Challenge-IV-10

neilmeister said:

Had to DoW Ramesses so he would like me enough to change civics for me.
Gave him Education to change to OR.

I had about ~90% of the votes already, so it wouldn;t have mattered where he put the missionary, as long as he did.

Click to expand...

Well Neilmeister... I honestly did not think it would be possible to even win a Deity religious victory as slowly as you did. And spreading TAP religion to civs in theocracy is really one of the skills essential to religious victories, I thought. But even knowing that, I managed to win even SLOWER than you did. 1830AD, to be precise.

This game was not as challenging as I had anticipated, actually. I did it on my second try... but I probably should have quit much earlier and did it on my third try instead of grinding it out. My first one I quit as soon as I lost the Oracle race... made me sad because I had such a beautiful start with food/commerce. Wasted time getting one worker tech, and lost the Oracle by 2t for it.

What went wrong for me in the second game happened very early, too. (1) I did not get any religion before Theocracy (by Oracle) and Christianity spread to my nearest neighbor Ramses as his first religion. Had to make Christianity TAP religion anyhow, because still had no other religion when AP was built by me. (2) I neglected espionage because all were already in favorite civics... this was the utterly stupidest thing, since I totally forgot I could change Ramses out of Christianity with espionage.

I was just Christianizing all my friends as much as possible, and took over the AP residency not before about 1400AD. (all my bhudism buddies). I still should have beaten you if I had remembered I could switch Ramses to Bhudism. Finally around 1750 I realized this (I was actuallty trying to switch Ramses to Free Religion... pretty desperate, I know!) when I saw option for my spies to change religion. First 3 spies failed but 4th did the trick. Now I'm the only candidate (as it should be) and even Ramses will vote for me.

Once I did that, I won, though it took two votes since Sury was at war with me at a high "pleased" rating when the first one happened. Thankfully, the UN stopped that war after he took one city. Next vote, even Sury casts his vote for me (I gave him 1000 gold to bump him to Friendly again). Viola... 1830AD victory, that should have been no later than 500AD if I didn't have my head up my rear end most of the game.

This one, along with the Noble space race game is one I will have to try again... its not a real big investment of time, at least.

I agree the keys to victory are:
(1) Oracle for Theology
(2) Common civics for +diplo
(3) common religion, too, if you want to win quickly.
(4) use espionage to change civics/religions as needed (you don't really need much tech... gifting resources or cash gets you the +4 fair trade just as easily). This is one of the very few types of games where using the espionage slider is highly recommended.
(5) Beeline alpha after Theo, trade up, but not Theo until spread TAP, if you can.
(6) beeline Optics after that. (I made some detours here, since Low Seas I thought galley access might be possible... nearly was, if I had settled an island and got a culture bridge it could have been, but I didn't... you take a big gamble thinking you can save turns without Optics).

Good luck, everyone... hopefully when I revisit this I'll see you all on the other side of 500AD.

Congrats kcd, this one was a very odd mixture of fun and extreme frustration.

I suppose I have indicated to everyone just how inexperienced I am at religous victories. The simple fact that you can spread religion, even when in Theocracy, was a complete revelation to me.

I still have no plans to play this again, so good luck beating my miserable date.

kcd_swede said:

I agree the keys to victory are:
(1) Oracle for Theology
(2) Common civics for +diplo
(3) common religion, too, if you want to win quickly.
(4) use espionage to change civics/religions as needed (you don't really need much tech... gifting resources or cash gets you the +4 fair trade just as easily). This is one of the very few types of games where using the espionage slider is highly recommended.
(5) Beeline alpha after Theo, trade up, but not Theo until spread TAP, if you can.
(6) beeline Optics after that. (I made some detours here, since Low Seas I thought galley access might be possible... nearly was, if I had settled an island and got a culture bridge it could have been, but I didn't... you take a big gamble thinking you can save turns without Optics).

Click to expand...

Allow me to add an alternative to points 1 and 6 above:

1) A Great Prophet can bulb Theology. This is helped by an one or more early Religions, completing one or more early Temples and hiring one or more Priests. Completing Stonehenge would be one way to generate Great Prophet Points without Religion spread. Since only 67 GPPs are needed at Quick speed to generate the first Great Person, it can be done quite quickly, certainly with respect to how long Optics would take in step 6. One caveat is one must complete both Meditation and Polytheism, before a Great Prophet can bulb Theology.

