SGOTM 15 - One Short Straw

OK. I'll play it forward to T18.

Here's the test save up to T18. I verified that the beakers, hammers and food were correct.

If you haven't played around in debug mode, it's pretty cool. For the most part, I let the AI do their thing while in Ctrl-Z debug mode, where you can see the whole map and what each AI is researching/building. If Toku or Hammy got off track, I would Alt-Z through the AI until I was controlling them and then I would change their build or tech. When the Toku worker didn't do what I wanted, I had to go into WB, remove the worker and add a new one to have him do what I wanted. Be sure to Alt-Z back to controlling OSS before you hit the red end turn button otherwise the computer will start contolling us...

Unfortunately, both Hammy and Toku are called Toku, so it gets confusing. Is it too late to change Hammy to Hammuragawa in the test save?

@LC For testing purposes, if you need to get the beakers right, you can always remove a river or add a cottage to fine tune the commerce. It would be good to have this correct in the test game if possible.

Okay, I'll play the test save forward.

Pretty sure this is basically right. The distance maintenance looks right. Toku managed to grow to pop4, but what the heck. We're guessing with some of his stuff anyway. I think Hammy got Sailing a turn too soon also, so somehow they're probably getting a few too many coins. No biggie. Teh purpose of the test save is to test our MM, not predict their behavior.

.

If Toku met someone on T12 who hadn't met any other AIs, then in 4t, that AI could have spent 20 on TOku and TOku would weight him 1, so on T17 he shift to 4:1, right?

Click to expand...

One point to note is that the AIs don't use Weighting numbers like "4" and "1."

Instead, they use numbers like "54," "72," and "18."

If you don't believe me, let the AI control our Civ for a turn after we have met both AIs and see just how crazy the Esiponage Weight values can look... or else, even better, just look at an AI using debug mode and look at their Espionage Weighting screen (Ctrl + e)... it will probably have crazy-looking numbers on it, too.

@LC For testing purposes, if you need to get the beakers right, you can always remove a river or add a cottage to fine tune the commerce.

Click to expand...

You can even go one step further. For example, you could add a Gold, Gem, or Silver Resource and a corresponding Mine to a square that you were going to work anyway, such as a Grassland Hills Mine... but since you'd be short one Hammer, you could then, say, add a Plains Hills square to the City Centre or a Forest to another square just to balance out the Hammers.

You can feasibly even add a Gold, Gem, or Silver Resource plus manually add a Mine to any square, including a Flood Plains square or whatever.

Basically, think of it as being:
- Food and Hammers are generally going to be right
- Finding ways of adjusting the Commerce
- When you increase the Commerce on a square (such as by adding a Gem Mine for a turn) whereby we lose a Hammer, then just make sure that you add a Hammer-producing item to another square and voila, you can easily manipulate the total Commerce. In fact, if you don't have a Library, Commerce can feasibly be temporarily added to or removed from any City that exists at the time, as long as you find a way to balance out the Food and Hammers properly.

Thanks for getting the test game ready for ZPV, LC.

ZPV, can you run through your mini-turnset with the new save just to make sure that the beakers turn out right. If so, I think everyone has agreed with it so you should be able to play forward.

Dhoomstriker said:

One point to note is that the AIs don't use Weighting numbers like "4" and "1."

Instead, they use numbers like "54," "72," and "18."

If you don't believe me, let the AI control our Civ for a turn after we have met both AIs and see just how crazy the Esiponage Weight values can look... or else, even better, just look at an AI using debug mode and look at their Espionage Weighting screen (Ctrl + e)... it will probably have crazy-looking numbers on it, too.

Click to expand...

You're right. The weightings in the test save towards us are 36 and 47.
I wonder if they're not being reset properly each turn.
edit: looks like that's it. So each turn the weighting changes by max(0,EPdeficit/8-attitudeVal).
I bet the function that they call to reset it to 0 each turn is a private function, being called from the wrong place.
So that means Hammi has stopped accumulating EPWeight against us, since he has an EP lead over us, while he continues to accumulate EPWeight against rival(s). We could calculate how many it would take to get us down to 1eppt.

edit2: ah, it only resets the weights to zero if an AI pushes the espionage slider. Otherwise they just keep accumulating.

Mitchum said:

Thanks for getting the test game ready for ZPV, LC.

ZPV, can you run through your mini-turnset with the new save just to make sure that the beakers turn out right. If so, I think everyone has agreed with it so you should be able to play forward.

Click to expand...

ok, will do.

Mitchum said:

Thanks for getting the test game ready for ZPV, LC.

ZPV, can you run through your mini-turnset with the new save just to make sure that the beakers turn out right. If so, I think everyone has agreed with it so you should be able to play forward.

Click to expand...

I've done that, with both the 2 turns at 0% and 1 at 50% variant.
The 2 at 0% gives us 7 more gold, in exchange for 9 fewer beakers in BW. I think that's a good exchange, so I'll look to spend the 2 turns at 0%.

Also: fortified:25% + hill vs a lion gives us 95.6% odds, so my plan is to stand and fight if I have to, and save the promotion for round 2 vs the next lion.

Sounds good to me. Good luck!

