.

No, it does not, assuming you mean with B=F and C=T that B is false and C is true. Consistent means possibly true. In your counterexample, you conclude from A and B both false that A and B are consistent. This conclusion is invalid.

My original claims stand. As said, that two sentences are consistent means that they are possibly true simultaneously. With B implies C I meant logically implies, I suppose should have said entails. This means that in all situations in which B is true, C is also true.

Here is your proof.

(1)Suppose A and C are not consistent, so there is no possible situation in which both are true.
(2)
Suppose B entails C.
(3)
And suppose moreover that A and B are consistent, so for some situation s, both A and B are true.

Since B entails C, in situation s, both A and C are true. This contradicts, assumption 1. So there cannot be a situation in which both A and B are true, so A and B are inconsistent.

 

[darski]

Regarding the aggressive settling. i did read an article wherein the author stated that when he settled that close, the Civ declared within 10 turns. I am trying to find the article/ post. I think it might have been in an SG I was reading from days long gone.

 

[Chamnix]

lurker's comment: I've seen so many people post explanations on why the AI chooses to declare war, but frankly I don't think anyone has any idea. If you found a city that crowds them, sometimes they'll declare, sometimes they won't; if you have no cities anywhere near them, sometimes they'll declare, sometimes they won't; if your military is much weaker, sometimes they'll declare, sometimes they won't; if your military is much stronger, sometimes they'll declare, sometimes they won't; if you have a luxury they lack, sometimes they'll declare, sometimes they won't; if you have nothing they want, sometimes they'll declare, sometimes they won't, etc.

Obviously it is chance-based, and people are guessing what increases the probability, but any theories are very much subject to cognitive bias. Personally, I don't think that that programming is that complicated - my theory is that it is attitude/aggression setting based and nothing else . Regardless, I never worry about it one way or the other since I don't think anyone knows the truth. I just play the game without basing any decisions on whether or not my actions will make the AI more likely to declare.

In Amsterdam, we need 2 more shields for the settler factory to work, 1 at size 4 and 1 at size 5. This means the second wheat should be mined and the BG just S should be mined. Our worker is mining the silks right now, it should stop doing that and instead road it (to help with the money situation). This means at least 15 worker turns (2 mines and 1 road), so I think after the granary Amsterdam should build another worker before going on settlers.

lurker's comment: I might be inclined to build 2-3 workers before a settler. Never underestimate the importance of sufficient workers very early.

IMO, Amsterdam should finish it's granary, knock-out a Settler then build barracks, after which it should be used primarily as a military pump, only creaming off Settlers as growth allows; I feel our expansion is going to come at the point of a sword, this game.

lurker's comment: Specialize your cities. You are already committed to a granary in Amsterdam, and it is high food. I would build nothing there except settlers and workers until I had plenty of each.

On a game related matter, TT - what do you want to do about 'demands'? I've played in some games where it's left to the player's best judgement, and others where there is a pre-agreed response.

In principle, I think we should cave to demands unless we are (1) ready and (2) willing to stand for the military consequences of a refusal. Right now, I'd say both criteria are off.

lurker's comment: I don't think there is any question - generally you have to refuse give in to early demands at these levels. Even if you think you are safe because the demander is far away, he is far more likely than you to be able to buy your neighbor into an alliance.

 

[Spoonwood]

Consistent means possibly true.

Well first off that's not what I meant by consistent... usually consistent means that two statements do not contradict each other... they have the same truth value. If A=B, then A does not contradict B. Second, you've jumped out of classical logic into a modal logic of sorts by talking about "possibly true". We need another sort of analysis for validity here.

With B implies C I meant logically implies, I suppose should have said entails. This means that in all situations in which B is true, C is also true.

It means more than that in classical logic. It means that FF, FT, TT work out as possible truth values of B and C respectively with B first and C second (FF means B=F, and C=F). The truth table for implication or entailment goes

P Q P->Q
F F T
F T T
T F F
T T T

If B entails C, then B can work out as false and so can C, since F->F works as a valid argument form. Consequently, the statement (A and C) can't get entailed by B->C alone, since if C=F, then (A and C)=F also.

 

[ThinkTank]

lurker's comment: I don't think there is any question - generally you have to refuse early demands at these levels. Even if you think you are safe because the demander is far away, he is far more likely than you to be able to buy your neighbor into an alliance.

I think you did not mean to say refuse here.

 

[ThinkTank]

Well first off that's not what I meant by consistent... usually consistent means that two statements do not contradict each other... they have the same truth value.

No, this is not what consistent usually means. We can have different uses of consistent, for sure, but read a textbook in logic for the definition of consistency, and you will find my definition.

Second, you've jumped out of classical logic into a modal logic of sorts by talking about "possibly true". We need another sort of analysis for validity here.

No, I have not. A row in a truth table can be referred to as a possibility, which is enough for this issue.

It means more than that in classical logic. It means that FF, FT, TT work out as possible truth values of B and C respectively with B first and C second (FF means B=F, and C=F).

