[BTS] SGOTM 13 - T'dr'duzk b'hazg t't

[Trystero]

OK - I will try some games with Bismarck as the German.

[neilmeister]

Originally Posted by Trystero

Really? I'm looking at the "Custom Game" set up screen and the B&S and M&S map types have no option for world wrap. Is this a variable that can be modified in a WB save?

I played a Medium & Small map a few weeks ago for a Gauntlet, and it was cylindrical. I don't recall the settings, maybe its Cylindrical by default.

The HoF rules state that cylindrical & toroidal are valid map options for 'XX & Small' maps. But I haven't sctually seen them.

[Trystero]

Yeah, I wasn't saying that those map types weren't cylindrical, I was just noting that there is no specific setting for world wrap with those map types at set up. Perhaps I'm trying too hard to glean information from the fact that the world wrap was specifically mentioned in the game settings.

[neilmeister]

Its always good to try reading the entrails.....

War of the Sixth Coalition (1813–1814), a coalition of Austria, Prussia, Russia, the United Kingdom, Sweden, Spain and a number of German States finally defeated France

So, it looks like we will have several German AI's, plus Victoria, Peter or Catherine, & Isabella.

[Trystero]

Possibly. But throw in this caveat (again from the maintenance thread):

Note that although the real-life Napoleon was exiled in 1814 AD, this game is a standard 4000BC start in all aspects other than the heavily modified map. Also, your opponents aren't exactly the countries of the sixth coalition - DS admits he took a bit of artistic license!

I had not considered multiple German leaders, but the qualification above doesn't make that seem too likely to me. But who knows?

Edit: Ah! Sweden! Now I see why you put Ragnar in your test game.

[Benginal]

Just ran a quick test game settling in place. My tech rate was pretty decent, but my hammers were just so incredibly poor. I got the oracle on turn 101 with the whip or turn 106 without it.

Build order: warrior -> wb1 -> finish warrior (no hammers lost) -> wb2 -> wb 3 -> one turn of lighthouse (to grow) -> worker -> finish lighthouse (chop) -> galley1 (whip) -> monument (chop) -> settler (whip) -> wb4 -> granary (whip into ...) -> Oracle (partial) -> galley2 (whip into...) -> Oracle

Tech order: Fishing -> Sailing -> Mining -> BW -> Mysticism -> Poly -> Priesthood -> Pottery -> Writing -> halfway into Monarchy.

Comments:

1. Settling in place does not get a lot of production.
2. Apparently we can go warrior, workboat, finish warrior with no hammers lost.
3. I built the settler late, but the first galley was able to get some good scouting done.
4. If we do want access to that copper, we MUST settle a city on that island where Neil put the copper. Obviously a lot depends on the real map, but if copper is where Neil put it, the only way to access it is by settling over there. This is another reason to settle on the PH. Note that putting a city over there is still possible because the 2-tiles away doesn't apply to different landmasses, but that's still a lot of overlap.
5. As of turn 100 I don't see any barb galleys.
6. Whip anger is going to be an issue so I think early Monarchy might be wise.
7. As Neil has already pointed out, the early lighthouse is HUGE early as it's giving us 3 food and makes the clams our three best tiles hands down.

Questions:

1. How do we feel about Meditation vs Poly for the Priesthood prereq?
2. How do we feel about one or two (or more) settlers built/whipped before the Oracle?
3. How do we feel about workboats vs. galleys being used for exploration?

[Benginal]

Continued on with the Test game up to 300 BC


A-M Builds: Forge (partial) -> Settler (whip) -> forge (whip) -> warrior -> warrior -> Colossus (partial) -> Settler (whip) -> Colossus (partial) -> worker (whip) -> warrior -> warrior -> Colossus (can finish on turn 152 with whip or by 157 without whip)

City 2 Builds: Granary -> Workboat (whip) -> monument (whip) -> workboat (whip) -> Library (whip) -> lighthouse. Note that I settled this city on the island right next to ours to the southeast so it has two clams in the first ring, but zero production.

Cities 3 & 4: Monuments and granaries. I'm letting the workers and workboats come from the first two cities.

Techs: Finished Monarchy, about to CoL. Confucianism went 380BC, so we just missed it.

Comments:

1. First Barb Galley appeared on turn 154, about 50 turns after we're going to get Metal Casting. So as long as we build and intelligently place a few triremes we should be fine.
2. The Colossus took a while but I didn't have copper hooked up, so hopefully it goes faster in the real game.
3. I'm going to be in favor of HR and lots of warriors. So we have to NOT research Hunting, assuming we get Copper.
4. We can do this game with very few workers.
5. I'm 10 turns away from a GP in A-M (89% prophet, 11% engineer) and 15 turns away in city 2 (100% scientist).