6) Depending on the specific Archipelago map and the locations of All AIs on that Map, it is possible that TAP Religion Missionaries can be transported to All AIs via Galleys. Also, as kcd_swede mentioned, a Galley can cross a Sea Channel larger than it normally can cross by settling a City on the Coast, generating Culture and popping to Cultural Radius 2 (BFC) or even more. A Galley should be able to navigate in Ocean dominated by your Culture. I believe that such use of Culture will not be required on a significant number of Archipelago. You will of course need to select Low Sea Level for the "Galley" option to be at all viable. Also, as kcd_swede mentioned/implied, a Sea Channel may be too big for Culture and one may need to resort to Optics and Caravels.

Sun Tzu Wu

Well, I tried this a few more times. That's a total of 8 times I've tried this, 3 of which are victories submitted (five of which are Oracle BIFAL before t40).

My 1700AD game, which was an improvement over my 1830AD game in terms of finish time, did not improve my standing in the challenge. The problem I had here was that I was too casual about spreading TAP. I was of course spreading to one city of each civ as fast as possible, but failed to identify my target voters early enough to spam their cities for vote padding. Anyhow, one civ was not located until fairly late, due to my own failure to fully explore the east coastline of Sury's landmass. Had several votes come up before I could win.

My 1490 game I corrected most of my errors from the previous two victories. I actually had a map which was galley accessible. But it was 680AD before a galley could reach one of the two far away civs because of the very long tortuous path there (discovered by my first workboat that started exploring right after the AP was built). If it takes 28 turns to get somewhere, you haven't really won anything by getting a galley there, since the second far away civ was reached by caravel at about the same time that galley arrived!

Unfortunately for me, that last civ was Hatty, and she self-spammed her cities with TAP religion, in a span of like 5 turns. That meant I not only had to get several civs to vote for me, I had to get them all good and infected in the majority of their cities. This took a lot of time, even though they were close. I got Ghandi, Ramses, Sury, and Pacal to vote for me. Gilga had converted to TAP religion and was my opponent, but he just had the one christian city and wouldn't have voted for me nor garnered any votes for himself anyhow. But it did double his vote total -- to 20.

Hatty had 126 votes, and abstained. (563 votes needed, 584 recieved). Getting her happy was going to be impossible if I had to keep her numerous enemies happy too. After placing my vote while in bhudism to get Sury/Ramses happy, I converted to TAP religion to double my votes. Just barely won. Ghandi (free religion, spy to switch him to bhudism failed) voted for me at +7 pleased, and Pacal voted for me at +6 pleased. If I failed to double or had bad hidden modifiers instead of the good ones... I'd still be chasing a finish that beats the previous #1 finish. As it stands, I'm not in last place on this one anymore, so I'll give it a rest for a while and start working on that Noble space race.

My second thought on this game is that anything before 1000AD will require a combination of superb gameplay and a bit of good luck. But STW has always amazed me on these kind of games, so I'm looking forward to having him blow my mind when he gets around to showing how its really done.

Non-Qualifying Game
My first attempt at this one was won at 560 AD. I didn't realize that goody huts were not allowed, so this game was accepted, but did not qualify for the Challenge series. I want to share my experience with this game because whoever draws a map like mine can win this challenge.

I missed the Oracle by a few turns but knew I couldn't pass up the chance to research Theology the old-fashioned way. Why do this? I had the fortune of drawing a Pangaea-like archipelago map. I had met land units from all of my opponents. I am sure there is a low likelihood of the happening, but it did. I knew I had to do whatever was in my power to win this one since no settlers or galleys would be needed. I ended up fumbling my way to a 560 AD victory on the Pangaea map. I won on the third vote.

Recent Submission
I have submitted a 940 AD game. This could have been won around 680-720 AD if I have won on my first vote. I had four opponents on "my island", and two others on islands. One was to the west and one to the east. I had a large "inland" ocean off to my capital's east, but was able to get a galley through an opponents city on a one tile land bridge to get access to my far east opponent. The ocean off to the east was contained by polar ice caps to the north, so galley's could not get around. My capital had non-irrigated corn and three gold. My capitol only got over population 4 for a short time, well after the Oracle. I was fortunate enough to found Confuciounism as well, but this turned out to not be necessary since my one neighbor that adopted Christianity switched to another religion. Of course, had I built the AP in Christianity, he may have stuck with it and caused me problems.