No, it does not mean more than that in classical logic, it means just what I said it does and what you say it does, since they are just different ways of saying the same thing. Entailment quantifies over all rows in the truth table.

The truth table for implication or entailment goes

P Q P->Q
F F T
F T T
T F F
T T T

I know the truth table.

If B entails C, then B can work out as false and so can C, since F->F works as a valid argument form.

Correct.

Consequently, the statement (A and C) can't get entailed by B->C alone, since if C=F, then (A and C)=F also.

That (A and C) is not entailed by B->C is correct, except that I never claimed that.

I happen to have a Ph.D. in logic. I also don't believe in arguments from authority. But I do know the textbooks. Be in for a long discussion.

 

[Chamnix]

I think you did not mean to say refuse here.

You are absolutely right - corrected .

 

[TheOverseer714]

Doctor ThinkTank. Makes this lowly uneducated clerk feel humble in your presence... Not completely, but I respect the huge amount of work, time, money, passion and dedication getting even a Bachelors Degree requires. My smarts, if there are any, comes from the school of hard knocks and being an insatiable, omnivororous bookworm. I'll let you handle the logic, the stuff you and Doug are arguing about is way beyond me. What I do know is planting a town that shares its first 9 tiles close to an AI may or may not cause a war, but it is a wasted settler at any level above Monarch, since the AI can flip it so fast. We can plant a city there at our own folly. So settle tight, prepare for war and cave to any demand that doesn't cripple us. And don't build cities for the AI.

 

[vmxa]

Lurker:

All of this was based on Bamspeedy article that was originally done for the unpatched version of PTW. It also is done without any reported access to the code. It says nothing about planting a town on top of the AI.

What we do know is that they will not like it and often will declare and it very common at DG or better, in the early stages as you are not going to be strong compared to them at that stage.

Further attitude is not a true harbinger of war. They can declare while Gracious to you.

 

[Spoonwood]

ThinkTank,

No, this is not what consistent usually means. We can have different uses of consistent, for sure, but read a textbook in logic for the definition of consistency, and you will find my definition.

Copi's _Introduction to Logic_ p. 421 has an 11 line proof, which has line 7 as C and line 9 as ~C. It then reads "In this proof, the lines from 1 to 9 are devoted to making explicit the inconsistency that was implicitly contained in the premises. That inconsistency emerge in lines 7 and 9, which assert C and ~C, respectively. One this contradiction has been expressed..." So inconsistency and contradiction work as synonyms, according to Copi, and by language consistency and non-contradiction also stand as synonyms. If two statements have the same truth value, they don't contradict each other... in two-valued mathematical logic. Wikipedia indicates that informal Aristotelian logic works a bit differently. Still, by Copi's definition if A=F, B=F, then A and B work out as consistent.

Copi could have spoken in an unusal sort of way, so you could still have spoken correctly. Still, there exists at least one widespread textbook that uses the term 'consistency' in such a way.

No, I have not. A row in a truth table can be referred to as a possibility, which is enough for this issue.

You didn't say possibility, you said "possibily true".

That (A and C) is not entailed by B->C is correct, except that I never claimed that.

No, you didn't really. Your statement came as ambiguous enough though that it could read that way. So much the worse for me for not reading more charitably here.

What I do know is planting a town that shares its first 9 tiles close to an AI may or may not cause a war, but it is a wasted settler at any level above Monarch, since the AI can flip it so fast. We can plant a city there at our own folly.

Best damn reasoning I've seen all day. Well, maybe not flip it on Emperor if you build culture quickly enough... but still... it's a risk and your point in general stands.

VMXA,

What we do know is that they will not like it and often will declare and it very common at DG or better, in the early stages as you are not going to be strong compared to them at that stage.

I don't see how they dislike such (this doesn't say they like such either)... especially since it doesn't cause their attitude to change. I haven't seen anything that indicates they'll declare more often other than say-so. Such a town works as more vulnerable for sure, in many ways, so I do agree it usually doesn't work as a good idea. I've seen that article referenced by C3C players, and it appears that they changed anything on AI attitude since then... although I could speak incorrectly on this.

 

[ThinkTank]

Still, by Copi's definition if A=F, B=F, then A and B work out as consistent.

You have not put forward Copi's definition of consistency. I'd be interested in that. In the mean time, I'll tell you. Consistency is either defined as not implying a contradiction or as having a model. It doesn't matter much, since having either definition you can prove that it is equivalent to the other one. What having a model means depends on the semantical system you're using, but for our case it just means that there is one row in the truth table in which the sentences in question all come out as true. There is no way in which a row in the truth table where all sentences are false proves consistency.

One dramatic way to see that is to consider that A is of the form (P and not P), so it is itself a contradiction, and B is of the form (Q and not Q), also a contradiction.

Now, by your alleged correct use of "consistent":

Still, by Copi's definition if A=F, B=F, then A and B work out as consistent.

it would follow that actually the conjunction of 2 contradictions comes out as consistent. That is clearly absurd.

 

[Bucephalus]

Turn 1.

Warrior (who appears to be loitering without intent) despatched towards Scandinavia, to try to get some intelligence.

