SGOTM 06 - Xteam

[leif erikson]

I just finished with WorldBuilder. The save attached below has all the information that has been revealed thus far built into it, terrain resources, etc. As I am still learning how to properly use WorldBuilder, the save has the Carthaginians nearby, the distance set by the random generation of the save. However, instead of Mali, the save placed Cyrus where I think Mansa should have gone. Mali is across the map and I couldn't get it to work right when I moved Mansa.

The save:

 

[DJMGator13]

Since this Warlords game is using the BTS dlls would that have modified the starting units or is that controlled thru a different part of the programming? In BTS, AIs start with an extra scout. Otherwise I'd say Gyathaar did some playing in the Worldbuilder.

 

[leif erikson]

Just for fun, I went into WorldBuilder again on each of the saves I created. I looked for the AI start locations and counted the tiles that were between them and us. The ratio is relatively constant at 14 by 7 tiles. I think those distances are relatively consistant with Scouts appearing at turn 10 and 13?

I would apply that in our case by looking 14 tiles east and 7 north or south, more or less, from our starting location. That's my guess on where the AI Capitals may be.

 

[Cactus Pete]

Agree that those scouts are not built (hadn't thought it through).

Wondering if you were using AW in your test games?

The more I ponder this, the more cautious I become.

"I would apply that in our case by looking 14 tiles east and 7 north or south, more or less, from our starting location. That's my guess on where the AI Capitals may be." I'm reluctant to rush across this distance with axes, especially given the quality of our capital. We should have little difficulty disposing of the AI on our continent before they get Alpha, much less longbows.

Since we are not likely to capture an enemy capital quickly, we should not do anything to greatly delay settling a second city, perhaps near the unimproved gold.

Do agree that having an axe wandering around an AI city is likely to ****** his advancement (and add to our knowledge).

 

[ShannonCT]

2 unpromoted axes are probably not enough if the AI has more than 1 archer so I guess everything depends on the expected window of opportunity. How much faster would we get the first axe if we postpone the 2nd worker?

If we switch to an axe as soon as copper is mined, We can have an axe on Turn 38. I agree that if we rushed with 2 axes and found two archers in the AI capital, we would have to refrain from attacking and wasting two axes. If we found an archer and a warrior ... that might be worth attacking with one and checking the odds for the second one if the first one dies.

EDIT2 - The power graph on the submission page looks interesting. Almost as though some teams are chopping Axes around turn 30?

Looking at the score graphs, it looks like most other teams have a higher population than we do, meaning they probably were slower in getting a worker. I dont know if it's even possible to have an axeman on turn 30.

There is some redundancy in the early axe plan. It isn't a complete loss if we come across an AI capital we can't capture, we can still park an axe in the vacinity and lock them down and deny worker improvements.

I ran a quick test game with huge fractal map, Prince setting with 10 Hannibal opponents. I artifically made contact with them all (DoW with AW setting) around turn 13 to mimic our current game. The first AI archer was built on turn 39. 6/10 of the AI's had an archer by turn 59. For us, if we have 2 axes at turn 45 we would need luck on our side to find a capital with warriors only by turn 60. The reward is large so it is worth testing/discussing further.

This would seem to be an argument in favor of sending out a single axe as soon as possible. If he finds an archer, he just sits tight and will be really tough to get rid of. The AI will just sulk. If he finds only warriors, then it's go time.

I ran a test with 10 Mansas. Made contact after 10 turns. First skirmisher was built on turn 42. By turn 59, 4/10 had skirmishers. The 6 that didn't have skirmishers were all coastal, had concentrated heavily on fishing and workboats, and hadn't even started building a skirmisher. This is the kind of Mansa we should be hoping for.

Just for fun, I went into WorldBuilder again on each of the saves I created. I looked for the AI start locations and counted the tiles that were between them and us. The ratio is relatively constant at 14 by 7 tiles. I think those distances are relatively consistant with Scouts appearing at turn 10 and 13?

I would apply that in our case by looking 14 tiles east and 7 north or south, more or less, from our starting location. That's my guess on where the AI Capitals may be.

Mansa's scout did appear from the NE, so you may be right that he is on the east coast to the north. If we were to send a lone axe out into the fog, this might be the best direction.

 

[leif erikson]

I find myself agreeing with CP. Two Axes ready at turn 45, requiring 14 turns to arrive at a city means turn 60.

Played another test game which confirms the distances. In this random game, the Civs on the island with me were Cyrus, Qin Shi and Roosevelt. Cyrus had Archers when I found his city on turn 50. Qin Shi just popped his first Archer on turn 58. Roosie's first Archer appeared on turn 69.

