SGOTM 08 - Fifth Element

BLubmuz · Sep 19, 2008

Guys, can we concentrate on those early TSs and stop after the Oracle is built to discuss the mid-game strategy?

No one spent a word on my tests: what about that strategy?

OK, now let me try to improve that 1320 BC.

WastinTime · Sep 19, 2008

I'll run a test game until Oracle is done late tonight, then I'll be ready to discuss your tests Blub

BLubmuz · Sep 19, 2008

in one of my attempts to improve, Oracle was built IADL in 1480 BC :eek:

and double :eek:

totally out of the blue, i always manage to built it, even in 1200.

OK, in some 12 attempts once it was BIADL in 1360, in this one in 1480.

If this happens even an MC sling is not safe.
Probabilities? hmmm, some 84% of success before 1280 is not bad.

With WB i've seen it was the damn Justinian. It seems that the religious civ often go for the Oracle.

greatbeyond · Sep 19, 2008

Honest, I don't really mean to be a wet blanket and I won't have time for tests until Sunday night, but....

Remember the point of thiis contest is to own the most wonders for the most turns. Letting others build them and then capturing them, in my estimation will be neither quick or a sure thing. If Astro is required, it may only be us, and Shaka on this side of the world, guaranteeing a long number of turns until we can even conquer them.

What if they don,t build many wonders?

I am not one to leave our fate in the AIs hands.

As Blubmuz has found out, the Ai can build the wonders we want before us. I was not able to get the GLighthouse in any of the games I tried it.

I was proven wrong several times in the past sgotms so keep that in mind.

Sweetacshon · Sep 19, 2008

dutchfire said:
I think the Colossus isn't majorly important, we have a lot of green land, and working grassland is usually better than working coast, so the Colossus won't give us that much extra.

Except when you're putting a city on a 2 tile island.

BLubmuz said:
in one of my attempts to improve, Oracle was built IADL in 1480 BC and double

totally out of the blue, i always manage to built it, even in 1200.

OK, in some 12 attempts once it was BIADL in 1360, in this one in 1480.

If this happens even an MC sling is not safe.
Probabilities? hmmm, some 84% of success before 1280 is not bad.

With WB i've seen it was the damn Justinian. It seems that the religious civ often go for the Oracle.

1480BC!? On Monarch? That's scandalous. On a test yesterday, I got the CS slingshot at around 1000BC.... I would've tested more, but got sidetracked by continuing to play the save (just...one...more...turn). It was suboptimal play, no doubt about it, but that's 3/3 CS-oracle tests for me. Your turn 67 sounds good, Blubmuz... I slipped sailing in to settle the island (city 2) pre-oracle to gain the marble, but found then that I was waiting for tech to catch up with the oracle at 1 turn for a while. Still, send that second workboat north combined with sailing makes it worthwhile in the trade commerce sense... or not.. I guess shaka is right there, but I thought that if we want to beat on him early, we shouldn't give him a leg up with a trade route.

On another point, further into the game, the pig city never really takes off due to it's v poor food situation. Should be good enough for a minor unit pumper, though. The northern city I cottaged, and it was great, but it could just as easily go farm/workshop if we need the :hammers:

Sweetacshon · Sep 19, 2008

greatbeyond said:
Remember the point of thiis contest is to own the most wonders for the most turns. Isn't it for the least turns?
Letting others build them and then capturing them, in my estimation will be neither quick or a sure thing. If Astro is required, it may only be us, and Shaka on this side of the world, guaranteeing a long number of turns until we can even conquer them. What if they don,t build many wonders?
Very true. A lot depends on the map, which is why I'd like that 2nd wb to be a scout. However, my first empirical "press enter and let the AI build" test showed the graph to flatten after about turn 200 (in terms of total wonders built in the world, meaning that any finish after that point should score similarlyas long as wonders keep getting built. The challenge, then, is how to achieve wonder building while we are conquering?? The only two answers I have are:

To mostly dominate the world, building all the wonders we can in a flurry at the end, just before finishing.

to keep an industrious civ (or 2) friendly and teched up so that they build for us and we steal them at the end after conquering the rest of the world

Anyway, the point right now is the oracle. Blubmuz, can you explain to me the need for the wheel before writing?

culdeus · Sep 19, 2008

greatbeyond said:
Remember the point of thiis contest is to own the most wonders for the most turns. Letting others build them and then capturing them, in my estimation will be neither quick or a sure thing.