Keys to Optimizing a finish for this Challenge game.
1. Choose the right opponents.
2. Choose a start location with at least 1 farm resources or 2 FPs and 2+ commerce resources.
3. Aggressive AI to slow Oracle building.
4. Play Pangaea-like Archipelago map (if you have the patience to play a zillion maps).
4B. Snaky island, low seas to maximize opponents on land mass.
5. Skip bronze working until after Oracle is built. It slows research down too much and allows AI to beat you to Oracle too often. (With luck and riverside gems, you can probably get BW in and still get the Oracle.)
6. If skipping BW, make sure you have a few non-forested hills.
7. Start with a coastal capitol.
8. The first build after the Oracle should be a Settler, who should be settled on the opposite coast (east vs. west) as the capitol. This is to avoid polar ice caps limiting galleys.
9. Get at least one explorer work boat out of your capitol after the setter.
10. Limit spreading your AP religion to your nearby neighbors until they have a religion, or spread prioritize spreading a non-AP religion to them ASAP.

Sun Tzu Wu said:

Opponent Selection:

Select 6 Peaceful Opponents, especially Mansa Musa. Put all Espionage on Mansa Musa, so you will eventually know what he is researching, so you can research something else. When he completes something, be sure you have something to trade for it, like Theology. Your other Opponents should probably include those that have a hidden +1 to +2 Diplomacy bonus of liking any Civ, including the Player. These opponents include Zara Yaqob (+2), Gandhi (+2), Hatshepsut (+1), Frederick (+1), Lincoln (+1). (I personally have had some bad luck with Gandhi, but he's probably still a good choice).

Sun Tzu Wu

Click to expand...

I believe STW may have taken this information from this Spreadsheet that many players use:



I believe there is an error. I believe the below is correct.
I believe Gandhi has only a +1 base attitude modifier.
Mansa Musa has a +2 base attitude modifier.

I took screenshots from Bat 3.0. It agrees with me. Note the (creative) attitudes: "+1 Your small civilization is no threat to us", "+1 developing nations should work together to catch up" and "+1 a first impression is a lasting one."

Sun Tzu Wu said:

Select 6 Peaceful Opponents, especially Mansa Musa. Put all Espionage on Mansa Musa, so you will eventually know what he is researching, so you can research something else. When he completes something, be sure you have something to trade for it, like Theology. Your other Opponents should probably include those that have a hidden +1 to +2 Diplomacy bonus of liking any Civ, including the Player. These opponents include Zara Yaqob (+2), Gandhi (+2), Hatshepsut (+1), Frederick (+1), Lincoln (+1). (I personally have had some bad luck with Gandhi, but he's probably still a good choice).

I would include Ramesses II (0) ...

Click to expand...

shulec said:

I believe STW may have taken this information from this Spreadsheet that many players use:

Spoiler :

Click to expand...

I do not rely on this particular chart, but I have created my own table based on this publicly available spreadsheet (see link below) that has the XML values in a spreadsheet format which is much easier to read than the raw XML files:

AI Leaders Personality (Vanilla, Warlords, BtS)

shulec said:

I believe there is an error. I believe the below is correct.
I believe Gandhi has only a +1 base attitude modifier.
Mansa Musa has a +2 base attitude modifier.

Click to expand...

I disagree. The iBaseAttitude value for Gandhi is 2 and for Mansa Musa the value is 1.

shulec said:

I took screenshots from Bat 3.0. It agrees with me. Note the (creative) attitudes: "+1 Your small civilization is no threat to us", "+1 developing nations should work together to catch up" and "+1 a first impression is a lasting one."

Spoiler :

Click to expand...

Again, I disagree. You are not seeing the iBaseAttitude value in these two screen shots; you are seeing something else. Which of the two attitude modifiers for Mansa Musa is iBaseAttitude? Neither of these two, in opinion, because both of them have value 1 and neither has value 2. What you are seeing are unrelated hidden attitude modifiers.

