Scouting

[Aeson]

I think this topic deserves it's own thread.

--------------

First I'll cover maps. When starting a game you get information about 9 tiles most of the time, and a location from the mini map. This gives you your latitude and longitude right off, and an idea as to what type of environment you're in. From this information it's often possible to figure out what type of gamestyle will work best, where the AI's are likely to be hiding out, and where natural barriers (coastline specifically) are going to reside.

This is all possible because the map generator, while not as consistant as some would like (getting stuck on that small island on a pangaea game will do it for most of us), follows definite rules for tile placement. Understanding these rules can give a generalized knowlege of the map even before it's uncovered. The best way to get a feel for the map generator is to just look at a ton of maps... but here are a few guidelines.

Maps generally follow gradients from tundra near the poles, grassland, plains, desert, plains (not as often), grassland, and jungle at the equator. These are of course horizontal bands, and their widths vary according to the weather settings. When you notice these bands tilted one way or another it's almost always because of a shoreline east or west (roughly mimicing the tilt of the terrain band). If you're on grassland, middle latitudes, with plains showing southeast, it's a good bet there is a northwest coastline nearby. Vertical bands are especially easy to spot. Another tell tale sign is starting where there should be jungle (equatorial) and seeing a lot of grassland/forests. Every once in a while there are gaps in jungles, but most of the time it means you have a coastline opposite the jungle that you do see. It's also a good sign that the Pangaea you're expecting isn't, or is of the long snaking variety.

AI placements are even easier to deal with and predict. Starting near the poles in the grassland belt is very common, as is just north or south of the equatorial jungle. The longitude is anyone's guess, but the latitude in this case seems relatively stable. For scouting purposes this is very important. The sooner contact is made, the more turns you have to buy workers (sounds like it's not such a good deal in PtW), and your starting techs have a better chance of being worth something in a trade. The AI's will do their trading without you if they make contacts first, leaving you out of the loop from the very start.

My first scouts (whether that's a scout, warrior, or whatever) almost always take off east and west, unless bounded by coastline. Once they run into an AI, then they head north or south (whichever has more potential) until they get to the jungle barrier, then follow along that edge to preserve movement points for mobile units, and to make contacts in any case. I try not to run into coastlines too much, and use the aforementioned terrain hints to better direct the scouts.

Like I said though, to really get a feel for the map generator, spend some quality time with an expansionist civ and ctrl-shift-q. I think understanding the map generator is going to be one of the most (if not THE most) important skills for serious multiplayer games.

As for random map settings, they become known very quickly if you know what to look for; latitude and width of terrain belts. So 'random' maps will definitely offer a strategical edge to those who can spot those things

Scouts Specifically:

I find them the most useful non-pop unit in a sedentary or no barbarian game. They won't win a game single handed of course, but they will get you the information needed to win, and do so earlier than any other unit. Jags are nice too, but the first scout beats them by 4-5 turns, and Jags often have to take time to heal up after popping barbarians from huts. Even in barbarian infested games, getting 4-5 Scouts out early should net contacts with all the AI's before they all get wiped out. They can also pass (almost) freely through AI territory, gathering information that might otherwise require trading maps to get.

Scout's movement is what makes them special, so don't waste it. Thats why folowing a jungle border is more likely than trudging up through it. You'll end up hitting more huts along the edge than you'd ever find in the jungle. Hopefully one of those huts will give an already slow moving warrior to do that.

Also when you note a hill or mountain on the first move, move parallel to it instead of towards it on your second move. The idea is to end your next turn with that elevated view, which will end your next turn whether your have 1 or 2 movement points left, so stay 2 moves away.

Opening Huts:

Thanks to information given by Firaxis, it's not a lot easier to get those early settlers (and more of them). Remember that before opening any hut (which you'd like a settler out of), make sure none of your cities are producing a settler, you have no active settlers, and your number of cities doesn't exceed the average number per civ. This really makes building an early granary more viable, as your number of cities will be lower for the first 40 turns or so (a granary starts paying off after the second settler or thereabouts). Being expansionist allows the first non-scout build to be a granary, which can be paying off before 2000BC.

If I run into a hut nearby my core, and think the AI's may not have as many cities as I do yet, I'll put off opening the hut. Also it's a good idea not to open up huts until after making a few contacts. That lets you trade your starting techs for other starting techs (which the AI is going to trade for anyways), and then you have a chance at more expensive, later techs from the same huts. I'll leave all the huts anywhere around me alone if I find a first settler and their aren't any AI's nearby.

Huge Barbarianless maps Specifically:

(that's really all you need to know about Scouts on that setting)

I was asked about how I use scouts in this setting, and it's basically the same as any other, only scale the number of scouts with the size of the map (15 or so on a Huge map). Often my first 10 unit builds end up being scouts, just use the luxury slider to keep citizens in line... finding those luxury resources as early as possible will more than make up for the lost gold later on. I send them out on straight lines, and try to never double back. That first scout should be as far away from the capitol as possible at all times, until the end of land is found in that direction.

