SGOTM 15 - Unusual Suspects

[kcd_swede]

I made a wb save with the unfogged tiles only being used. Made sure that "New Random Seed on Reload" is checked, but move warrior different each time just in case. I used all random leaders and civs, and shuffle map.

Played to hinduism founding 7 times; 5 times with agri gambit, and twice without going after Polytheism at all to see when the AI get it. In the first five cases I open wb after the test to see who has teched what.

1) bhudism gone 3400bc (t15), agri/poly hindu 3160bc (ibt t21-22) (1AI has meditation)
2) bhudism gone 3440, agri/poly hindu 3160bc (3 AI have Meditation)
3) same as #1
4) same as #2
5) bhudism gone 3440bc, agri/poly hindu 3160 (2 AI have Meditation)

6) bhudism gone 3440bc; hindu gone 3320bc (teched agri/masonry)
7) bhudism gone 3440bc; hindu gone 3040 (teched agri/AH/poly).

It would appear that if we take Poly as second tech then the AI are less likely to try for Poly. At least in this test. Note, that's 5 of 5 founding hindu with poly as second tech!

More tests with different maps/opponents/civ combos would help us see if this is enough of an indication that the gambit is worth the risk. Its a very big advantage over teams that go poly first, if it works.

View attachment 311405

 

[kcd_swede]

Tried two more maps built from scratch.

Hemispheres: impossible to found Hindu... even if you go poly first!

Pangea:
1) 3400bc bhuda/ agri+hindu 3160bc PASS
2) 3440bc bhuda/ 1t poly, then agri+hindu 3160bc PASS
3) 3720bc bhuda/ agri/masonry hindu founded by AI 3400bc test
4) 3280bc bhuda/ 3360bc AI founds hindu teched agri/poly FAIL

Each case of course had 8 different random leaders/civs.

This is very interesting. If an AI goes for Poly first, we lose the game NO MATTER WHAT WE DO. If it is possible to get Poly as first tech, in a randomized game it seems likely that we would also get it as a second tech.

There was two of my 3 test maps that had an agri gambit fail, and in those fails we would have failed with poly first, too!

Does anyone think Neil left it random as to whether founding hindu will work or not? If we could found it as our first tech, its looking more and more likely we could also found it as our second tech.

But then again, Neil surely did not randomize our leaders/civs, and could have manipulated it so Poly first works but poly second won't. I doubt it though... that's asking a lot of the AI.

Unfortunately, the code-divers on other teams might already know exactly how to manipulate AI so that they won't be guessing.

Oh well... enough for me for today.

 

[AAA]

Tried two more maps built from scratch.

Hemispheres: impossible to found Hindu... even if you go poly first!

Pangea:
1) 3400bc bhuda/ agri+hindu 3160bc PASS
2) 3440bc bhuda/ 1t poly, then agri+hindu 3160bc PASS
3) 3720bc bhuda/ agri/masonry hindu founded by AI 3400bc test
4) 3280bc bhuda/ 3360bc AI founds hindu teched agri/poly FAIL

Each case of course had 8 different random leaders/civs.

This is very interesting. If an AI goes for Poly first, we lose the game NO MATTER WHAT WE DO. If it is possible to get Poly as first tech, in a randomized game it seems likely that we would also get it as a second tech.

There was two of my 3 test maps that had an agri gambit fail, and in those fails we would have failed with poly first, too!

Does anyone think Neil left it random as to whether founding hindu will work or not? If we could found it as our first tech, its looking more and more likely we could also found it as our second tech.

But then again, Neil surely did not randomize our leaders/civs, and could have manipulated it so Poly first works but poly second won't. I doubt it though... that's asking a lot of the AI.

Unfortunately, the code-divers on other teams might already know exactly how to manipulate AI so that they won't be guessing.

Oh well... enough for me for today.

Nice work. The muddiness continues

 

[ChrisShaffer]

Are you working a 1C commerce tile in your tests?

I wonder if Neil nerfed the AI BFC for those starting with mysticism, do they have zero commerce tiles in their inner ring?

 

[Xcalibrator]

The best way to do so is as Xcal says...

That would be a first!

The earliest Wonders are all stone, and I think (?) we'll have plenty of time to quarry the marble in time for the Oracle. And if we don't actually want Henge and/or Great Wall, we can partially build them for failure gold with a 2:1 hammer:gold payoff--not critical in the very early game when we can run 100% research, but it would give us a leg up after founding another city or two, and with No Tech Brokering it's important to be first to a tech. And an argument in favor of the Great Wall is that we could settle a Great Spy, letting us keep tabs on the research of a couple AIs so we know exactly when to trade techs to them.

Nice testing, kcd! If only we knew what it means.

Edit: Lack of commerce at the start doesn't matter much because the capital's commerce (8 cpt?) dominates.

 

[ChrisShaffer]

I think I agree with whoever said we don't want Stonehenge or The Great Wall. We will have early religion, making Stonehenge less useful. And I have very bad memories of our Wall experience last game. And, of course, with 8 AI on the map, unless the map is huge, it's unlikely that barbarians will be much of a problem.

