From demigod to klutz

[Lanzelot]

@Lanzelot: As this is a huge map, your main error is too slow expansion. Corruption here is super low, so founding more cities is very profitable. A good goal would be to found 56 cities as 56 cities would have less than 50% rank corruption as a republic with VP and with courthouse and police in the city and even without the police station it would still be 47. While having less than 56 cities getting more via war has an increased priority. When you already have enough cities the civil build up is more of a priority.

Had you founded 30 to 40 cities instead of 20 leaving Despotism earlier might have succeeded.

Well, try and do it faster.
And also: everything 3-4 tiles south-west of the capital was jungle. I wonder how you want to build significantly more than 20 cities with that kind of land.
And also: not the number of towns is important, the number of citizens is. At 390 BC I have ~70 citizens, 1 settler and ~30 workers. And my big army will soon capture enough AI towns...

BTW: of course in my previous post I forgot to mention that now since we have Feudalism, we should of course first upgrade our swordsmen to medieval infantry before going to war. Enough cash for that is available... 20-25 MI will do quite a bit of damage...

 

[walletta]

Well, try and do it faster.
And also: everything 3-4 tiles south-west of the capital was jungle. I wonder how you want to build significantly more than 20 cities with that kind of land.
And also: not the number of towns is important, the number of citizens is. At 390 BC I have ~70 citizens, 1 settler and ~30 workers. And my big army will soon capture enough AI towns...

BTW: of course in my previous post I forgot to mention that now since we have Feudalism, we should of course first upgrade our swordsmen to medieval infantry before going to war. Enough cash for that is available... 20-25 MI will do quite a bit of damage...

Will you target India?

 

[justanick]

And also: not the number of towns is important, the number of citizens is.

The number of properly used tiles is, that includes the city tiles.

 

[Lanzelot]

Will you target India?

Would be the obvious target. The other two neighbors that would be close enough are Carthage (Numidean Mercennaries...) and Rome (Legions...). Fighting those early on is a bit more difficult. (And both of them are pretty strong by now, they both probably already had their GA: Rome had a short war with Spain, though not sure if they had iron connected at that time, and Carthago had been slugging it out with England for several centuries. My scouts up there have already observed Numideans killing English archers and horsemen, so they definitely had their GA.)

 

[walletta]

Would be the obvious target. The other two neighbors that would be close enough are Carthage (Numidean Mercennaries...) and Rome (Legions...). Fighting those early on is a bit more difficult. (And both of them are pretty strong by now, they both probably already had their GA: Rome had a short war with Spain, though not sure if they had iron connected at that time, and Carthago had been slugging it out with England for several centuries. My scouts up there have already observed Numideans killing English archers and horsemen, so they definitely had their GA.)

Yes, agreed. India is best, but not when they have just got the elephants, unless it's a weak India. Of what will your army consist, approximately? Also, Carthage is quite far from your power house economy so sending units up there will be time consuming and possibly weakening should Rome etc decide to pounce. Nor do you have horses, another drawback in the start position is that the nearest horses, while obtainable, are a long way away and that part of the empire, as I found, is very hard to defend, especially if one falls behind in tech (I lost four well-developed cities there in two turns to a stack of German cavalry which consumed almost all my knights in the process).

I have moved on, through my double-javelin thrower-army-disaster, to a new game featuring more rivers and forests which I have chopped into many libraries as I came out of anarchy into republic respectably early (5th century BC). I found a good path was:

warrior code, trade something for alphabet, then research writing, philosophy, code of laws as a free tech and then shoot for Republic (if I go for CoL first then someone always wins the race to Philosophy). You can basically buy and trade for everything else in the course of getting those. In fact, I end up with a lot of money at the end of all this. Now thinking of war with England but without repeating the previous mistake of failing to prepare adequately. I have iron, two sources of horse and two of wine with no one to trade with yet. I know only three other civs and need to get some curraghs going once I plant a town by the sea, probably the next objective while growing some pop. and watching the money roll in, per the Justanick method.

 

[Lanzelot]

The number of properly used tiles is, that includes the city tiles.