EDIT - Checked the land area around the cities. The more sea based food resources the city had, the more behind they seemed. Cyrus had mostly land tiles with good food resources, Rice and Cows with a Sugar and a Clam and a river. The other two had no river.

 

[DJMGator13]

Looking at the score graphs, it looks like most other teams have a higher population than we do, meaning they probably were slower in getting a worker. I dont know if it's even possible to have an axeman on turn 30.

Could they be trying an archer rush? But they would have to research 2 techs first. I don't have my Warlords tech cost in front of me but in BTS that would be a combined 100 beakers for the 2 techs versus 120 for BW.

 

[Cactus Pete]

If the AI has revolted to slavery, I'd expect him to whip an archer as our axe(s) approaches. Has either revolted yet?

If we switch to an axe as soon as copper is mined, We can have an axe on Turn 38. I agree that if we rushed with 2 axes and found two archers in the AI capital, we would have to refrain from attacking and wasting two axes. If we found an archer and a warrior ... that might be worth attacking with one and checking the odds for the second one if the first one dies.

I think the odds would be at least 60% of winning -- 20% on the first turn and over 50% on second with slightly wounded axe vs. wounded but promoted archer. Even so, I don't like the risk reward.

This would seem to be an argument in favor of sending out a single axe as soon as possible. If he finds an archer, he just sits tight and will be really tough to get rid of. The AI will just sulk. If he finds only warriors, then it's go time. Sending out a single axe certainly has merit. We might consider sending one in the most likely direction and send a warrior in the other, hoping to send a chariot not too long afterwards in the other as well.

I ran a test with 10 Mansas. Made contact after 10 turns. First skirmisher was built on turn 42. By turn 59, 4/10 had skirmishers. The 6 that didn't have skirmishers were all coastal, had concentrated heavily on fishing and workboats, and hadn't even started building a skirmisher. This is the kind of Mansa we should be hoping for.

Did you use AW and did you make sure the AI had made contact with a rival on same continent?

Mansa's scout did appear from the NE, so you may be right that he is on the east coast to the north. If we were to send a lone axe out into the fog, this might be the best direction.

Here's a first draft of a non-rush/axe-harassment scenario:

Put off second worker production for a few turns.
Put citizen on bananas, and work on a warrior 7 turns until copper is mined
Once mine built, switch to axe and move worker to chop forest.
Shortly before chop comes in, send one warrior out to explore in direction don't plan to send axe, if no enemy units are threatening
We should get axe and then warrior with chop and be very close to size two
Move worker to another forest
Grow city to size2 and begin a second axe to the point that we can pop rush or use chop for it (rather than second worker) if necessary, then switch build to second worker and move first worker to cows after he completes his chop

Once axe is off to harass, it might be smart to keep a warrior fortified in London to discourage AI forays.

In a couple of my test games, AI warriors did enter city limits with bad intentions, though much more often they go around. I know this was before 2000BC, but don't recall exactly when.

 

[Jimmy Thunder]

I ran a test with 10 Mansas. Made contact after 10 turns. First skirmisher was built on turn 42. By turn 59, 4/10 had skirmishers. The 6 that didn't have skirmishers were all coastal, had concentrated heavily on fishing and workboats, and hadn't even started building a skirmisher. This is the kind of Mansa we should be hoping for.

I just ran two test games, Prince, Big and Small, Always War, with 10 Mansas, made contact after 10 turns.

TestOne:
First skirmisher built on turn 46. By turn 59, 4/10 had skirmishers. Only one was still building skirmisher. Of the non-skirmisher Malis, 3/10 had triple warriors in capital and 3/10 had double warriors.

TestTwo:
First skirmisher built on turn 43. By turn 59, 4/10 had skirmishers. Four were building their first skirmisher. Of the non-skirmisher Malis, 6/10 had triple warriors in capital.

So we have more than an outside chance of crushing a capital around turn 60.

I like ShannonCT's idea of a single early axe. I'll do some more tests assuming we build an axe on turn 38 and have it next to a Mali capital 15 turns later. Of the warrior only defences I'll get a feel for the possible combat success. Remember, even if we find archers our rush isn't in vain.

Shutting down one (or two) civ(s) while we build a 2nd (or 3rd city) before making axes in earnest is also a viable route.

 

[ShannonCT]

If the AI has revolted to slavery, I'd expect him to whip an archer as our axe(s) approaches. Has either revolted yet?

Checking the turn log, no, neither had revolted after I met their scout, and I dont think they could have revolted before turn 13.