Whatever the weightings are, it boils down to 5eps or 4:1, or 3:whatever... We can possibly learn something about the other AIs using these numbers. For example, we know that Toku's and Hammy's attitudes toward us are -4. When Toku switches to 4:1 on T17, his attitude toward some AI has certain limits. That AI probably knew only Toku or Toku and one other, so he's either putting 5eppt on TOku or 2/3eppt on him. That gets up to 8/8 in 2-4t. Let's say it's 5eppt with Attitude +1.

Let's say it's 3eppt with Attitude +1.
Let's say it's 2eppt with Attitude +1.
So perhaps he met a second AI during these turns, but that would have required some perfect timing to coincide with turning off points for the first one.

This looks pretty complicated, but at the same time, it seems to be telling that the assignment to us only decreases over time. It's also telling that it decreases only 1ep at a time. It's also telling that the AIs' attitudes for each other increase every 10t. Somewhere in all this, there's some information.

Okay. So are we sure we want to go Masonry? The alternative BW-Writing is quite strong on research. If we're sure about Masonry, then let's go for it. We'll see what happens.

What I gather from all of this is that how the AI assigns epp is just too complicated.

LowtherCastle said:

Okay. So are we sure we want to go Masonry? The alternative BW-Writing is quite strong on research. If we're sure about Masonry, then let's go for it. We'll see what happens.

Click to expand...

Well, I thought we were all on board with the SH gambit for easy border pops in all of our cities. If we are, then we need Masonry next if we want the 100% stone bonus. If we're not, then which of the many test runs do you propose that go BW -> Writing?

Mitchum said:

Well, I thought we were all on board with the SH gambit for easy border pops in all of our cities. If we are, then we need Masonry next if we want the 100% stone bonus. If we're not, then which of the many test runs do you propose that go BW -> Writing?

Click to expand...

Masonry now enables us to pop Delhi's borders on T62 to acquire the marble.

The alternative might be to forget about getting the marble and chopping SH in Zlatorog, for example. That risks missing on SH, but it also keeps our gene pool clear in Delhi.

Each library + 2 sci adds about 10.5 bpt. That's a great way to boost our science. The question is, do we want to do that in an impure gene-pool city?

We can put libraries in gems or a gold city with similar results. Delhi can crank out workers and settlers and then the Oracle for a pure GPro gene pool. A GPro can bulb Civil Service... or did we mess that up already?

EDIT: For the CS bulb to work, we'd have to skip Masonry.

Anyway, I think we're burnt out on testing and the SH variant looks interesting, so we might as well move on to avoid losing more interest.

Ok, I'll do it in the morning.

While Bronze Working is nice, I wasn't finding us whipping much (and probably not at all with us having settled the Northern Gold instaed of Gold-1E) and Chops weren't doing all that much early on for us.

It will be very nice to get Bronze Working, though--does going for Masonry prevent us from researching Bronze Working for a very long time?

I also like the flexibility of skipping Fishing and settling Cow + Deer or no 5th City at all without having to settle Marble City. Of course, this fact may change if Copper pops up down there, but for now, Marble City is an investment that would take a while to pay off.


So, hopefully, we'll be getting Bronze Working within a reasonable number of techs, but even if not, getting Stonehenge in Delhi for the Cultural Border pops is going to provide an interesting enough game.

I also don't mind risking getting a Great Prophet--if you're planning on making an Academy, getting a Great Scientist as your first Great Person is very good, but after the first one, the returns aren't always amazing unless you can find a way to fund your Science Rate. Great Prophets can be of use in keeping your Science Rate up, thereby giving you extra use out of Libraries and Academies and allowing you to get a higher return on bonus Flasks.

So, sure, Masonry next is fine by me.

ok, playing now.
Turnset report t32-39
t32:
<Press enter>
IBT:
Toku founds Osaka at corn-1E.
t33:
there's no useful fogbusting adjustment to make
t34:
research to 100%
IBT:
Hammy revolts to Slavery.
t35:
our northern warrior has healed and starts moving south.
IBT:
A Japanese worker appears, on the corn.
I'll pause for thought.
------ continued after post 800 -----
t36:
DoW Tokugawa. Steal his worker.
IBT:
He doesn't attack. His power doesn't change.
t37:
I move our warrior into the Lion's Den (see post 801). Oh well. Promote and cross fingers.
IBT:
The lion attacks, and we lose.
t38:
Pause briefly to get my thoughts together.



t38: cont'd after post 825
I move our warrior east towards Osaka to threaten it.
The Worker retreats to fogbusty-safety.
IBT:
Osaka grows to size 2, and Toku builds an Archer.
The wounded Lion moves directly East of our Warrior.
t39:
save and upload

I was thinking that another reason to let him settle there was that he would offer us a worker. Hm...we don't have any strategic resource and BW isn't coming for a while. We haven't researched hunting so can't squeeze in archery. We probably won't be able to capture another unit of his, so DoP won't come for a while and won't be cheap, either. If we have a unit within two tiles of Osaka, is there any chance he'll offer us DoP for less than a city?

Also, what has his hpt been recently? Has his capital grown to pop4?

My preference is to capture the worker, but we don't want to be suicidal.

Edit: Also, what units do Hammy and TOku now threaten us with?