The extra city tile is compensated for by the fact that a settler costs two citizens. So that evens out to exactly zero. Basically all my 70 citizens are working fully improved tiles. But if I had more towns, they would only be "jungle camps". Or I would have had to build more workers to clear the jungle faster, but that would then also have reduced the number of settlers I would be able to produce.

Ok, there is probably space for 3-4 more useful towns up north, but as I had enough good tiles for my citizens to work on closer to the core, I did not feel the urgency to settle those. And I repeat: as long as I have not seen the "many small towns approach" actually be significantly faster, I don't believe it...

(With "significantly" I mean at least 10 turns. Of course lack of concentration in the MM or a mishap with barbs can easily cost a turn or two. But that wouldn't necessarily prove the superiority/inferiority of one or the other approach. And also the "luck factor" shouldn't count. For example, if you finish Philosophy at around the same time +/-1 turn, and none of the AI has it yet at that time, that is purely luck. Or as you can see, when I finished Philosophy, I was able to buy Literature and Map Making for it. However, CoL was not yet discovered. If one of the AIs had gone for CoL instead of Lit or MM, I could have bought that and would have been 12 turns faster. This is also purely "luck".)

 

[Lanzelot]

India is best, but not when they have just got the elephants, unless it's a weak India. Of what will your army consist, approximately?

India still doesn't have Feudalism, so they would have to get three more techs, before they can build those nasty Elephants: Feudalism, Monotheism and Chivalry. That will take some time, in any case longer than my remaining preparations will take. (I expect to hit them within the next 10 turns.)
I currently have 17 swordsmen and 10 warriors. Some of the warriors are still far away, mapping the English territory or returning from the south, so I expect to field ~20 Medieval Infantry.

warrior code, trade something for alphabet, then research writing, philosophy, code of laws as a free tech and then shoot for Republic (if I go for CoL first then someone always wins the race to Philosophy).

Yes, this is also something that struck me while playing this map. I was much surprised. Usually I play standard maps (can't remember the last time I played huge, must have been 10 years ago, when I didn't have a wife and kids yet...), and with a good start position on Demigod I manage the "full slingshot" in about 50% of the cases. But here it was absolutely impossible to get Philo first (3 civs had done it before me), not to mention the full slingshot... (Ok, the start position was not that good, but I would have expected that getting Philo first would always be possible...)

So it looks like the huge map actually favors the AI. Can it be, that only the human player pays the increased tech price, while the AI still pays the "standard" price?! Another explanation would be: on standard maps the AI soon runs out of space and then gets into wars with their neighbors, which slows them down... Here on the other hand they had lots of space around them, so they were able to take full advantage of their better expansion-capability (a settler costing only 14 food and 21 shields for them).

 

[walletta]

India still doesn't have Feudalism, so they would have to get three more techs, before they can build those nasty Elephants: Feudalism, Monotheism and Chivalry. That will take some time, in any case longer than my remaining preparations will take. (I expect to hit them within the next 10 turns.)
I currently have 17 swordsmen and 10 warriors. Some of the warriors are still far away, mapping the English territory or returning from the south, so I expect to field ~20 Medieval Infantry.

What about spears to hold the captured towns? Do you figure on not using catapults either? I must admit, I never grasped the importance of upgrading swords to MI but, without elephants, I can't see India having anything to stand in their way other than numbers and she probably hasn't got those either. Do you know a sure way of making India declare war on you? It seems to make a big difference to the people.

Yes, this is also something that struck me while playing this map. I was much surprised. Usually I play standard maps (can't remember the last time I played huge, must have been 10 years ago, when I didn't have a wife and kids yet...), and with a good start position on Demigod I manage the "full slingshot" in about 50% of the cases. But here it was absolutely impossible to get Philo first (3 civs had done it before me), not to mention the full slingshot... (Ok, the start position was not that good, but I would have expected that getting Philo first would always be possible...)

So it looks like the huge map actually favors the AI. Can it be, that only the human player pays the increased tech price, while the AI still pays the "standard" price?! Another explanation would be: on standard maps the AI soon runs out of space and then gets into wars with their neighbors, which slows them down... Here on the other hand they had lots of space around them, so they were able to take full advantage of their better expansion-capability (a settler costing only 14 food and 21 shields for them).