If we can't capture every single AI city on monarch with only a hint of resistance then something has gone horribly wrong.

greatbeyond · Sep 19, 2008

You're right Sweetacshon,kind of. For two teams that have the same number of wonders built at exactly the same time the winner would be the ones who have the earliest win. IIRC, that's where the curve comes into play. If we get lot of early wonders, we get more points than a team that captures all the wonders later in the game. To really simplify, 5 wonders for 100 turns is the same as 100 wonders for 5 turns. Of course I could be screwing up the winning criteria. I would have to look back and see who it was that checked this out. We want to shoot for the exact sweetspot of the curve.

greatbeyond · Sep 19, 2008

culdeus said:
If we can't capture every single AI city on monarch with only a hint of resistance then something has gone horribly wrong.

I don't have any doubt we can capture them, but at what turn and how many wonder points will we have?
If we build very few ourselves, we lose credit for all the turns we could have had them. Because of the way the wonders are spread out, in a perfect game there is a point where, if you could win anytime you want, playing any further doesn't gain you a better win.

I'm thumbing all this in, so I don't have strength for a long analysis nor civ to test it.

dutchfire · Sep 20, 2008

@greatbeyond: Just to be clear, it's wonders at the end of the game / total turns, so it doesn't matter if we get the wonders now, or the turn before we win the game.

BLubmuz · Sep 20, 2008

Dutch preceeded me, GB the victory will go to who will have the best rate turns/wonders or viceversa.

Now, if you like to talk about mid-game strategy, here are some of my thoughts:
CS sling or not, if we have no contact with other AI than Shaka we can/must research on our own: BW, sailing, pottery, all the early techs, then the path to literature.
In the meantime we'll think how wipe Shaka ASAP.
After Literature we'll go for optics, then EDU.
We must build the pyramids, and we'll probably loose the GLH and the ToA, for sure we'll loose the GW and SH.

Probably, if we succeed on the CS sling, the colossus can be ours.

Our UU is not that great, it's a WE which attacks first mounted units in the field, not inside a city. Never experimented with that, even if i played a good game with that leader.
Don't underestimate the AI on Monarch, remember that all our SG were on monarch before the last one.

Sweetacshon said:
Anyway, the point right now is the oracle. Blubmuz, can you explain to me the need for the wheel before writing?

Yes, to keep the worker busy, but let me experiment some more.

Mesix · Sep 20, 2008

I agree with Dutchfire. The number of turns that the wonders are held makes no difference in the outcome. It is the wonder points/number of turns. What the earliest finish tie breaker means is that if we get 40 wonders (40*5=200 points) by turn 200 and another group gets 30 wonders (30*5=150 points) by turn 150 we would both have a score of 1.0. In that case, the team that finished earlier (turn 150) would win regardless of if they built all the wonders early or captured them all on the very last turn of the game.

BLubmuz · Sep 20, 2008

Well explained, Mesix, but:
both Alan and Gyathaar posted there's not a tie-breaker.

Thus, in the example you posted, both the teams will win, or whatever.

Anyway this is to post another couple attempts: writing before the wheel is totally useless, better is worst. Why? the capital is busy producing WBs, and the library will be delayed to max 1-2 turns, while the worker stays idle.

Remember that almost every road asks 3 turns: 1 to arrive, 2 to build, 'coz the most part of our terrain is forest, jungle or hills.

Interesting my last attempt: i built a settler before the library, and the Oracle was built in 1240 BC, with CS in the same turn. Those 3 turns of delay compared to my best attempt (1320 BC) let city 2 buid a settler after library, and city 3 is up and running in 1200 BC.

One more attempt: switching settler/library Oracle and Cs was done in 1240, but city 3 is delayed of some 5 turns.
Another one: this time i has 2 events: Mother Lode

and Herbalist - went for the mid, no success, but pop killed, 1 angry citizen, Oracle in 1160 and no academy yet.

This makes only 2 attempts failed out of total 16: 12.5% chance of fail.

Mesix · Sep 20, 2008

Nice job with the test games. I haven't had much time to play this week. Work was very busy and I didn't get any reading done for school...so it was actually homework. I may play a test game or two tonight if I finish up the last of my homework early ehough.

I'm okay with getting the wheel first as long as it doesn't delay the library. To keep the worker busy (while researching the wheel after writing in my earlier test game), I had him build a mine south of the capitol. Then I had him move to the swuare where the gold mine was. He arrived as writing was finished and built a road connecting the cows and gold to the capitol before building a road to the 2nd city location. I can see the benefit of researching the wheel earlier and connecting up the resources. I just wonder if the worker won't remain idle at the 2nd city location waiting for the settler to finish.

I think that getting a library up in the capitol as soon as possible will help us to reach any goals in the early game. Additional research as well as an early academy will pay dividends throughout the game.

