SGOTM 10 - Xteam

[ShannonCT]

As this is a new version, are there any "improvements" to AI behavior that may cause us grief?

There are changes to AI behavior, but nothing regarding peace making.

 

[Cactus Pete]

gandhi is still ~20 turns from sending out a settler. If we wait for a chance to capture a settler heading south, we can probably pick up 2-3 workers when we really need them. If we start the war now, we probably can't get more than 1 worker why wouldn't we be able to pick off settlers just as in your scenario? and we would need to keep the war going for 30-40 turns to achieve the objective of keeping gandhi confined to the north. I was hoping for a quick war of 10-15 turns where we wouldn't give gandhi much time to organize a counterattack. playing the devil's advocate: If we can prevent ivory from being hooked up, not sure he'd be much of a threat with no advanced units and limited tiles improved.gandhi will talk 6 turns after being declared on and has the highest possible probability of making peace.

. . . . . . . . . . .

 

[ShannonCT]

. . . . . . . . . . .

AI tend not to send settler parties toward enemy units. So I don't what our chances are of capturing a settler when we're already at war.

If we wait for a settler party heading south before we declare war, we'll have time to build some warrior MPs/fogbusters, then hook iron, then build some good counter units for pillaging. A stack of knight/pike/Xbow can pillage one tile per turn. The combination of those three units will be very hard for Gandhi to stop.

If we send in a LB now, it could be killed in the open by another LB and put us in a bad spot.

 

[Cactus Pete]

AI tend not to send settler parties toward enemy units. So I don't what our chances are of capturing a settler when we're already at war. Might be able to deal with this by stationing units on forest (ideally on opposite side of cities from direction we'd like settler to go) to see inside city, and then try to anticipate where settlers would go and put knight out of sight in that direction.

If we wait for a settler party heading south before we declare war, we'll have time to build some warrior MPs/fogbusters, then hook iron, then build some good counter units for pillaging. A stack of knight/pike/Xbow can pillage one tile per turn. The combination of those three units will be very hard for Gandhi to stop. That should certainly allow us to pillage at will, but more concerned with stopping settlements, and it means building multiple units. Also, would Gandhi put both a WE and a longbow with his new settlers?

If we send in a LB now, it could be killed in the open by another LB and put us in a bad spot.

Could happen. Certainly a need to test what Gandhi is likely to do, but if ivory is not improved and there are no more than 2 longbows in either city right now, would think our longbows could get to good defensive tiles (and stifle the fast workers) before Gandhi would be willing to risk attack. Might work to double them on one good tile. Concerned about Gandhi's willingness to chop units, while our hands are somewhat tied.

 

[ShannonCT]

I tried some tests settling near Gandhi. I gave myself a Great Spy to spy bomb him, so I could see what he was building.

Delhi builds LB, barracks, LB, worker. Bombay builds a settler (slowly at Pop2). Delhi's 2nd LB (6th overall) comes around T20-22. The first built LB will escort the worker, although sometimes the fast worker gets ahead of the LB. If I declare war around that time, Gandhi will interrupt the worker build and start building the best available unit in Delhi. If our LB is on open terrain, it will be attacked if Gandhi has 2 roaming LBs in the area.

 

[Cactus Pete]

I tried some tests settling near Gandhi. I gave myself a Great Spy to spy bomb him, so I could see what he was building. That seems a clever and highly useful tactic.

Delhi builds LB, barracks, LB, worker. Bombay builds a settler (slowly at Pop2). Delhi's 2nd LB (6th overall) comes around T20-22. The first built LB will escort the worker, although sometimes the fast worker gets ahead of the LB. If I declare war around that time, Gandhi will interrupt the worker build and start building the best available unit in Delhi. If our LB is on open terrain, it will be attacked if Gandhi has 2 roaming LBs in the area.

Most interesting . . . lots of questions: Were you able to capture the first worker? Does the settler build get interrupted as well? When does the ivory get improved? Does our lb still get attacked if he has some defensive bonus? Will a lb paired with another lb or just an explorer get attacked in open terrain? Have you played far enough to find out where the settler is sent both in and out of war, and any info on Gandhi deciding to attack St. Pete?

BTW, appreciated the info posted earlier on Gandhi's tendencies.

Note to leif: In 22 turns Bombay will expand, and the two forests outside Moscow's fat cross to the SW will no longer be available for chopping.