Curragh 01 spots another landmass to the NW; I think I may suspend ir's circumnavigation to investigate the possibility of contacts.

Turn 2.

It apears the other landmass is about three tiles of empty plains.

Turn 3.

To the North, Warrior01 meets the Iroquois; they are up BW and CB, and have an astounding six cities.

The Granary completes in Amsterdam, it's upkeep pushing us into negative gpt. If I can't find a way to balance the books, I will have to disband someone. That gpt payment for the 'Wheel' is crippling the economy.

Turn 4.

Portugal is either at war with the Vikings, or they are coming for us; there are a stack of 5 Warriors at the border.

Have to raise lux to prevent riots in Amsterdam; I now have to suspend research on Writing to avoid having to disband. Trouble is, when the Settler pops, gpt will drop further still until it Settles in four turns.

Turn 5.

Our Worker finally finishes mining the Silks; frankly, I don't understand why it was doing that when we have a pressing need for it to be roaded, and an undeveloped BG adjacent.

Portugal sends its Warriors over the Scandinavian border and into the fog of whatever city has settled near the Horses. If it auto-razes, it might be prudent to take advantage with oue imminent Settler; let's see what happens.
Rotterdam copletes it's worker; agonize over build - we need a Rax there, but we are also desperately short on Workers. Decide to go with the Woker.

Turn 6.

Down to 1g, -2gpt. The only thing to be done is to take a citizen from Amsterdam, and make him a taxman.

Turn 7.

Curragh01 spots sea tiles, and goes on a suicide run.

Portugese Warriors return in our direction now.

Turn 8.

Silks connected; Amsterdam completes Settler; Amsterdam is not yet capable of running as a settler factory, and we need the crowd control, so I set build to Archer.

Turn 9.

Curragh sinks.

Turn 10.

Den Haag is at last founded, solving the immediate unit support. issues.

 

[Bucephalus]

I'm having a few Internet connection problems this morning - I'll upload the save later.

 

[ThinkTank]

Yes, let's get on with the game! Our position remains difficult, regarding almost all aspects of the game, and the curragh sinking does not help. Looks like you did what had to be done, except maybe this:

Turn 5.

Our Worker finally finishes mining the Silks; frankly, I don't understand why it was doing that when we have a pressing need for it to be roaded, and an undeveloped BG adjacent.

There was this remark by me after doug's turnset, but I suppose that got lost in the side discussions, for which I am partly to blame.

In Amsterdam, we need 2 more shields for the settler factory to work, 1 at size 4 and 1 at size 5. This means the second wheat should be mined and the BG just S should be mined. Our worker is mining the silks right now, it should stop doing that and instead road it (to help with the money situation). This means at least 15 worker turns (2 mines and 1 road), so I think after the granary Amsterdam should build another worker before going on settlers.

 

[Bucephalus]

Damn...sorry, TT; you're right, I did allow myself to be distracted by the 'aggressive settling' red herring.

The Curragh was no big loss, really; it was due to meet the other in not too many turns, and I figured either we meet someone, or we save on unit support.

I suggest that we be not in too much of a hurry to play the next set - I think a cohesive strategy is required to ensure that we make the right choices.

 

[ThinkTank]

Damn.sorry, TT; you're right, I did allow myself to be distracted by the 'aggressive settling' red herring.

No problem, that happens to the best.

I suggest that we be not in too much of a hurry to play the next set - I think a cohesive strategy is required to ensure that we make the right choices.

Agree. I'll look at the save tomorrow.

 

[Spoonwood]

it would follow that actually the conjunction of 2 contradictions comes out as consistent. That is clearly absurd.

I do NOT think that the conjunction of 2 contradictions as mutually consistent an absurdity. In other words I deny your purported absurdum as actually absurd. In 2-valued logic, both contradictions have truth value of false. So, both contradictions imply each other. In some sense they fit together, since they logically imply each other. In a multiple-valued logic the statements (A and not A) and (B and not B) could both have truth value of .999999. In such a case even though both statements qualify as contradictions... they both at least seem consistent since they have equal truth values near 1.

Copi didn't have an explicit definiton of consistency. You can glean from the text given that Copi means inconsistency as any (derivable) instance of (A and not A). Consistency, comes as the antonym.

 

[darski]

I do NOT think that the conjunction of 2 contradictions as mutually consistent an absurdity. In other words I deny your purported absurdum as actually absurd. In 2-valued logic, both contradictions have truth value of false. So, both contradictions imply each other. In some sense they fit together, since they logically imply each other. In a multiple-valued logic the statements (A and not A) and (B and not B) could both have truth value of .999999. In such a case even though both statements qualify as contradictions... they both at least seem consistent since they have equal truth values near 1.

Copi didn't have an explicit definiton of consistency. You can glean from the text given that Copi means inconsistency as any (derivable) instance of (A and not A). Consistency, comes as the antonym.

lurker's comment: I think this belongs in OT discussions. Y'all have gotten way off the track of this game.

 

[DWetzel]

Player's comment: I'm agree with that.