BLubmuz · Sep 20, 2008

The best attempt privileged the research, not sure where i lost some turn in the following ones.
Having city 2 up soon will slightly help the research (trade routes) despite we go @90%. It also lets us improve the ivory, to reach fast size 7 in capital. Needless to describe what this implies for research and production.
The worker will never be idle: we need to connect the mine where he stands when the Wheel arrives, then the forest 1SW of the capital, then the gold and cows. After those he will jump on the hill which will connect city 2, and the settler arrives on location in 1 turn. After that he can go on ivory or pigs, road then connect to the network. He can also build a mine on the hill 2S2W of capital before city 2 is up. As you see this is a lot of work, so my vote on wheel first.

Now i'll try to repeat my best attempt, but keeping track of any MM and worker job.

Alright, here is another attempt on 1320

Spoiler :

This last one required a lot less MM than the previous.
I think the Autolog is self explaining.
For what to do with the worker, see above.

I needed 3 more attepmts to arrive here, so i think our chance of success is around the 90%.

Now:
Votes, please.
I'll post the key points on the 3rd post, we'll vote and in case i'll change them accordingly.

WastinTime · Sep 20, 2008

I ran test game where Oracle was built in 525 BC. CS-slingshot seems worthwhile.

I can't find even one small thing I don't like about the plan in post #3...I better look again.

Mesix · Sep 20, 2008

Are you building the Oracle in the capitol? Would it be possible to build it in city #2 and chop rush it? If we use this method then the capitol would be free to build more workers, another settler, and military units sooner. It would also make the 2nd city a GP center.

Another question I have is which city will build Pyramids and become GE center? What is the timeline to get Pyramids before the AI?

BLubmuz · Sep 20, 2008

Mesix said:
Are you building the Oracle in the capitol? Would it be possible to build it in city #2 and chop rush it?
no, we can't research BW and stay on this plan
but i can try qo squeeze a test on this
If we use this method then the capitol would be free to build more workers, another settler, and military units sooner. It would also make the 2nd city a GP center.
see above

Another question I have is which city will build Pyramids and become GE center? What is the timeline to get Pyramids before the AI?

The only way to have Pyramids built on a decent date (for us, usually AI don't prioritize them) without stone is to build them in our bureaucratic capital!

I know this throws specialized cities in the toilet, but i can't see alternatives.

EDIT: made a try, oracle built in 950 BC (production problems).
2nd try, 975 BC (research problems)
I'll try to improve, this is too risky.

There's a possible alternative: research sailing and masonry (instead of BW), build a galley and a 2nd settler, settle on marble.
But i guess this will delay more, due to research.
Unless we LB Mathematics.

Mesix · Sep 20, 2008

I ran a test. I was able to build the Oracle in the 2nd city by turn 76 chopping 3 forest tiles. The Pyramids are difficult to build in the 3rd city without getting MC since we can't build workshops. Chopping 3 forest tiles for the Pyramids does not get them very far. There is only one mine that can be built in the BFC of the 3rd city. The additional benefit to getting BW is that we can find out if we have bronze nearby and start building troops to take out Shaka sooner rather than later.

For what it is worth...here is the autolog from my test game. If you look at turn 19 you can see my research path. I'm sure there is some micromanaging that could improve the results a bit. One idea would be to delay the settler for city #3 until the capitol grows to 7 pop to run an additional scientist.

Spoiler :

Since both plans (with and without BW) require fishing>writing and 2 WB>library, I think that we can agree on the short term path for the next turn set or 2. I also think that it is important to get more info on the actual map that we are playing (meet other AI players, scout more map tiles, etc.) before we plan too far ahead.

After we get the GS for the Academy, what do we do with other GS from the capitol?

I think that your plan is really well worked to squeeze out that CS slingshot, but at what cost? Will we still be able to get the Pyramids? How will the GP points from the Oracle pollute our pool in our capitol?

BLubmuz · Sep 20, 2008

3rd test: 1025 BC, but LB Math.
Still too risky, and we forget the Academy for awhile.

The nice side of this attempt is that by that date we got 4 cities up and running.

Chopping a wonder without the appropriate resource is almost useless.

Yes Mesix, your attempt is very close to my 2nd one.

But i agree on your proposal: let's proceed with the actual game, after we voted for the plans of next turnset.
Since we'll surely stop after that, we can discuss possible variants due also to a better knowledge of the map.

When a TS plan will be approved, i'll change in this way on post #3, with its eventual variations.

After the votes Culdeus is UP

SGOTM 08 - Fifth Element

HoF Quattromaster

Deity

HoF Quattromaster

King

In the foetal position

In the foetal position

Emperor

King

King

Deity

HoF Quattromaster

Obergruppenführer

HoF Quattromaster

Obergruppenführer

HoF Quattromaster

Deity

Obergruppenführer

HoF Quattromaster

Obergruppenführer

HoF Quattromaster

Similar threads