I never managed to get to Philo first after taking time out to research CoL.

Btw. one difference in our play is that I was moving the starting settler one square to the NW at the start to get off a bonus tile and on to plain grass, luckily finding a hut and then scoring one of the following, which I rank in order of preference:

A a free warrior (to make contact with other civs 5 turns earlier)
B ceremonial burial
C 25 gold
D a map of the region

I wonder whether the better players would put D at the top. It wouldn't surprise me.

 

[justanick]

The extra city tile is compensated for by the fact that a settler costs two citizens. So that evens out to exactly zero.

Given that the city tile consumes no food but gives food it does not even out for long. If that extra food of 56 cities instead of 20 cities is used for 36 additional scientists that is 108 additional beakers per turn. If this map had no barbarians this might make the difference between the slighthot and waiting 30+ more turns for republic.

And I repeat: as long as I have not seen the "many small towns approach" actually be significantly faster, I don't believe it...

That is understandable.

Yes, this is also something that struck me while playing this map. I was much surprised. Usually I play standard maps (can't remember the last time I played huge, must have been 10 years ago, when I didn't have a wife and kids yet...), and with a good start position on Demigod I manage the "full slingshot" in about 50% of the cases. But here it was absolutely impossible to get Philo first (3 civs had done it before me), not to mention the full slingshot... (Ok, the start position was not that good, but I would have expected that getting Philo first would always be possible...)

That shocked me, too. 2 days ago i wanted to play this map. Then i noticed that this a huge map, which reduced my motivation a lot. Then i saw that you failed at getting philo first. So even if i could get philo first chances for a slingshot are not very good, which decreases fun a lot more.

So it looks like the huge map actually favors the AI.

I beg to differ. It favours Demigod+ AIs in the ancient age if goody huts are allowed. Compared to a standard map the amount of goody huts is increased by a factor of 2.56, so AI will get almost 2.56 more techs from goody huts and thus also techs, that are much more expensive. Furthermore the human player has to cope with the fact that researching those techs will cost 66.7% more, so in total that is factor of ~4 in favour of AI, but it only applies during the ancient age.

Can it be, that only the human player pays the increased tech price, while the AI still pays the "standard" price?!

I am rather sure AI has to pay the full price of 40 beakers per base point as it should at regent. As this is demigod the human player has to pay 57.14 beakers per base point, but this is not what you referred to.

Another explanation would be: on standard maps the AI soon runs out of space and then gets into wars with their neighbors, which slows them down...

I would argue in the opposite direction. The map size itself does not favour AI, the increased amount of civs does. This increases the amount of tech trades that is utilized by AI a lot. If you have 15 enemies instead of 7 chances are one of them will have philo before you and the trading between them makes things even worse.

Start a huge map with no barbarians at all and only 1+7 civs and the larger map will favour the human player. Well, at least if he plays in the proper way. That is to be read my way.

 

[walletta]

In that case, I wonder why I was advised against playing demigod on tiny maps? I enumerated several advantages of doing so before but can't remember them all know. They included:

1 greater chance of a GSL
2 greater chance of the slingshot
3 greater chance of building wonders
4 easier to handpick opponents with the wrong characteristics for the map (another shameless admission)

 

[justanick]

In that case, I wonder why I was advised against playing demigod on tiny maps?

You were? The smaller the the map the earlier AI will attack you, while on bigger maps it is easier to win. But this easier is very much counteracted by increased adminstrative hassle. So i cannot advise to big maps althouth they are indeed easier to win. The way i see it the world sizes would need to be renamed:

tiny(60x60) -> small
small(80x80) -> standard
standard(100x100) -> big
large(130x130) -> too big
huge(160x160) -> much too big.

1 greater chance of a GSL

Only true under some circumstances. More civs increase the chance someone will have a tech before you. If however you manage to be much bigger than you biggest enemy your chances for SGL increase. Big maps make it more likely that there are some backwards civs, but those are not relevant for SGLs anyway.