I have just doubled checked the BtS XML file where the iBaseAttitude values are stored:

<C:\Program Files\Firaxis Games\Sid Meier's Civilization 4\Beyond the Sword\Assets\XML\Civilizations\CIV4LeaderHeadInfos.xml>

It has the iBaseAttitude values that I stated early in this thread (See quote at the top of my post).

Whenever I suspect that some "spreadsheet" value I've been using is incorrect, I check with the XML source file that contains the suspect value to verify its correctness.

Of course these XML values are useless without an understanding of how they are used:

AI Attitude Explained

Sun Tzu Wu

Sun Tzu Wu said:

I do not rely on this particular chart, but I have created my own table based on this publicly available spreadsheet (see link below) that has the XML values in a spreadsheet format which is much easier to read than the raw XML files:

AI Leaders Personality (Vanilla, Warlords, BtS)



I disagree. The iBaseAttitude value for Gandhi is 2 and for Mansa Musa the value is 1.



Again, I disagree. You are not seeing the iBaseAttitude value in these two screen shots; you are seeing something else. Which of the two attitude modifiers for Mansa Musa is iBaseAttitude? Neither of these two, in opinion, because both of them have value 1 and neither has value 2. What you are seeing are unrelated hidden attitude modifiers.

I have just doubled checked the BtS XML file where the iBaseAttitude values are stored:

<C:\Program Files\Firaxis Games\Sid Meier's Civilization 4\Beyond the Sword\Assets\XML\Civilizations\CIV4LeaderHeadInfos.xml>

It has the iBaseAttitude values that I stated early in this thread (See quote at the top of my post).

Whenever I suspect that some "spreadsheet" value I've been using is incorrect, I check with the XML source file that contains the suspect value to verify its correctness.

Of course these XML values are useless without an understanding of how they are used:

AI Attitude Explained

Sun Tzu Wu

Click to expand...

I am glad to have your input on this. I am going to do some game tests to convince me one way or another. I have to admit that I don't know or understand the first thing about the XML data and these things.

Until I convince myself of this or that, I will ask you to agree to disagree. Now I apologize what I said in my previous post is misstating things. But this is my assertion:

If Gandhi has a net of +6 attitude toward me, he will not vote for me in an AP diplo victory, but Mansa will vote for me at +6.

I plan to prove or disprove this based on game play. I will then report back.

In the event that a holdout AI is in Theocracy and you don't want them to use a gifted missionary in a large city, can't you build a growth inhibited city (on tundra, for example), spread the TAP religion to it and then gift the city to the holdout AI? Or does that not work on Deity?

thbrown81 said:

In the event that a holdout AI is in Theocracy and you don't want them to use a gifted missionary in a large city, can't you build a growth inhibited city (on tundra, for example), spread the TAP religion to it and then gift the city to the holdout AI? Or does that not work on Deity?

Click to expand...

This can definitely be done. But... I find that if I get to a civ and they are in theocracy, it takes too much time to get a settler built and to a location to build a city for gifting. I think this gambit is a great approach to use if you have an opponent like Tokugawa who will not open borders. This circumvents the need for open borders as well it will give you the +4 fair trade bonus.

shulec said:

I plan to prove or disprove this based on game play. I will then report back.

Click to expand...

I have run some tests and I am now more confused than ever.
Zara had games where he had +6 and voted for and against me without predictability.
Gandhi never voted for me at +6
Mansa never voted for me at +6.

In one game, engineered to have a AP religious victory unrealistically early, I couldn't get votes at +7.

I guess I need to study the post that STW referenced on AI Attitudes.

shulec said:

I have run some tests and I am now more confused than ever.
Zara had games where he had +6 and voted for and against me without predictability.
Gandhi never voted for me at +6
Mansa never voted for me at +6.

In one game, engineered to have a AP religious victory unrealistically early, I couldn't get votes at +7.

I guess I need to study the post that STW referenced on AI Attitudes.

Click to expand...

As I understand it (not perfectly, to be sure, but fwiw...) there is the FIXED hidden modifier for each AI, plus a hidden modifier randomly assigned to each AI at the start of a game (which is fixed, but only for that particular game). So Mansa and Ghandi might have different thresholds in different games. Note, you can't do anything (legal) about the random hidden modifier... the best you can do is select the AI with the most advantageous constant hidden modifier and hope for the best.

