SGOTM 09 - Xteam

[DJMGator13]

Good progress everyone.

An airship would allow us to scout for and weaken any approaching barb before it hits our borders. This might delay it from actually attacking one of our cities while we are running a minimum number of defenders.

Without the ability to granaries right now we may want to nearly complete a barrack or a warrior and then poprush it so that we start with a hammer overflow towards the Oracle build. (You all may have done this in the practice games but I wasn't sure.)

I won't have time to run through the practice game until Friday night and by then Fred will have probably posted his pre-plan.

 

[PaulisKhan]

If someone is near a computer can they test this (it will be superior to the warrior overflow method), I'll be home in around 6 hours to try it myself otherwise

1) Chop a forest or two into the oracle (and invest some build hammers) in the NON oracle intended city that has marble hooked up(ideally on the first turn it becomes available)

2) Cancel the build and complete the oracle in the intended oracle city.

I'm 99.9% sure that we get to cash in the hammers from the city that didn't finish the oracle.
Doing this gives us a 1.5x multiplier on our hammer-gold conversion rather than the 1.25x multiplier for a warrior under police state (or do we get 2x multiplier from marble, I'm ashamed to say I'm not sure >.<).

I suspect that delaying oracle by a turn or two to invest more hammers could actually be worth more than the couple of commerce we lose from the delay in electricity (it would let us run at 100% for several more turns which should recoup the investment).

-edit- thinking on the plan some more: hook marble city up to capital, chop two riverside capital forests into oracle and invest build hammers until oracle almost complete, then cancel build and chop oracle into marble city.
This nets us 1.5x(2x?) hammer cost of oracle in gold for only a couple of turns lost electricity commerce.
That will fuel our research until the next wonder investment comes through.

 

[Cactus Pete]

"It may be to our advantage not to connect copper immediately after building the mine because we want to build more warriors for fog busting and later MP duty. Warriors are much cheaper than phalanx's and spears. If we don't connect the copper, we need to get an airship out quickly to give advance notice of potential problems. There could be another advanced barb city in the western fog.

It's probably too early to build barracks. We are not anywhere close to attacking someone so it's probably better to invest our hammers in something immediately useful. Yes, unless it's the only good option for storing hammers (they don't deteriorate quickly, and we'll need the rax eventually). There aren't many alternatives except building warriors/workers and settlers. And airships or possibly Oracle in a second city, as PK suggests. This is also one reason why I would like to get to Pottery faster (So that we can build granaries). Perhaps postpone Wheel even further [But Fred, you (unlike G22) are building the Oracle in the capital, which requires roads to hook up the marble. Are you arguing for building the Oracle in Marbletown to avoid researching The Wheel? If so, like your thinking, but it won't work -- need The Wheel for Pottery.]and substitute with Pottery?

It would be very interesting to see if it's possible to get some gold from overflow without delaying the Oracle significantly. The mechanism for this is as follows: Build a warrior until he is almost complete. Then whip him and finish a forest chop simultaneously. This generates so much overflow that some of the hammers will be paid out in gold instead of overflowing to the next build." This could work for building anything and might be worthwhile, but the more interesting possibility is taking advantage of the marble multiplier (if that works).

I did a test of my own with more emphasis on tech rate. Here's the result:

Test game w. gold chop

I managed to get the Oracle in turn 43 and chop some gold to fuel research. This enabled me to get Pottery. AH is due in 2 turns and Writing in 5. There is, of course a trade off compared to e.g. SCT's save where a Settler is available along with an almost completed Airship. I believe either SCT or myself could have gotten Pottery researched as well if we had delayed AH. I opted for AH first because the sheep needs to be pastured, my cities were already at max size (didn't get a forest preserve created), and wanted to know where the horses were before the next settler was put in play.I think I'm willing to do the trade in order to get granaries and Libraries faster. When Writing is in the bag we are ready for the next expansion phase. Fast Writing is important if we want to avoid getting a GProphet from Oracle.