2 greater chance of the slingshot

That is true, but goody huts and increased number of civs counteract it. Especially goody huts are almost game breaking so better deactivate barbarians.

3 greater chance of building wonders

Only true if you are leading in tech or there are less enemies. Those 2 conditions may contradict each other.

4 easier to handpick opponents with the wrong characteristics for the map (another shameless admission)

If there are more enemies there will be more enemies that have not the wrong characteristics for the map.

The map size is of very secondary importance. Donnot let yourself be distracted by it.

 

[walletta]

There only need to be a few well-adapated civs for them to take off and be away before you can do anything about it. This is another disadvantage of a huge map - lack of control over the farther reaches of the game where things play out among the AIs and an uncontrollable monster emerges.

ETA and on the GWs, even with a tech deficit, there are so many of them to go round that the 3 AIs cannot build them all. Hence, even with a lag, you can usually knock up something in the ancient age. The GSL thing is also important. In my one and only (unofficial DG) win I managed to erect the Temple of Artemis (with the GSL), the Great Wall and Sun Tsu's art of war which meant every new city was founded with a temple, walls and barracks. How cool is that?

 

[Lanzelot]

Given that the city tile consumes no food but gives food it does not even out for long. If that extra food of 56 cities instead of 20 cities is used for 36 additional scientists that is 108 additional beakers per turn.

I don't think you will be able to build 56 cities, even using your approach. Remember: after the initial 7-10 towns you immediately get into jungle, and in a jungle town you can't build settlers. (Growth stops at size 2, unless you clear a jungle first to provide at least one +2fpt tile. So you need to build lots of workers, which in turn initially reduces even more the number of settlers you are able to build.) I think that the 20 towns I managed under these circumstances are quite ok. I can imagine that using your approach you will probably end up with 30 towns, perhaps not even that much. And it won't be that many in the early phase where it counts most for the Philo-race?!

So it looks like the huge map actually favors the AI.

I beg to differ. It favours Demigod+ AIs in the ancient age if goody huts are allowed. Compared to a standard map the amount of goody huts is increased by a factor of 2.56, so AI will get almost 2.56 more techs from goody huts and thus also techs, that are much more expensive. Furthermore the human player has to cope with the fact that researching those techs will cost 66.7% more, so in total that is factor of ~4 in favour of AI, but it only applies during the ancient age.

So you agree that the increased goody hut rate and the 66.7% increase in tech cost are a big factor in favour of the AI. But both of these are a direct consequence of the hug map size, so the huge map favors the AI.
q.e.d

I would argue in the opposite direction. The map size itself does not favour AI, the increased amount of civs does. This increases the amount of tech trades that is utilized by AI a lot. If you have 15 enemies instead of 7 chances are one of them will have philo before you and the trading between them makes things even worse.

The increased amount of civs definitely increases the probability that one of them will get Philo before you do. Especially as there are also 2.56 times the number of huts from which Philo can be popped...

But I can rule out an increase of tech trades in this case: on a standard map the AI will meet much earlier, so tech trading starts earlier. On Pangaea you have 7 civs trading with each other pretty fast, and on continents like two groups of 3-4 civs.

Now look at this map: in the north we have a group of 3 (Carthage, England and Germany), in the south a group of 4 (India, Rome, Byzanz and Spain), and on the other continent a group of 3 (Babylon, Greece and Persia). 4 more civs I haven't met yet, most probably they form a fourth group on a third continent. Even now in 390 BC there is only one contact between these 3 groups: England has met Byzanz. And this is pretty recent. For most of the game, the 3 groups have been isolated from each other, which is in fact the reason why I was able to gain such huge amounts of techs and gold whenever I discovered a new tech.
So on this map we have a similar situation as on a standard continents map: groups of 3-4 that are isolated from each other. The differences are only: a) there are more groups (but this does not necessarily increase the tech pace inside each group) and b) the members inside each group meet each other a bit later than on standard (which actually slows down the tech pace a bit in the beginning!)

Of course all that may be different on a Pangaea map.