Fun Fact: I just popped tin in the capital. Too bad THAT won't happen.

[neilmeister]

Very nice use of the whip!

I don't whip enough probably.

@ Comments:

Yeah, not enough hammers is what turned me off SIP. I like the idea of Monarchy after Writing, because it fits in well with our - hoped for - early high population cities and Colossus strategy. Lighthouses will probably be the first build of most new cities. A-M will need to go Forge-Colossus, we will need to see if cities 2 and later 3 can contribute settlers/WB's by then.


@Questions:
1. I normally do Poly out of habit I suppose, as I am normally pursuing a space VC and want the GLib. We need Meditation for Philo, but normally by then it is a 1 turn tech. But, as it should shave 1 or 2 turns off the time we start building Oracle, then I say yes to Meditation.
2. I'm not so sure, we don't know where city 3 is going to go yet, and we have only 1 worker and 1 WB for city 2. I might try whipping a WB or 2 into the Oracle after settler 2.
3. WB's for exploration, normally. So much cheaper than a galley, which barbs eat at Emperor anyway.

My thoughts, I want to tech towards Maces and Cats, as always, but we don't know how close the neighbours are.


edit: X-post with Benginal. I will play my test out to 300BC and post the results.

[Alamankarazieff]

oop, too much tapas. Yes I was talking normal speed. Back online tomorrow. Itching to run some tests.

[Trystero]

I started running sets of test games to see when the AI would build the The Oracle. I also noted when other wonders and religions went for reference. This first set of tests used a default Archipelago map setting (landmass type: archipelago, rather than snaky continents or tiny islands). All other settings were identical to the actual game.

I did not do any city/worker micro during these games, or exploring, or settle a second city, and I followed roughly Neil's build/tech order. My primary intent here was to follow AI behavior given some assumptions about their identities and the map. I specifically avoided trying to build wonders myself.

For the first 3 tests I used Catherine, Frederick, Isabella, Joao, Victoria, and Willem as the AI opponents.

Game 1
Religions: Bud: t12, Hind: t24, Jud: t60
Wonders: SH: t93, Oracle: t102

Game 2
Religions: Bud: t13, Hind: t25, Jud: t69
Wonders: Oracle: t94, GW: t103

Game 3
Religions: Bud: t13, Hind: t28, Jud: t85
Wonders: SH: t97, Oracle: t97

For the next 3 tests I substituted Bismarck (Expansive/Industrious) for Frederick (Philosophical/Organized).

Game 4
Religions: Bud: t12, Hind: t24, Jud: t74
Wonders: SH: t70, GW: t77, Oracle: t109

Game 5
Religions: Bud: t13, Hind: t26, Jud: t81
Wonders: SH: t92, GW: t105, Oracle: t122

Game 6
Religions: Bud: t12, Hind: t52, Jud: t78
Wonders: GW: t119, ToA: 119, Oracle: t128

Given the caveat that these are small sample sizes, the presence of an Industrious leader did not appear to make The Oracle go earlier (It actually went later if Bismarck was on the map ). This map script gives some isolation, but lots of room for expansion. None of the AI's were stuck on small islands when I checked the saves in WB, so that did not necessarily constrain the AIs. Most had 2 or 3 cities settled by turn 100 or so.

I also only saw a barb galley in one of the games (game 5, on turn 84). Granted I did not do any exploring, so the barbs had to find me. But even so, I did meet at least one AI in each game.

In Neil's two tests, we finish The Oracle on turns 103 and 107. It's hard to say what effect the map would have on AI wonder building, but under these settings we'd miss out on The Oracle in the first test case (with Frederick rather than Bismarck on the map). Again, I don't know how much weight to give that conclusion, given all the potential variables.

Also note that I didn't edit the saves at all, so no AI started with the Great Lighthouse, nor was it built in any of these test games (although I did note it being constructed when I checked in WB at the end of a couple of games). I am not sure whether starting with the GLH would alter AI wonderbuilding. They'd certainly tech faster.

I will try the same AI combinations with a Medium & Small map tomorrow. I might also run a couple more games with these settings to see if the German leader really does make a difference. I'd also be glad to hear any other suggestions for changing variables.


@Benigal: Nice test game! I will have to look at it carefully. I am inclined to agree with you however that HR will be a priority. And I normally use WBs for exploration. Mostly because we can get one out faster (we don't need Sailing to be done, and they are fewer hammers than galleys).

[neilmeister]

I'm hoping to get it under t100, with some whips and with working some of the coastal squares instead of the mines. Either way, it seems like we cannot be sure that we get Oracle. I still think it is worth the risk. We would get it 4/6 times with your tests.