Did you use AW and did you make sure the AI had made contact with a rival on same continent?

In my 10 Mansa test, everyone met each other on turn 10.

Here's a first draft of a non-rush/axe-harassment scenario:

Put off second worker production for a few turns.
Put citizen on bananas, and work on a warrior 7 turns until copper is mined
Once mine built, switch to axe and move worker to chop forest.
Shortly before chop comes in, send one warrior out to explore in direction don't plan to send axe, if no enemy units are threatening
We should get axe and then warrior with chop and be very close to size two
Move worker to another forest
Grow city to size2 and begin a second axe to the point that we can pop rush or use chop for it (rather than second worker) if necessary, then switch build to second worker and move first worker to cows after he completes his chop

Once axe is off to harass, it might be smart to keep a warrior fortified in London to discourage AI forays.

I'll test this out and compare dates with our other option. Sounds reasonable though.

 

[Cactus Pete]

"I like ShannonCT's idea of a single early axe. I'll do some more tests assuming we build an axe on turn 38 and have it next to a Mali capital 15 turns later. I suspect that one axe will be able to eventually overcome even three warriors, but I'm skeptical that archers won't be whipped. Can you test for this, somehow putting AI in slavery?Of the warrior only defences I'll get a feel for the possible combat success. Remember, even if we find archers our rush isn't in vain. Yes, a single axe rush with harassment as an alternative may be a good compromise.

 

[Capt Buttkick]

lurker's comment: Looks like a great start
I don't think I can contribute with much in this game. Looking forward to joining you in a future BSGTOM, though

@Leif: Go Vikings!

 

[Frederiksberg]

Sounds like the axe rush has some chance of success with speed as the key factor so maybe we should do some more testing with the single early axe. I guess that the AI may not be smart enough to switch the build to archer and pop rush if it's building something else like settler. What is your experience with this?

 

[rrau]

I haven't done any tests, but my gut feeling is taking out an AI early is worth delaying a second city. With the number of AI we've met so far, we're going to get dogpiled. If we can reduce the number of AI that will really help our game later. In my test games, we can only have 3 or 4 cities without building courthouses and that's just not enough to really put out the units when all the AI and barbs are attacking. Even trying to build more cities with courthouses doesn't work because they're dead weight for so long (draining both treasury and units to defend). I think our expansion will need to come from whipping courthouses in high population AI cities after we capture them, but that will be a ways down the road.

 

[leif erikson]

There are a lot of ideas floating in our thread. I would like to summarize a bit and see if we can focus a bit more.

First, an early goal is to cripple, or eliminate, the two AI civs we have met.
An Axe rush has been proposed as well as a compromise of a couple of single Axes, each headed for the two AI capitols.
Second, we have to develop our Capitol as it has the potential to be a very powerful city. What is the cost to development of our civ by going for an early score? Is the trade off worth it?

What else do we need to test?

I guess that the AI may not be smart enough to switch the build to archer and pop rush if it's building something else like settler. What is your experience with this?

I think this is true enough. In one of the test games, Roosevelt had three Warriors in his capitol while I had a Warrior nearby. He was last to Archers because a Settler showed in his capitol before the Archer appeared. That told me he was working on the Settler and couldn't start the Archer until he finished it.

If we try an Axe rush, it would be against one of the two civs?

 

[leif erikson]

lurker's comment: Looks like a great start
I don't think I can contribute with much in this game. Looking forward to joining you in a future BSGTOM, though

@Leif: Go Vikings!

You could if you could find that Warlords disk!!

How come they didn't put a Santa hat on that rocker?
Better PM Mistfit and see what he can come up with.

to you and yours Capt.

 

[Frederiksberg]

Here's a first draft of a non-rush/axe-harassment scenario:

Put off second worker production for a few turns.
Put citizen on bananas, and work on a warrior 7 turns until copper is mined
Once mine built, switch to axe and move worker to chop forest.
Shortly before chop comes in, send one warrior out to explore in direction don't plan to send axe, if no enemy units are threatening
We should get axe and then warrior with chop and be very close to size two
Move worker to another forest
Grow city to size2 and begin a second axe to the point that we can pop rush or use chop for it (rather than second worker) if necessary, then switch build to second worker and move first worker to cows after he completes his chop

What is your plan for the 2nd axe? I don't think we need it for defense that early - in my test game the AI didn't start sending stacks of archers before around 1000 BC. A 4th warrior might be good enough if we only intend to use him to fog bust a 2nd city site. On the other hand, if we have two axes outside an AI Capital we might be able to attack and destroy settling parties since they are usually composed of 2 archers and a settler.