 

[High King J.J.]

I second moving the warrior. Don't ask where.
I second founding a city. Don't ask on which PH.
I second researching polytheism. Don't ask when.
I second early rushing if necessary. Don't ask me to play when rushing.

If others agree, I don't mind playing recklessly. I finally envision a useful role for myself: if we take risks and lose, you can blame me for our demise. If we take risks and it pays off, you can take the credit for brilliant playing.

Spoiler :

Just my luck: for the first time on this decade I have something to do in RL. And it's when I take part in a SGOTM.
Maybe quit work to play? Maybe not. Tried that with Civ3, still couldn't finish many Gotms (but that was the fault of Civ3. With Civ4 I actually finish a game every now and then.)
Too tired to make sense. Must

PS. Maybe settle on marble. Build pyramids on second city built on stone, use GEngineers to rush wonders. Be the envy of the world. Be destroyed when other civs all attack because we have built all useful wonders and neglected military. Get more GGenerals due to Gwall. Get to blame an insane team member for this idiocy.
Maybe I'd better now.

Spoiler :

Good (k)night.

 

[SteelHorse]

Lets assume neilmeister took bad luck out of the equation for all teams that SIP and beeline Poly. (Otherwise there would be serious griping from those teams that did that and still lost).
The two most obvious ways to do would be:
1. Start none of the AIs with Myst, or
2. Start none of the AIs with access to more than 1 single gold tile.

So the question is what is the likelihood that we can research Agri + Poly under the options of SIP or Settle in 1 turn and build worker or grow onto more tiles during the time it take for the AI to get Poly in case 1 and case 2.

Given that Myst is cheaper than Agri we always will lose going Agri->Poly v AI going Myst->Poly all other things equal even ignoring 1 turn to settle delay.

With 8 AI in the game the odds intuitively feel high (bad for us) that at least one of them will go Myst->Poly. And given our plan to spend 1 additional turn to settle on marble (or stone), my gut tells me to go Poly first. I am all for taking good risk, but not sure this is a good risk to take.

 

[AAA]

Lets assume neilmeister took bad luck out of the equation for all teams that SIP and beeline Poly. (Otherwise there would be serious griping from those teams that did that and still lost).
The two most obvious ways to do would be:
1. Start none of the AIs with Myst, or
2. Start none of the AIs with access to more than 1 single gold tile.

So the question is what is the likelihood that we can research Agri + Poly under the options of SIP or Settle in 1 turn and build worker or grow onto more tiles during the time it take for the AI to get Poly in case 1 and case 2.

Given that Myst is cheaper than Agri we always will lose going Agri->Poly v AI going Myst->Poly all other things equal even ignoring 1 turn to settle delay.

With 8 AI in the game the odds intuitively feel high (bad for us) that at least one of them will go Myst->Poly. And given our plan to spend 1 additional turn to settle on marble (or stone), my gut tells me to go Poly first. I am all for taking good risk, but not sure this is a good risk to take.

I agree. If I had to guess, my guess would be that neilmeister has included a number of religious leaders (to make the must stay hindu significant) but has stripped them of Myst. But they are likely to go for religions ASAP so we should grab Poly first.

 

[Deckhand]

It is likely that we are isolated (or semi-isolated). When we finally meet the other civ's they will be entrenched in buddhism (or whatever) and not impressed with our hinduism. Spreading hinduism to local civ's won't work if a) there aren't any or b) we rush them.
What does this speculation mean? Religious victory won't be easy. We will need astronomy.
The No Tech Brokering could be there to help us: if we are isolated, then we can't trade with anyone. The NTB keeps the AIs from running away from us tech wise.
I wonder how many techs and cash we will be able to get for optics?


Steelhorse, good analysis that myst+poly is faster than agri+poly. Not really consistent with kcd's test results. We shouldn't trust him anyway.

 

[kcd_swede]

We shouldn't trust him anyway.

That's right... if I had made the map, you would have at least 50% chance of suffering conquest defeat at the hands of a mobile artillery while you are researching polytheism.

I need to look more at the test games I played. It is not always a Myst civ that gets the early religions, nor is it always the high religion flavor leaders that do so. It seems to be more determined by what resources they have nearby, and what worker techs they lack to be able to develop them. THATS the only correlation I can see so far as to whether an AI will beeline an early religion. But in every game at least one civ will always aim for a religion first.

If there is more than one civ that already has worker techs to keep its starting workers busy, then one of those civs may aim for Poly while the others aim for Med. Depends whether they have a 2comm tile or not whether Hindu goes first or Bhuda goes first.

Does any of this really matter? Probably not. We have no (permissable) way of knowing what Neil did. We can guess. Guess wrong, and we play for Wooden Spoon Avoidance instead of Laurels. Guess right, and we still have to play well to beat some very good teams.

I'm sure Neil tested whatever he did many many times. In fact, the save has not been released yet. He could still adjust the saves based on team discussions so far and what folks are planning to do. That's power, baby.

 

[ChrisShaffer]

Guess wrong and we quit and watch the other teams...