We'll need to get a library built in 7 or 8 turns (have a total of 17 turns before the GP -- not exactly sure how this works) after the Oracle is completed and probably have to run 2 scientists immediately until GS. Argues for Writing before Pottery. Not easy in any case. Will it be worth it?

 

[Cactus Pete]

I've managed to get the Oracle on turn 42. Yes, and not in the capital. There is much in common with mine and SCT's saves. Only chopping two forests, plus a forest preserve is certainly an improvement on my attempt, but small cities (will be relatively fewer hammers and commerce in the future) and fewer units isn't. Also there is merit in Fred's suggestions.
G22

Doubt any of us has optimized. Hope you'll have another go.

 

[Cactus Pete]

Comparing SCT's test and mine, there is considerable similarity. He has chopped one more forest and not built any roads, but he has created a forest preserve where one must go (bad mistake to have chopped this forest, as I did) and has an additional windmill improvement.

I think that, if I build one less road, I can leave the critical forest unchopped for a later preserve (utilizing a forest next to the city). The question, then, is whether the remaining roads are more valuable at this stage than the preserve and the windmill. Suspect the improvements are the better value, but not certain, as lack of roads means lack of mobility. Barbs could threaten.

Also would argue that getting the airship out earlier, which SCT apparently could have done but didn't, provides advance information on any activities around Marbletown. Can't afford to put that city at risk, and it has only warrior defence.

Possibilty of chopping for gold should be explored as well.

Hope Fred will borrow from all three test attempts and try his hand at putting the Oracle in Marbletown.

 

[ShannonCT]

If someone is near a computer can they test this (it will be superior to the warrior overflow method), I'll be home in around 6 hours to try it myself otherwise

1) Chop a forest or two into the oracle (and invest some build hammers) in the NON oracle intended city that has marble hooked up(ideally on the first turn it becomes available)

2) Cancel the build and complete the oracle in the intended oracle city.

I'm 99.9% sure that we get to cash in the hammers from the city that didn't finish the oracle.
Doing this gives us a 1.5x multiplier on our hammer-gold conversion rather than the 1.25x multiplier for a warrior under police state (or do we get 2x multiplier from marble, I'm ashamed to say I'm not sure >.<).

I suspect that delaying oracle by a turn or two to invest more hammers could actually be worth more than the couple of commerce we lose from the delay in electricity (it would let us run at 100% for several more turns which should recoup the investment).

-edit- thinking on the plan some more: hook marble city up to capital, chop two riverside capital forests into oracle and invest build hammers until oracle almost complete, then cancel build and chop oracle into marble city.
This nets us 1.5x(2x?) hammer cost of oracle in gold for only a couple of turns lost electricity commerce.
That will fuel our research until the next wonder investment comes through.

Yes, this works, and can be done so that Oracle still completes in Marble City on turn 43. The difference is that we'll need to chop 4 forests into Marble City so that it can be produced there immediately after Athens stops building it. Delaying the Oracle is too risky. Spending more worker turns on chopping and roading will mean we will be delayed in building watermills and preserves. If getting this extra gold means Athens takes longer to work a 6th tile, a 7th tile, an 8th tile, etc, is it worth pursuing?

 

[PaulisKhan]

I didn't do any real micro as this was more a proof of concept rather than optimised strategy but the results are positive.

Test 1 went as expected
1) Chopped two forests into oracle in capital along with natural build
2) cancelled build and chopped two forests into oracle in marble city and finished it.
3) next turn recieved 120ish gold in the bank from capital hammers, fueled research for a good period of time

Test 2
1) deposited hammers into stonehenge in all 3 cities in the period that I was wanting them to grow or when they didn't have anything that was vital at the time, can mix and match cities depending on which one has nothing better to do (obviously this works better with a marble WW)
2) Stonehenge BIDL, recieved gold from all three cities corresponding to the number of hammers deposited into the wonder in that particular city.

excellent results and possibly the best use of hammers when we want a city to grow.
For infrastructure we could just use the whip as the food->hammer conversion will usually be better for that.