Once axe is off to harass, it might be smart to keep a warrior fortified in London to discourage AI forays.

In a couple of my test games, AI warriors did enter city limits with bad intentions, though much more often they go around. I know this was before 2000BC, but don't recall exactly when.

I saw an enemy warriror entering cultural borders around 2700 BC in my game and he was a woodsman 2 which created a lot of problems with worker protection due to the double move in forests.

There are 6 rivals in total and if more of them are close we should normally meet them real soon. Unless, of course, Gyathaar has modified the map such that nearby AI capitals block the access of civs further away. In which case we don't get fewer opponents by taking out an AI early.

 

[ShannonCT]

I suspect that one axe will be able to eventually overcome even three warriors, but I'm skeptical that archers won't be whipped.

It is highly probable for an axeman to beat three warriors fortified in a capital. In about 5 tests, the worst case was one where the probabilities for defeating the 1st, 2nd, and 3rd warriors were 99.3%, 97%, and 87% respectively. If you can pick up 1 XP from killing an animal on the way there, you're in even better shape because you get a promotion after the first warrior kill. And if the AI capital is on a hill, it's also not a problem because then the first warrior kill guarantees a promotion, and the second warrior kill often gives another promotion.

Sounds like the axe rush has some chance of success with speed as the key factor so maybe we should do some more testing with the single early axe. I guess that the AI may not be smart enough to switch the build to archer and pop rush if it's building something else like settler. What is your experience with this?

It seems that if Mali has researched archery, then it will start building a skirmisher, and if it has researched BW and adopted slavery, it will whip a skirmisher if threatened. We will be able to see if Mali revolts to slavery. Our success really depends on Mali not having researched both archery and BW by the time we get there. And in our 10 Mansas tests, if Mansa starts on the coast, he almost invariably goes for fishing first and builds a workboat. The landlocked Mansas had archers much earlier, and were adopting slavery much earlier. Of course we don't know whether Mansa is coastal or not, and we probably won't find out until our axeman gets there.

As has been said, if we find skirmishers in the Malinese capital, it is not a waste to have built the early axeman. 1 axeman fortified in a forest is going to be a lot harder to get rid of than 2 warriors.

So we have to decide now whether to:

1) mine the copper first and get an axeman out on turn 38

2) mine the copper first for the purposes of early expansion, but foresaking the early axe rush

3) mine the gold first to get Animal Husbandry faster.

I'm biased toward using the copper somehow. This will probably give us a nice advantage over other teams who didn't get BW or a worker so early.

 

[Frederiksberg]

So we have to decide now whether to:

1) mine the copper first and get an axeman out on turn 38

2) mine the copper first for the purposes of early expansion, but foresaking the early axe rush

3) mine the gold first to get Animal Husbandry faster.

I'm tempted to go for option 1 because the potential reward is high. It seems that getting a 2nd axe does not improve our chances much because he comes late and unpromoted axemen do not have good chances against skirmishers. So maybe we should just build a single axeman now and then basically return to our original plan of fast expansion and tech. We will be facing opponents with lots of units combined and it would be nice to ensure that we have the quality advantage.

In my test game warriors were adequate for locking down the AI workers. I simply positioned the warrior in a forest 2 tiles from the city and they were never attacked. Positioning a warrior adjacent to the city was more dangerous and the AI would every now and then attack with an archer. Attacking and destroying settling parties require more units - at least two axes to dispose of the typical settler,archer,archer stack.

 

[Cactus Pete]

I'm tempted to go for option 1 because the potential reward is high. It seems that getting a 2nd axe does not improve our chances much because he comes late and unpromoted axemen do not have good chances against skirmishers. So maybe we should just build a single axeman now and then basically return to our original plan of fast expansion and tech. We will be facing opponents with lots of units combined and it would be nice to ensure that we have the quality advantage.

I'm in agreement on option 1.QUOTE]

"What is your plan for the 2nd axe? I don't think we need it for defense that early - in my test game the AI didn't start sending stacks of archers before around 1000 BC. A 4th warrior might be good enough if we only intend to use him to fog bust a 2nd city site. On the other hand, if we have two axes outside an AI Capital we might be able to attack and destroy settling parties since they are usually composed of 2 archers and a settler."

My plan for the second axe is not to finish building him for ten turns, unless we are threatened. His partial build is an insurance policy that we would cash in inten turns. Once he is bult to the point of completion in one turn with a pop rush, then we go back to a second worker and get our capital to size three with cows, copper, and gold.