 

[ShannonCT]

Most interesting . . . lots of questions: Were you able to capture the first worker? Does the settler build get interrupted as well? When does the ivory get improved? Does our lb still get attacked if he has some defensive bonus? Will a lb paired with another lb or just an explorer get attacked in open terrain? Have you played far enough to find out where the settler is sent both in and out of war, and any info on Gandhi deciding to attack St. Pete?

These tests were on a new random map that didn't match the real map. So I rebuilt a map with matching terrain and got some different results:

Delhi built cat, worker, settler
Bombay built missionary, barracks, elephant, theater
Original worker hooked cows, then chopped a forest on the silks and improved the silk. Then both workers built roads to Bombay and hooked southern ivory on T20.
He popped a GA in Bombay and settled it there.
First settler went north to get iron (~T30 with 2 chopped forests) with longbow and elephant escort.
Second settler went south to get wheat (~T50) with pikeman escort.
He also captured a barb city with his elephant.
Around T55, Gandhi had 4 LBs, 1 cat, 1 elephant, 1 pike. He was building another elephant and researching Education.

 

[Mad Professor]

These tests were on a new random map that didn't match the real map. So I rebuilt a map with matching terrain and got some different results:
[snip]
He popped a GA in Bombay and settled it there.

Ouch. We don't want him taking big chunks out of the fat crosses of our cities... Convincing him not to build that theatre would be of some benefit, if we can.

 

[Frederiksberg]

First settler went north to get iron (~T30 with 2 chopped forests) with longbow and elephant escort.
Second settler went south to get wheat (~T50) with pikeman escort.
He also captured a barb city with his elephant.
Around T55, Gandhi had 4 LBs, 1 cat, 1 elephant, 1 pike. He was building another elephant and researching Education.

I think the settler dates and the places selected by Gandhi for new cities are quite important. If they are fairly stable we can determine if there is any gain at all from starting a war. Most likely we can't prevent Gandhi from settling the iron spot and I doubt that we can get get the wheat inside our cultural borders anyway. So we could let him have those spots without any real loss. The question is if we are fast enough to claim the remaining available city sites.

We also need more information. If there are attractive city sites on nearby islands there is no need at all to fight for the not-so-fabulous sites on the western half of our home island. Let's get some work boats out scouting.

 

[ShannonCT]

In the test I did, there was one turn where Gandhi left two workers undefended on one of the silk forests. That is the only place where we would have a good chance to worker steal early. Why don't I move our LB toward that area. If we can catch the workers there, we could consider declaring war. Otherwise, we can wait for the settler that he'll probably send south around T50.

 

[Frederiksberg]

In the test I did, there was one turn where Gandhi left two workers undefended on one of the silk forests. That is the only place where we would have a good chance to worker steal early. Why don't I move our LB toward that area. If we can catch the workers there, we could consider declaring war. Otherwise, we can wait for the settler that he'll probably send south around T50.

Sound like a good idea. I must admit that I don't find an immediate DOW very appealing because it's unclear to me what we gain from it. I prefer to acquire some more knowledge of the surrounding islands before any decision is made.

 

[Cactus Pete]

No guts, no glory, Fred. There are some agressive players on the teams that have been beating us. Think we need to thoroughly explore this option.

Appreciate the time and effort and skill that went into testing, SCT.

In the test I did, there was one turn where Gandhi left two workers undefended on one of the silk forests. That is the only place where we would have a good chance to worker steal early. This is before they roaded to Bombay and hooked up the ivory? If undefended for only one turn, that implies that an Indian unit was on an adjacent tile. Would he attack our longbow (and two workers on a forest tile?) and possibly kill him and get both workers back? Odds on this low enough so we would not need to keep the explorer around to stand with our longbow? Fear Gandhi will not leave workers unprotected if our lb is threatening. If steal not available, could we then declare anyway and prevent workers from getting to ivory?Why don't I move our LB toward that area. If we can catch the workers there, we could consider declaring war. Otherwise, we can wait for the settler that he'll probably send south around T50.

Wondering if our situation at T50, with Gandhi having hooked both ivory and iron, will be that good, unless we greatly distort our development. Right now we may have a greater advantage.