[GusSmed]

4-5 scouts!? I'm boggling. Of course, perhaps that's because I generally play with Raging barbarians, but on Standard maps, I find that I might buy 1 more scout, and then I'm too busy with other stuff to waste shields on them.

Your points about proper opening of huts are interesting. I've always popped them open as soon as possible, but perhaps the method you describe will make Expansionist more worthwhile.

These days I typically build a Warrior or Scout to scout, a Warrior to keep the second pop point content, Temple, then Settler. The first Warrior may come back for a few turns to keep the 3rd pop point content while the Temple is building, depending on whether I have a luxury in range of my starting city.

I do this primarily to get that early jump in culture. Typically this lasts until about 300 BC or so, even if I build no more culture generators, so cities I take in my first couple of wars remain mine.

Are you saying you build 4x scout, then granary, then settler with an Expansionist civ?

- Gus

[Aeson]

4-5 scouts!? I'm boggling. Of course, perhaps that's because I generally play with Raging barbarians, but on Standard maps, I find that I might buy 1 more scout, and then I'm too busy with other stuff to waste shields on them.

I tend to not build Temples at all, at least until expansion is over with. 1 temple = 30 or 60 shields, one or two is enough to cover the expense on the Scouts.

Are you saying you build 4x scout, then granary, then settler with an Expansionist civ?

On a huge map with no barbs, yes, and usually about 15 Scouts overall. With barbs, I try to get some Warriors out early as well, though there seems to be enough time before barbs start showing up to get a couple Scouts out first. On small and tiny maps it's probably best to only build Warriors, as the extra movement point isn't going to help as much. The starting Scout should be able to cover enough ground.

[ducki]

Fantastic information about the map generator that I never noticed. Thanks a lot Aeson!

Another option, instead of Ctrl+Shift+Q might be (if you're not on Mac) to repeatedly generate maps in the Editor and look at how the Terrain Bands flow.

I never thought of it other than to realize that if I followed a River, I'd end up in either mountains or a coast. Great insights, thanks again!

[punkbass2000]

Indeed, veryy good info. Aeson
I'll never look at scouts the same way again.

[Sheik]

On a huge map with no barbs, yes, and usually about 15 Scouts overall.

That's interesting. I have always made my number one priority building more cities. What is your timeline for city building look like? At what time do you want to have 3 cities? 4? 5? (Sorry if this off topic).

[Aeson]

In most terrain you're going to be able to produce 2-3 Scouts before the food is available to start the first Settler anyways. That's 3-4 Scouts right off including your starting one. On a Huge map, it's likely you're going to get at least 1 Settler from a hut if you follow the hut opening guidelines. Because of the way it works, you can only get those Settlers if your number of cities is at or below the average number of cities per civ, so overexpanding early on can actually be counter productive. It cuts out your chance to make up whatever production and pop you've saved in the interim.

Use this 'down' time to focus your production elsewhere, and build up some pop in your capitol. The luxury slider is key here. There really isn't anything lost by having 3 pop that requires 2 luxury commerce compared to 2 pop requiring 1 luxury commerce, only gained production and population. The tile worked by the 3rd pop should provide it's own commerce necessary or better. This works until you get to about half the number of optimum cities, where corruption starts to really hurt the effect of the luxury slider. Once I get to that number of cities, it's best to focus on getting a luxury hooked up (happiness unaffected by corruption), or start focusing on some garrison units for those cities. The core cities can continue to rely on the luxury slider, and don't need to be defended as the perimeter cities will do that for them.

As for a timeline, I like to double my number of cities every 16 or so turns. The first doubling is almost always by a Settler from a hut when played this way. In the last 5 maps I've played through that far, I've gotten a Settler every time by ~2500AD. That puts the expansion a little behind, but the Granary will start paying off shortly thereafter, making up the difference. Plus I'll still have more chance for a 2nd and sometimes 3rd Settler from a hut by delaying my number of cities and leaving huts nearby my territory.

It usually works out pretty much even by 10AD, where ~100 cities (7 doublings in 108 turns) is my goal (either through peaceful expansion and/or conquest). The difference is in the number of units and improvements that I've been able to build is more by not wasting chances at Settlers from huts, and in the extra population points (and thus commerce) the Granaries have provided.

I'd have to look at the saves, but in my most recent game, I think I came close to doubling the number of units that I had at that point in my old 1.17f Iroquois game, and that was with 40 shield pop rushes.

[ducki]

Originally posted by Aeson
Because of the way it works, you can only get those Settlers if your number of cities is at or below the average number of cities per civ, so overexpanding early on can actually be counter productive. It cuts out your chance to make up whatever production and pop you've saved in the interim.