[drlake]

Comment on exploring: I normally play Emp, and Galleys are far better for exploring a map like this than WBs. The key is to defend with them, not attack. That gives them better than a 50% chance of survival, which beats the hell out of WBs. The only real advantage of WBs is that you can get them out earlier, but I doubt we'll be exploring very far at that stage if we're going for the Oracle, because we can't afford the early production diversion.

Of course, if we wait long enough, Triremes are far better for exploring, and will still be early enough to win circumnavigation if it is possible on this map.

[neilmeister]

I think we can whip an exploring WB into Oracle. Early WB is to try and get contact with as many AI as possible, and crucially for this map, circumnavigation, before the barb galleys appear.

[Trystero]

I tried 2 more tests with each German leader. I figure 5 attempts with each would give a somewhat statistically significant outcome.

With Frederick:

Game 1
Religions: Bud: t14, Hind: t30, Jud: t92
Wonders: SH: t66, GW: t87, Oracle: t111

Game 2
Religions: Bud: t13, Hind: t44, Jud: -
Wonders: SH: t69, GW: t87, Oracle: t99

With Bismarck:

Game 1
Religions: Bud: t13, Hind: t25, Jud: t82
Wonders: GW: t98, SH: t108, Oracle: t119

Game 2
Religions: Bud: t13, Hind: t25, Jud: t121
Wonders: GW: t70, SH: t78, GLH: t:117, Oracle: t124

In summary, the Oracle build times were:

Frederick: 94/97/99/102/111
Bismarck: 109/119/122/124/128

The differences between the two leaders are pretty remarkable and counter-intuitive. Especially since in the Frederick games it wasn't always the Germans who build The Oracle. Again, this map type leaves ample room for AI expansion (most had 3 cities by turn 110 or so), so I think that aspect at least will be similar to the real game. What appears to be important is which AIs are on the map. I made educated guesses, but Dynamic Spirit could have chosen to make the Romans or the Vikings part of the 6th Coalition I suppose.

Originally Posted by neilmeister

I'm hoping to get it under t100, with some whips and with working some of the coastal squares instead of the mines. Either way, it seems like we cannot be sure that we get Oracle. I still think it is worth the risk. We would get it 4/6 times with your tests.

I agree it seems worth the risk. And yes, getting it before turn 100 would greatly improve our chances. Pre-turn 95 would be ideal, but I'm not sure that's possible given the scarcity of trees to chop (at least that we can see).

[Brian Shanahan]

Sorry for the lack of input so far, have been preparing for a speech with Toastmasters. I'll run a few tests myself to get a feel for the strategy tomorrow, and give my thoughts then. Looks interesting so far though.

[neilmeister]

I got it on t92, by doing BW & Pottery after Priesthood. Whipped Oracle on the turn I learned Pottery.

On the minus, only have 2 cities, not three, and A-M is pretty whipped down. Copper didn't show up until well after the settler was out.

[Benginal]

Originally Posted by neilmeister

I got it on t92, by doing BW & Pottery after Priesthood. Whipped Oracle on the turn I learned Pottery.

This is a good idea. I'm guessing you didn't even have to worry about getting the Oracle too quickly given how weak our hammer situation is. But this way we guarantee getting an early Metal Casting which I think is worth not getting an early third city.

Originally Posted by neilmeister

On the minus, only have 2 cities, not three,

I'm not convinced this is really that bad. Obviously we'll have to do some scouting in the real game to reach a decision. I think an early second city to take over worker/settler/workboat/galley/trireme production is good while our capital builds oracle -> forge -> colossus. But an early third city isn't vital I don't think.

Originally Posted by neilmeister

and A-M is pretty whipped down.

Don't worry about it. The past few months have seen me fall in love with whipping. You can pretty much whip as much as you want without consequence in HR. I mean to a point, don't go overboard. But usually the 100+ turns of whipping unhappiness is worth it.

[Benginal]

Question for Neil: Did you go with medi or poly on that t92 test?

[neilmeister]

Meditation, settling on the PH.


ediy: Yeah, I think it works well.

[Trystero]

Originally Posted by Benginal

Don't worry about it. The past few months have seen me fall in love with whipping. You can pretty much whip as much as you want without consequence in HR. I mean to a point, don't go overboard. But usually the 100+ turns of whipping unhappiness is worth it.

Yes - If I recall correctly in SGOTM12, some of the top teams whipped much more than we did in the early going. I'll have to look back at those threads....

Regardless, our start will have plenty of food and limited hammers, a situation which seems to call for heavy use of the whip.

I do think that early expansion is important, but I'm going to agree with Benginal, in that the timing of the third settler is going to depend on what our scouting finds. If we can get the Colossus, by way of Oracle > MC, it would be worth delaying our 3rd city.

Edit: And nice work Neil. Finishing The Oracle on t92 seems close to a sure thing.