If the strategy sounds good then we'll need to work on optimising it for the next turnset.

 

[Grover22]

PK,

Test 2 seems far easier to pull off to me. I'll save test 1 for the experts. I wound up building and finishing a barracks in Athens during my try just so that the hammers would have somewhere to go while Athens grew. Stonehenge is a better idea.

I'll try to run an improved set tomorrow morning, using Stonehenges as hammer dumps and not building so many roads.

G22

 

[PaulisKhan]

I take on board the comment about the opportunity cost of the worker actions and alternative build actions.

The research it is worth is the difference in the max tech rate, and the tech rate we'd be forced to use to not go negative (takes a dive when we settle the 4th city).
I think it's worth the 4-5 turn delay to the windmill.

 

[PaulisKhan]

PK,

Test 2 seems far easier to pull off to me. I'll save test 1 for the experts. I wound up building and finishing a barracks in Athens during my try just so that the hammers would have somewhere to go while Athens grew. Stonehenge is a better idea.

I'll try to run an improved set tomorrow morning, using Stonehenges as hammer dumps and not building so many roads.

G22

The downside of this is that we don't have stone, we're not industrious and we aren't able to adopt Org Rel, so we're just getting a 1:1 conversion rate on our hammers.
It will definitely work great when we start unlocking the marble WW's though, I accept that we probably have to build enough units early on to fogbust and defend against barbs too.

Test 1 can carry a turn or two of risk, and require some extra chops/worker turns, but we get the benefit of the 1:2 conversion rate of hammers to gold which is where we have a chance to pull ahead of other teams and stay competitive with the best.

-edit- realistically, what is the earliest that the Oracle will be built by the AI?

 

[ShannonCT]

The downside of this is that we don't have stone, we're not industrious and we aren't able to adopt Org Rel, so we're just getting a 1:1 conversion rate on our hammers.
It will definitely work great when we start unlocking the marble WW's though, I accept that we probably have to build enough units early on to fogbust and defend against barbs too.

Test 1 can carry a turn or two of risk, and require some extra chops/worker turns, but we get the benefit of the 1:2 conversion rate of hammers to gold which is where we have a chance to pull ahead of other teams and stay competitive with the best.

-edit- realistically, what is the earliest that the Oracle will be built by the AI?

Turn 44 seems to be the earliest. I've seen 3 games with these settings where Oracle went to the AI at turn 44.

I'm thinking that the delay in building more mills and preserves will have Athens working 1 less tile than it could be for a prolonged period. If that causes us to lose 3-4 commerce per turn for 20 turns, we'll be giving back most of our gains.

I believe what the competition is going to hinge on is getting to Rifling quickly and having the cities and population and workers to produce a paratrooper army quickly. 100 gold from Oracle is a pretty small fraction of the beakers we need for Rifling. If this causes any delay in horizontal and vertical expansion, I don't think it's going to be worth the opportunity cost.

 

[Cactus Pete]

"realistically, what is the earliest that the Oracle will be built by the AI?" Even if the probability is only 10% (might know better if we look at who gets Fishing as an initial tech) that the AI will build on turn 44, question whether the risk (little chance to win the game) is worth the reward. We will be able to run close to 100% science for a while without extra gold. Won't the Stonehenge sink and the extra gold from more improvements and city growth allowed by preserves pretty well cover the problem over until there is another opportunity to get an excellent return chopping gold?

 

[PaulisKhan]

You've talked me out of it, leaving post up for posterity.

True, but the low cost early techs generally take longer to research than the high cost techs later and the sooner we bring on things like forges, oxford and courthouses, the sooner we get to rifling.