In conclusion I think that the reasons are not as obvious as it may at first appear. More likely it is a combination of several factors that make this map so hard to play:

• (relatively) poor start position for the human player
• lots of goody huts (direct consequence of map size)
• lots of space for the AIs to fully unfold their increased growth potential compared to the human player (direct consequence of map size)
• the additional tech prize increase due to map size makes the total number of beakers that the human player has to "catch up", even bigger: e.g. a tech for 300 beakers would cost the AI 210 beakers, so the human player has to catch up 90b. (Which may be easy to do with better strategy and better micro management). However if that same tech costs 500 beakers (66.7% more), then the AI gets if for 350b, so the human player has to "catch up" 150 beakers! (Which is of course much harder, as the number of towns and citizens that the human player can generate in a given amount of time, does not depend on map size, so is the same in the early phase, no matter whether we play standard or huge). (direct consequence of map size)

 

[Lanzelot]

So even if i could get philo first chances for a slingshot are not very good, which decreases fun a lot more.

Well, I guess that is a matter of personal taste... For me it is no fun, if everything goes "too smoothly". It's more fun when my plans fail and I have to fight a tough uphill-struggle...

 

[justanick]

So you agree that the increased goody hut rate and the 66.7% increase in tech cost are a big factor in favour of the AI. But both of these are a direct consequence of the hug map size, so the huge map favors the AI.
q.e.d

Well, i do reject the assumption of activating barbarians in the first place. That is the players fault. But even with huts activated this advantages is annihilated in the middle ages. In the long run large maps tend to be easier.

lots of space for the AIs to fully unfold their increased growth potential compared to the human player (direct consequence of map size)

I am not sure that is an advantage for AI. True, there is 28% more space per civ. But techs are 66.7% more expensive, so once AI is in the middle ages the human player can close in more easily than on smaller maps simply because AI will need more turns per tech. Also lower corruption tends to favour the human player.

the additional tech prize increase due to map size makes the total number of beakers that the human player has to "catch up", even bigger: e.g. a tech for 300 beakers would cost the AI 210 beakers, so the human player has to catch up 90b. (Which may be easy to do with better strategy and better micro management). However if that same tech costs 500 beakers (66.7% more), then the AI gets if for 350b, so the human player has to "catch up" 150 beakers! (Which is of course much harder, as the number of towns and citizens that the human player can generate in a given amount of time, does not depend on map size, so is the same in the early phase, no matter whether we play standard or huge). (direct consequence of map size)

I estimate that a human player can in the mid run generate a higher amount of beakers per turn compared to the tech cost at larger maps. There is more space that is not claimed by AI and that, that is claimed by it, might be easier to take. Larger maps offer more opportunity to make AI marginalize themselves.

 

[Lanzelot]

I aggree to all of the above, in the long run the human player will overtake the AI and a large map makes this easier. (One reason being that the human player has more time to develop a powerful empire undisturbedly and does not run the risk of being overrun by the AI's starting units...)
However, in the current discussion we only consider the early phase of the game, basically the time until the discovery of Philosophy (or until the discovery of Republic, if we want to make the period a bit longer), and during that time the large map seems to favor the AI (or to make it harder for the human player, depending on how you want to look at it). Not?

 

[walletta]

Weird. I started a standard size demigod continental Maya game, managed the slingshot and went into anarchy in 1500 BC (no ***t) came out, played a few turns and realised I was still in anarchy! Two things are possible: (1) I clicked the wrong button or (2) there is some earliest possble date barrier that I don't know about.

ETA, I meant to say I was still in despotism, not anarchy. That really would be a bummer.

 

[justanick]

I bet on 1. Or at least not 2.

 

[Buttercup]

 

[walletta]

Another question, Lanzelot, if you don't mind. Assuming you play on (I appreciate you may not be interested) would you divide your army or send it forward in one great, irresistible lump? Also, do you check to see whether India is building/has built any wonders and, if so, prioritise your attack accordingly?

It just occurs to me that the timing of setting up an embassy might be usefully co-ordinated with an invasion, in order to get that one turn look at what is being built in the capital to see whether it's a GW and, if so, how many turns before completion because it might be worth delaying until it's finished. A good plan might be to place embassies in Civ's that share a frontier with the target Civ in order to bribe them into joining in so as to weaken the target further.