 

[Frederiksberg]

No guts, no glory, Fred. There are some aggressive players on the teams that have been beating us. Think we need to thoroughly explore this option

You have to do better than that . Surely the level of aggression had nothing to do with how we did in our recent SGOTMs. If anything we have been more aggressive than most - if not all - of the other teams.

What I am looking for is a coherent plan indicating how the war will play out, how many hammers we will have to divert to it and what the potential gain is.

 

[Cactus Pete]

You have to do better than that . Surely the level of aggression had nothing to do with how we did in our recent SGOTMs. If anything we have been more aggressive than most - if not all - of the other teams.

What I am looking for is a coherent plan indicating how the war will play out, how many hammers we will have to divert to it and what the potential gain is.

Points well taken, though I'm tempted to quote another oft used cliche' relative to how long a coherent plan remains that once the enemy is engaged.

 

[ShannonCT]

This is before they roaded to Bombay and hooked up the ivory? If undefended for only one turn, that implies that an Indian unit was on an adjacent tile. Would he attack our longbow (and two workers on a forest tile?) and possibly kill him and get both workers back? Odds on this low enough so we would not need to keep the explorer around to stand with our longbow? Fear Gandhi will not leave workers unprotected if our lb is threatening. If steal not available, could we then declare anyway and prevent workers from getting to ivory?

Wondering if our situation at T50, with Gandhi having hooked both ivory and iron, will be that good, unless we greatly distort our development. Right now we may have a greater advantage.

Yes, they built a road on both silks as part of the road to Bombay, before hooking ivory. The only extra unit Gandhi had built at that point was a cat. If he only has one spare unit to attack with, he probably wont attack a LB in a forest. My experience tells me that the AI does not consider the threats to its workers from civs it is at peace with.

If the workers have already roaded those tiles when we get there, it will probably be too late to prevent the hooking of ivory, although maybe not too late to pillage it before an elephant can be built.

If Gandhi plays like he did in my test, he wont have many reserve units at T50.

 

[ShannonCT]

I played up to T19, with some very illuminating turns.

On T15, I spotted an Indian caravel next to the Chinese borders. Gandhi seems to be racing for circumnavigation, heading east. I've bought his map a couple times and revealed some Chinese, English, and American territory. We have no chance for circumnavigation trying to chase Gandhi's caravel east. I think we have to go west and keep buying Gandhi's map every couple turns for 5 gold. If we head north and then west, we can contact the Chinese and English on the way and perhaps buy their maps. A workboat from Moscow can contact the Americans. Every American leader is willing to trade maps at Cautious. Victoria and Mao are willing to trade maps at Cautious too.

Our LB has come across a single Indian worker who has just finished a silk plantation and road on the eastern silk. The road isn't connected to anything, so the worker could be stolen without danger. But if we go to war now, it would probably kill any chance of getting circumnavigation. Gandhi's decision to build an early caravel seems to have come at the expense of a second worker. He seems to have no roads other than on the cows and eastern silk.

 

[Cactus Pete]

Will be a few hours before I can look at save and respond. Hope others will be able to comment in the interim.

 

[ShannonCT]

I notice Gandhi has also reveled part of the islands to the west of our continent. He might have a workboat out exploring as well as the caravel. The map is 64 tiles wide and Gandhi's map has already spanned 36 tiles. With a workboat heading west and caravel heading east, he could circumnavigate in as few as 6 turns if the workboat can find landmasses. At most, he will circumnavigate in 9 turns if his caravel doesn't get blocked by a landmass. Beelining west, our caravel would take 9 turns to get to the current fog line in the west. That means we probably can't circumnavigate without map trading. So maybe the best chance we have is to meet the Americans, Chinese, and English ASAP with our caravel and hope that map buying will get us circumnavigation.

 

[leif erikson]

How about if we head east to meet America and see what lies further east. Not sure I see what going north will get us, unless England has maps of the area to the west?

 

[ShannonCT]

How about if we head east to meet America and see what lies further east. Not sure I see what going north will get us, unless England has maps of the area to the west?

For the purpose of getting the circumnavigation bonus, going east wont help us. Gandhi's already ahead there. And going west is probably not going to work either unless Gandhi's caravel gets blocked by a landmass. The best bet seems to do as much map trading as possible. The English or Chinese might have built a caravel or workboat and sent it west. Maybe the Americans sent a caravel east ahead of Gandhi's.