Correct me if I misread this: It sounds as though you are trading definite expansion (in the form of built settlers) for potential expansion (settlers from huts) coupled with increased exploration and maximized potential trading.

Put another way, you are gambling on Hut Settlers with cheaper, profitable Scouts. On the off chance that you don't pop any Settlers, you end up having to build them anyway, but with larger populations, they build faster, right?


I'm about to start a game geared at trying to generate 5 leaders and was considering using a Military/Expansionist civ... maybe I can test out two theories at once.

Thanks again for the great info!

[ducki]

A secondary question about Scouts, Huts, and Settlers.
With PtW, you can pop a Town - does anyone know if the Settler-in-production flag has anything to do with this?

[Aeson]

Here's a comparison of the two games. The difference in tech is due to rule changes. The AI used to get a tech bonus, now the player gets an increase in tech cost. That and map size. Huge is smaller now, so less huts available. The difference in number of cities is mostly due to using ICS in one and not the other, and map size again played a role as there's not anywhere near as much grassland in the smaller huge maps.

As you can see, % of the landmass controlled is about the same either way. Number of units, and overall commerce, is where a difference shows up (captured catapults not taken into account). Some of this can be attributed to having Industrious in one game and not the other, but switching to a Republic in the Iroquois game would at least somewhat make up for that. It's not a huge advantage, but an extra 10-20 Knights can make a lot of difference, so why not take it?

ps. The number of Scouts produced was about the same in each case, I just lost a lot of them trying to pillage and cause havoc with various AI's after the exploration was done with in the Orangesoda game.

[Aeson]

Oops, forgot the picture!

[Aeson]

Originally posted by ducki

Correct me if I misread this: It sounds as though you are trading definite expansion (in the form of built settlers) for potential expansion (settlers from huts) coupled with increased exploration and maximized potential trading.

Early on, yes. The 3000BC Settler may or may not be there in this case, but a Granary and extra Scouts will.

Put another way, you are gambling on Hut Settlers with cheaper, profitable Scouts. On the off chance that you don't pop any Settlers, you end up having to build them anyway, but with larger populations, they build faster, right?

Yes, with the Granary, the expansion rate should catch up around the 3rd or 4th Settler produced by that city depending on terrain. Considering that you can have 10 or more Settlers produced in the expansion phase (even if that's including raze/rebuilding) by this one city, it should always pay off in the end.

That should leave other cities able to build more units (and specialize to do so... mining cows ect.)

One drawback of a granary first is you lose a lot more by doing a Palace jump. If it looks like that would need to be an option, you can always build regularly, and let the second city build the granary and start pumping out Settlers. I think having 2-3 granaries in your first 5 cities is a good idea on an uncrowded map (I like my 8 AI on Huge). Just the first one if you're with a lot of AI's or have cultural linked starting locations on.

[alexman]

Originally posted by Aeson
I think this topic deserves it's own thread.

Yes, and the expansionist trait deserves a better reputation!

Aeson, an original, well-thought, experience-backed, and well-written post as usual

Let me just add that your scouts' use doesn't end when you have contacts and knowledge of your continent's map. If it's still early, I put my scouts in galleys and send them exploring overseas after they're done at home. You will often find land with unpopped huts overseas, even later in the game, even on already settled continents. Also, spying on other civs is valuable, and scouts are not perceived as a big threat by the AI.

What about exporing at sea? Have you noticed any patterns of the map generator that could be helpful when looking for another landmass?

[ducki]

Originally posted by alexman
What about exporing at sea? Have you noticed any patterns of the map generator that could be helpful when looking for another landmass?

Actually, I noticed what seems to be a "pattern".

You know how it's generally believed that Earth once had a Pangea, and the tectonic plates drifted apart?
You can look at South America, Brazil specifically, and see how that sort of fits as the "missing" piece in Africa.

Look for things like that on your own continent to give you a guess as to where to start looking for passages. Large concavities and convexities are prime spots to search, I've found.

Nothing statistical, mind you, it just seems to work out that way. Look for the place that seems to be "missing a piece" and then go to the "point" or peninsula to it's north or south. Most of the time I'll find a passage, or at the very least, an Island.


I'm sure Aeson's got more, but that's the thing I've noticed.

[Aeson]

I could only say what Ducki said... it does seem that opposite coasts will mirror each other. As for the width of the gap inbetween, I can't think of any way of deducing it, or even if there will be a mirror shore at any given location. Of course if you're on one landmass near the east/west border (of the mini-map), it's much more likely the other landmass(es) will have a crossing opposite that border. Not always of course... one of the AU games had a crossing through the east/west border didn't it?

In naval exploration I tend to just follow some very basic rules. Mainly, scout the perimeter first, and take off (with much praying) in a likely direction that shows a thicker sea border...

It's a good question though, the map generator certainly may have some sort of tell on that count. I just haven't figured one out.