SGOTM 22 - Anti-TSR

[Pollina]

If Pollina is right about Z/G already having built a settler, then barbs might start spawning on T25 or whatever and invade as soon as they arrive at our borders. That would even preclude earliest settler using the 4 warriors as garrisons.

Did I write that? If so, then I probably meant that they built a worker. It is certainly possible that some AIs are already building a settler, especially if they started with a pre-improved food resource and no stone/marble. Basically, the higher our odds for GW are (AI without stone), the earlier we can expect a barb invasion.

if (getElapsedGameTurns() < ((GC.getHandicapInfo(getHandicapType()).getBarbarianCreationTurnsElapsed() * GC.getGameSpeedInfo(getGameSpeedType()).getBarbPercent()) / 100))
{
bAnimals = true;
}

Confirms that human barb won't spawn before t25.

 

[LowtherCastle]

Did I write that? If so, then I probably meant that they built a worker. It is certainly possible that some AIs are already building a settler, especially if they started with a pre-improved food resource and no stone/marble.

I think you're right. It was worker, not settler. 10hpt was not happening. It was 6hpt.

Basically, the higher our odds for GW are (AI without stone), the earlier we can expect a barb invasion.

THat's what I'm thinking.

 

[WastinTime]

The more I think about the mapmaker's intentions the more I think that GW may be available after all. BSP is not exactly known as an evil mapmaker who sets up traps, nor is he one to restrict our options. More like the opposite, actually. He seems more interested in providing many viable options and making it hard to pick the best one. Luring us towards GW with Raging Barbs and No Tech Trading and then denying us access to it (well after we have committed to that path), is not really a typical BSP move. It's more like something we would do in his place. If his intention was to keep us from GW, he probably would have given us a more obvious hint (like he did with The Oracle), or made sure it is built well before any team would commit to the GW path.

+1
+1

 

[BSPollux]

GET OUT OF MY HEAD!
STOP TRYING TO READ MY MIND!

Spoiler :

 

[Mitchum]

The more I think about the mapmaker's intentions the more I think that GW may be available after all. BSP is not exactly known as an evil mapmaker who sets up traps, nor is he one to restrict our options. More like the opposite, actually. He seems more interested in providing many viable options and making it hard to pick the best one. Luring us towards GW with Raging Barbs and No Tech Trading and then denying us access to it (well after we have committed to that path), is not really a typical BSP move. It's more like something we would do in his place. If his intention was to keep us from GW, he probably would have given us a more obvious hint (like he did with The Oracle), or made sure it is built well before any team would commit to the GW path.

That's too funny. The more I think about it, the more I think BSP did NOT want us to get the GW. The earliest it can be had is T41 or there about, right? So in order to make it fair, BSB would have to ensure that either a) the GW always goes later than T41 or b) the GW always go much sooner than T41. That way, either all teams can get it if they try or no teams can get it if they try.

If you're with me so far, by giving the AI Masonry, isn't he tipping his hand a bit? If he wanted us to have a chance at it, I don't think he would have given them Masonry, right? So my best guess is that he's done everything he can to make sure that the GW gets built way before T41.

 

[WastinTime]

Most teams won't know AIs have Masonry and they'll think they have plenty of time to get GW. They might even go Archery->Masonry and still get GW!

If the AI doesn't have stone, then it's not likely GW would go before 41.

If anything, giving them Masonry was to trip up teams like us that think we're so smart. We're falling into BSP's trap for teams that over-analyze the data.

If we take the small detour towards GW and fail, it's nothing to be ashamed of.
But if we talk ourselves into being the only team without GW, we will be the butt of many jokes.

In my test games, sure the stone-enhanced AI often get GW around T35, but more often, even with stone, I see henge and/or GW going T49-T51. That would allow teams to go
Archery->Masonry. Rough calculation: you would get GW T43 with Archery first.

 

[WastinTime]

fyi, no team has played past T12, so we're not behind.

xteam hasn't even started. Maybe they're debating skipping AH and starting with Masonry!

 

[Pollina]

GET OUT OF MY HEAD!
STOP TRYING TO READ MY MIND!

NEVER!!

That's too funny. The more I think about it, the more I think BSP did NOT want us to get the GW. The earliest it can be had is T41 or there about, right? So in order to make it fair, BSB would have to ensure that either a) the GW always goes later than T41 or b) the GW always go much sooner than T41. That way, either all teams can get it if they try or no teams can get it if they try.