Yes, kcd is right. There is a random modifier rolled for each AI in every game.

neilmeister said:

Yes, kcd is right. There is a random modifier rolled for each AI in every game.

Click to expand...

I love that I can learn something new with this game every day! Thanks for setting me straight guys!

shulec said:

If Gandhi has a net of +6 attitude toward me, he will not vote for me in an AP diplo victory, but Mansa will vote for me at +6.

Click to expand...

Remember that these are the sums of visible attitude factors only. You need to add in all four hidden attitude factors to see whether you get the +8 required by an AI to vote for your RL DV.

Please look at the Attitude Explained link I provided earlier and again below. Until you begin to understand the inner workings of the AI Attitude system you are truly only guessing, but there are only three other hidden attitude factors in addition to the AI's hidden iBaseAttitude toward the Player as explained in this link:

http://forums.civfanatics.com/showthread.php?t=204328

I have probably analyzed this thread more than most, yet I do not claim to completely understand it. I suggest you focus on only the four hidden attitude adjustments, since the F4 table will provide the rest. In the case of Gandhi, I believe his settings for these three other hidden attitude adjustments often net -1, so 6 + 2 - 1 = 7 and that's not enough for him to vote for your RL DV at visible +6 Diplomacy.

In my first post on this subject in this thread I stated that I did not like using Gandhi as an Opponent, because I didn't have that much luck with him despite his +2 iBaseAttitude value.

You can believe what you want about AI Attitude, but I do believe that its all explained in the thread I've quoted for anyone with the motivation and analytical skills necessary to understand it. However, I also understand that some players wish to preserve the mystic of Civ 4 BtS and avoid all XML and source code reading to understand the game better as that's not the way the developers would want Players to play the Game. On the only hand, I have limited hours in which I'm free to play Civ 4 BtS. I can't afford to play hundreds of Games to understand the nature of the AI opponents, so I feel no guilt (nor should I) whatsoever in using the threads that explain it for me such as the "AI Attitude Explained" link above.

Sun Tzu Wu

kcd_swede said:

As I understand it (not perfectly, to be sure, but fwiw...) there is the FIXED hidden modifier for each AI, plus a hidden modifier randomly assigned to each AI at the start of a game (which is fixed, but only for that particular game). So Mansa and Ghandi might have different thresholds in different games. Note, you can't do anything (legal) about the random hidden modifier... the best you can do is select the AI with the most advantageous constant hidden modifier and hope for the best.

Click to expand...

neilmeister said:

Yes, kcd is right. There is a random modifier rolled for each AI in every game.

Click to expand...

AI Peaceweight:

There is a -3 to +3 random Peaceweight modifier that is added to the Peaceweight of each AI as defined in the XML, but it is not directly added to AI Attitude. It does have strong indirect effects on AI Attitude of course.

AI Attitude Explained:

For AI Attitude per se there are three other hidden Attitude factors in addition to iBaseAttitude, but they are definitely not random, though they definitely appear to be random. The "AI Attitude Explained" thread explains how they are computed.

A "Random" Simple Model:

I suppose you could think of these three other hidden Attitude factors as random, though they really are deterministically computed from relevant values in the game. Statistically, I believe these three values can be -1, 0 and +1 and added together generate a bell curve of -3 to +3 where -3 and +3 are somewhat rare, -2 and +2 are uncommon and -1, and +1 are common and 0 occurs more often than any of the others. If this characterization of these three hidden factors as a random bell curve helps you play the game better, I'm all for it.

The Real Story: Obfuscated Source Code:

However, the thread above does define the true nature of these three hidden attitude factors and that nature is an ugly computation. Anyone who doesn't like obfuscated source code should run in horror; I'll lead the way out; I dislike reading obfuscated source code. I put a hex on ... for your reminding me of it.

Sun Tzu Wu

I suppose you could think of these three other hidden Attitude factors as random, though they really are deterministically computed from relevant values in the game. Statistically, I believe these three values can be -1, 0 and +1 and added together generate a bell curve of -3 to +3 where -3 and +3 are somewhat rare, -2 and +2 are uncommon and -1, and +1 are common and 0 occurs more often than any of the others. If this characterization of these three hidden factors as a random bell curve helps you play the game better, I all for it.

Click to expand...

Thanks... that is helpful. It sounds logical, too.