Like you've done, for my own games I decide on the major keystone events and decide how to get there fastest.
The phase of lost commerce due to a missing windmill is ended pretty much as soon as we pause growth in our capital to build another settler/worker (allows the improved tiles to catch up to pop again)
This basically means if we build a settler or worker in our capital any time in the 25-30 turns following the oracle then we're better off with the gold.
With 4 cities it's worth around 9 beakers per turn to us (42 bpt@100% vs 33bpt@80%)

The point you make that does have me worried is the AI nabbing oracle before us. If we miss out on that then it's game over for us. No electricity means we'd never catch up to the other teams.
Turn 44 seems incredibly early but if you've seen it then we can't afford to take the risk of delay.

I trust your judgement so decide for yourself if the strategy is worth it or not.

There will almost certainly be times later when we want to invest hammers into incompleted wonders though, so don't take it off the table completely.
ToA is a good example of a WW that could be whored around to our cities when we are wanting to grow them.


-edit- I've been convinced.
Focus on turn 42 or better Oracle.
Sink hammers into SH at a 1:1 in cities when we have nothing better to build, or warrior whip overflow at 1:1.25 if we're prepared to sacrifice the pop.

 

[Frederiksberg]

Don't have much time to test today, but will do some more tomorrow.

I'll investigate some of the alternatives like:

• Building Oracle in Marble city
• Putting hammers into SH. Or we could research Poly instead of Med and put hammers into ToA.

What is the additional cost of researching Writing before Pottery? (I.e. with one less prerequisite).

 

[PaulisKhan]

Think Oracle in marble city is the main strategy now rather than the alternative

I always tech poly on my way to oracle, didn't realise people used the Med route >.<


Are we better off putting hammers/chops into warriors instead of henge for the 1.25 multiplier until we get access to marble wonders?

 

[Frederiksberg]

I always tech poly on my way to oracle, didn't realize people used the Med route

Med is cheaper. If we go Poly we can put hammers into ToA. Or will the AI choose not to build it now that it's obsolete?

Are we better off putting hammers/chops into warriors instead of henge for the 1.25 multiplier until we get access to marble wonders?

We can gain a little less than 2 turns by chopping one forest and whipping one pop into an almost finished warrior. It could pay off because the important thing is to get fast to "multiplier techs" like Pottery (multiplies growth), Writing (beakers), MC (hammers), Education (beakers), etc. Astronomy has increased importance because it enables both Observatories and Laboratories.

Still, there could be better ways of using those worker turns. I will have a closer look at the other test games (and new ones if some of you have time) on Friday before I write a plan.

 

[ShannonCT]

ToA is a good example of a WW that could be whored around to our cities when we are wanting to grow them.

Putting hammers into SH. Or we could research Poly instead of Med and put hammers into Parthenon

It turns out that we can't build either of those wonders, or GL, because Scientific Method has obsoleted them.

What is the additional cost of researching Writing before Pottery? (I.e. with one less prerequisite).

Knowing Preisthood, or AH, or Pottery is enough to research Writing, and knowing the bare minimum number of prereqs mulitplies beakers (the number in the top left corner) by 1.2. Each additional prereq adds 0.2 to the multiplier. So with Preist, AH, and Pottery, the multiplier would be 1.6.

 

[ShannonCT]

I always tech poly on my way to oracle, didn't realise people used the Med route >.<

I usually do to, because of the possibility to build ToA or Parthenon, and because it's a prereq to Mono and Lit. But ToA, Parthenon, and GL are all disabled in this game. The only things useful up that branch are the Epics.

 

[Cactus Pete]

"We'll need to get a library built in 10 turns (have a total of 17 turns before the GP ) after the Oracle is completed and probably have to run 2 scientists immediately until GS. Argues for Writing before Pottery (probably not). Not easy in any case. Will it be worth it (probably so)?" Woke up this morning knowing this needed editing. Sleep-deprived thinking is my excuse.

 

[Cactus Pete]

I usually do to, because of the possibility to build ToA or Parthenon, and because it's a prereq to Mono and Lit. But ToA, Parthenon, and GL are all disabled in this game. The only things useful up that branch are the Epics.

Yes, and Poly costs more beakers, which are helpful in getting to Masonry in time for Oracle.