If you're with me so far, by giving the AI Masonry, isn't he tipping his hand a bit? If he wanted us to have a chance at it, I don't think he would have given them Masonry, right? So my best guess is that he's done everything he can to make sure that the GW gets built way before T41.

I had hoped that BSP wouldn't give us access to GW, because I always found RB with GW kinda cheesy. I fully expected him to pre-settle the AI's capitals and give them the wonders he didn&#8217;t want us to have. Or have the AI start with GEngineers, so that all ancient wonders would go on t1. He didn&#8217;t do any of that, obviously. He only gave us a hint that the AI started with PH, which pretty much implies that The Oracle is unavailable. He didn&#8217;t tell us that the AI has Masonry. Wouldn&#8217;t he have made it a lot more obvious, if his intention was to keep us from getting GW, too? Not doing so and luring us into some kind of trap without fair warning is evil, which is not BSP's style, judging from his previous scenarios.

He may have given the AI Masonry to give them access to marble and early Oracle, and to make the other ancient wonders go earlier than they normally would in a Monarch/AggAI/RB game. I just no longer think that he made it impossible to grab them. We can't know for certain, obviously.

What would our decision be, if we thought our odds for GW are 80-95%? Risk it?

 

[Mitchum]

What would our decision be, if we thought our odds for GW are 80-95%? Risk it?

Yes. I'd probably risk it at 50% given DV's decision tree.

Regarding Masonry, I forgot that you need it to have access to marble to speed up the Oracle. So, you're thinking that BSP told us about Priesthood to scare us away from the Oracle but he then had to give them Masonry so that they had access to marble just to make sure that they built is soon enough?

I don't have time to test right now but if we give all of the AIs PH, Masonry and Marble (no Stone) with some pre-improved tiles and/or a worker, what kind of wonder dates for Oracle, SH and GW would we expect to see? Does having marble and no stone make the AIs favor the Oracle over the other two?

 

[Pollina]

I don't have time to test right now but if we give all of the AIs PH, Masonry and Marble (no Stone) with some pre-improved tiles and/or a worker, what kind of wonder dates for Oracle, SH and GW would we expect to see? Does having marble and no stone make the AIs favor the Oracle over the other two?

One of the biggest factors in AI decision making when choosing to build a wonder is the number of turns they need to finish it. That's why they won't usually go for ultra early Pyramids, even when they have stone. A wonder that takes more than 10-15 turns (iirc) to finish has a very low priority, and is thus very unlikely to be built early on. Stone or no stone makes a huge difference, unless the AI has a lot of production in their city.

I can run a couple quick tests, if you want some numbers for your scenario.


Edit:

Test results until t41, with marble and 1-2 pre-improved resource tiles for each AI:

test1
Oracle t22 Zara
no SH or GW in construction

test 2
Oracle t24 Zara
SH in construction, delayed by barbs threatening city. Completed t53.
no GW in construction

test 3
Oracle t30 Gilga
SH in construction at t41 (Saladin), but delayed by barbs threatening city.
GW in construction (Hannibal hooked up 2nd city stone)

 

[Mitchum]

Awesome. Thanks for running the tests!!

Based on that, the key is not whether or not the AIs have Masonry. They key is whether or not BSP gave them stone... and if he did, will they pick GW or SH.

I think LC was able to figure out what tiles an AI might be working. Do we need to dig into that further or is it just a guess about how evil BSP really is?

 

[da_Vinci]

I like you already dV... what would be your 1 dimension inclination...

if you had to play right now, would you try for Masonry/GW or not?
whatever it is, we'll do the opposite....j/k

If I had to play before reading any of this, I would have figured I'd want to get GW, would not have inferred any strange AI bonuses regarding extra techs or tiles, and would have pursued a path to GW.

Then I might well have some garbage to take out.

After reading this, my brain is too scrambled to decide, But LC has me pretty scared about a GW chase or maybe it's just his subliminal messaging; GW ain't happening.

dV

 

[da_Vinci]

Here's another way to think of it...

This game is going to play vastly differently for GW and nonGW teams.
And make no mistake, this ain't no easy Monarch barbs. The AI is artificially far away. This could be the worst quantity of barbs you've ever experienced.

Seems like BSP would want all teams to get it or all not get it.

So, Assume no one will get GW
If we try for it and fail, we can recover. We'll survive the barbs. It's somewhat like if they popped a free tech from a hut and we didn't.
If we ignore GW, we get a slight advantage of those that tried for GW.

Now, Assume everyone will get GW
If we go for it too, we're equal to every other team.
If we ignore GW, we're the only team playing a very different scenario.

This answers dV's question about which choice leaves us most lagging the other teams. Being the only team without it.