 

[kcd_swede]

Guess wrong and we quit and watch the other teams...

While it might seem tempting to give up... I think we'd play it through fo rthe excercise and to get a feel for what other teams faced. Beating them in finish date and/or score would still mean enough for me to keep going, anyhow.

I've been reading through the discussion thread again for clues to what Neil is thinking. Mylene brought up the agri gambit there, so we aren't the only ones who have thought about it.

Neil has said we shouldn't faff about in founding Hindu... whatever that means... don't tell me if its inappropriate for a family forum. He said you "probably" want to get Poly as soon as possible. He also jokingly said that SIP and Poly first gives us "at least 50%" chance of getting Hinduism.

Seems kind of weird to have an "all-or-nothing" choice from a blind position right at the start.

I would say that it would have to be almost unanimous vote of team members to do the gambit. I doubt we'd get that... so maybe its time to start planning the best game we can get going Poly first. At least we all get to play a whole game that way.

I think warrior first, switch to worker at size 2, put a couple turns into a mine while waiting for agri to finish, is the way to go.

I'd suggest to move the warrior 1NE, and the settler to the marble hill on turn zero. If we see an oasis or riverside gems in the BFC (2+ commerce tile), we could reconsider the gambit for another 70 posts before deciding and settling t1.

If settling t1 won't get us poly, then settling t0 won't either. Who knows... maybe thats the right guess, and we defeat some of our more aggressive competition in the first 20 turns.

 

[ChrisShaffer]

If we settle on either marble or stone, we can build two warriors prior to pop2, which would give us much better scouting than we've gotten the last two games.

I suggest moving the warrior 1NE and posting screenshots. If nothing appears special, then move the settler to the marble hill. Someone already posted that the marble has better rivers, which I agree, and we seem to have decided that we will have a second city on/near the stone soon enough. So, if warrior 1NE doesn't reveal anything special, then yes, settler to marble.

 

[Xcalibrator]

Agreed. But let's not get too committed to settling on the marble. Yes, it has more river tiles visible, but the stone has 3 (not counting the stone tile itself) and (I think, from drawing things out) at least 5 more that we can't see, while the marble has at least 4 more--plenty of river tiles either way. And if we want to build the Henge or Great Wall (probably only for bonus-boosted failure gold), settling on the stone will let us start work on them sooner, and there's more urgency on those than for the Oracle. By the time we're ready to work on the Oracle we'll almost certainly have settled a 2nd city, and can quarry or settle on the marble.

Edit: And if we can trust the original game-announcement screenshot, there's more land to our north than our south, so a capital on the stone will be better centered in terms of distance maintenance.

 

[ChrisShaffer]

And if we can trust the original game-announcement screenshot, there's more land to our north than our south, so a capital on the stone will be better centered in terms of distance maintenance.

I think if we can trust the screenshot, there is "some land" to our north, not necessarily "more land." We could, for example, be in the exact center of a pangea.

 

[Rusten]

I don't agree with going for stone and GW/early fail gold.

Fail gold: Superb a little into the game, and needs to be utilised a lot this game, but at the start you want to expand to all the good spots. It's not worth it for the first turns, hammers are too precious. You want some cities first, and by then we will have stone anyway.

Completing SH: Not sure how I feel about this yet, but if the land is good early expansion should beat it easily. Not a big fan of the prophet GP [note: something like the Oracle is late enough so that we can get a GS 1st. SH would be built so early that I think GProphet will pop 1st).

Completing GW: Bad IMO. We will get a great spy who's not going to be useful. It does ease expansion a little, but nothing good empire management can't handle, and the GSpy is a big problem. EP is not worth it against emperor AIs.

 

[Trystero]

Accessing Stone (or Marble) for wonder-building also requires Masonry, even if we settle on top of it. In my experience Stone is never connected fast enough to be much benefit in getting SH at higher levels. It's much more useful if we were going to build the Pyramids, but those usually don't go too early. Given that we already have to research Polytheism early, I don't think also prioritizing Masonry just for the hammer bonus to SH would be a good use of our beakers.

I'm currently in favor of trying to get a GPro for the Hinduism shrine, and maybe another to bulb Theology if we want the AP. But I agree a shrine's usefulness would be diminished if we were isolated.

 

[SteelHorse]

Agree with Rusten on Stone Wonders, and Trystero on Hindu Shrine and AP.

Unless we are isolated, which I am not convinced we are, 8 AI means lots of early warring to get space/cities. This may turn into another Dom/Conquest fight, but neilmeister has seemed to set the rules that would eliminate the AP cheese condition, but allow a ligit AP victory.

One thing that, is not to difficult to do with these conditions is to snap up a LOT of religions. In the test games, other that Bud, it was not hard to get most of the rest. Perhaps we should also keep cultural in mind, especially if we are isolated.

In summary, path could be:
Isolated: Cultural
Pangea/Single Continent Map: Conquest/Dom
Multiple Continents: AP

Figuring this out early will be important.

And of course, most importantly, warrior NE.

 

[Deckhand]

And of course, most importantly, warrior NE.

OK, now I can sleep.