It's a two by two table! I am an epidemiologist, so now we are all set!

Great Wall Possible and Chase Great Wall: Sweet!

Great Wall Possible and No Chase of Great Wall: Ouch2! How much worse are we than the chasers who got it?


Great Wall Impossible and no Chase of Great Wall: Sweet!

Great Wall Impossible and Chase Great Wall: Ouch! How much worse than if we ignored it? How far behind the non chasers


One way to think about this is how to minimize risk? Which ouch (mismatch of choice to scenario) puts us most behind other teams? Make a choice which avoids that ouch.

Alternatively, if one wants to max gain, which sweet puts us more ahead?

Consider that for a sweet to put us ahead, no one else can also be sweet (all must do other than us). For an ouch to put us behind, only one team needs avoid the ouch. So maybe minimizing ouch is better than maximizing sweet?

dV

 

[Mitchum]

So epidemiologist is the study of sweet vs. ouch? You learn something new every day!

Regarding this turnset, I think you should play it, DV. You don't need to do a ton of play testing to be able to do it. You just have to herd this group of cats and get us to agree on a path forward. Pushing the mouse around (not the one the cats are chasing) is the easy part!

 

[da_Vinci]

Here is a new thought ... if we get GW, AI may get a lot of archer promotions defending the barbs. If we don't get it, one AI will have fewer of those archer promotions. Does that matter at all?

Do we have a clear sense of whether Warrior ring defense is better or worse than archer defense? If we can get the choice down to three scenarios: Get GW, Fail GW and the best Ignore GW, does that make decision easier?

If Ignore GW has a value of 10x, How relatively bad is bad is fail GW? 5x? 1x? worse?

And if Ignore GW has a value of 10x, how much better is Get GW? 20x? 100x? more?

I expect we have different opinions, but it would be interesting to get each person's value set (among those who know enough to have a set, which leaves me out)

dV

 

[da_Vinci]

So epidemiologist is the study of sweet vs. ouch? You learn something new every day!

Regarding this turnset, I think you should play it, DV. You don't need to do a ton of play testing to be able to do it. You just have to herd this group of cats and get us to agree on a path forward. Pushing the mouse around (not the one the cats are chasing) is the easy part!

Well, we tend to study the predictors of bad outcomes as opposed to good outcomes, so sweet vs. ouch is not far off!

Also, in certain epi analyses, all the action is in the discordant cells (matched pairs who have different outcomes within the pair), which is like this situation.

Before you vote for me to play now, I have to confess that I am the guy who left a city un-garrisoned and lost it to a loaded barb galley early in the ren start Stalin game ... I could really test your no complaints rules ...

dV

 

[LowtherCastle]

​

+1
+1

-48
-653

​

 

[Mitchum]

Here is a new thought ... if we get GW, AI may get a lot of archer promotions defending the barbs. If we don't get it, one AI will have fewer of those archer promotions. Does that matter at all?

Do we have a clear sense of whether Warrior ring defense is better or worse than archer defense? If we can get the choice down to three scenarios: Get GW, Fail GW and the best Ignore GW, does that make decision easier?

If Ignore GW has a value of 10x, How relatively bad is bad is fail GW? 5x? 1x? worse?

And if Ignore GW has a value of 10x, how much better is Get GW? 20x? 100x? more?

I expect we have different opinions, but it would be interesting to get each person's value set (who knows enough to have a set, which leaves me out)

dV

If I understand this properly (I'm not a "Sweet / Ouch" guy), I would say:

Fail GW on T40 = 1x
Fail GW on T30 = 4X
Fail GW on T20 = 8X
Ignore GW = 10X
Get GW = 25X

Those are all brown numbers... and I'm definitely not the right person to ask!

 

[LowtherCastle]

You're all missing the key epidemiological question: On which turn, pre-T41, of barbs invading do we sh!t our pants? In lay terms, what do we do when we see a barb archer at our borders and it will take 4/3/2 turns to build an archer?

In LT terms, if I knew failing GW was merely WT:blahblahblah, then I'd happily take the risk. Problem is, you folks seem to be ignoring that failing GW equals having barbs stone-age us.

 

[LowtherCastle]

So the real question is, at what percentage of likelihood do we go for GW:

10% GW - 90% stone-aged
40% GW - 60% stone-aged
60% GW - 40% stone-aged
80% GW - 20% stone-aged

Just so you know, I'm not joking. This is precisely how I see it, except for this detail:

1% GW - 99% stone-aged

Frankly, I wouldn't be surprised to see some teams die.