Tutorial: Mapmaking for Civ 4

[Charriu]

Next up some topics about the map at large:


Terrain Distribution

One thing you're missing is Peaks.  Don't need many, maybe just 5-10%.  Actually, I'd probably go for 5% everywhere except High Desert and High Jungle, where I'd go for 10%.

Also - although the tundra looks really nice, it's really hard to balance; I would either reduce it to a couple tiles, or eliminate it entirely (to be added to offshore islands or the like).  Fits better in a SP game where people can just ignore the land until they've settled everything else.  I guess you could keep it if you changed the tundra region to just a nominal amount of tundra, 10% or less, just enough to be noticable.

(...)

On forests and hills - be generous, especially with the forests.  I would sprinkle at least 25% forest into all biomes, except High Desert, and possibly aim more for 40-50% in the non-Jungle areas.

I would probably aim for 25% hills, uniformly in all biomes.  Maybe 30-35% in jungle/desert.  Players like hills too, although unlike forest you can have too many hills.

Also, can you sprinkle in some Oasis tiles into the desert and desert transition areas - maybe 5% in transition and 10% in High Desert?  That will cover all the terrain features except floodplains, which can't really go in until rivers are placed anyway.

Finally, from a 'natural' perspective - can you add some nominal amount of desert to the grassland and plains biomes?  Maybe 2% in grass and 3% in plains?  Just enough to add a little interest, not enough to shift the value of the terrain really.

Source

Luxury placement

In order for luxury placement to be "fair," either 1) every player must have ~equal access to wine and equal access to the same number of Calendar luxuries, OR 2) one must assume the necessary tech paths are balanced enough that wine is ~equivalent in value to a Calendar luxury and players need only have access to equal numbers of any Classical luxuries.  In case 1, we would need to make sure everyone without a wine start can get access to another wine tile early (among other things) but would have to do that anyway unless we made wine unavailable on the starting landmass (as in PB37).  In case 2, there is essentially no problem.

Dunno if this argument is sound though.  My entire MP experience of late has been bouncing ideas back and forth with JR4 in his PB37 thread, which has been tremendously fun and enlightening for me, but not the same thing as actually playing.

Source

[Charriu]

Next we have some posts around the topic of starting locations:

Resource placement in the initial BFC

(An observation for future mapmakers: if you do not want people to move away from pre-designed capital spots, ruining the map balance, you should give them decent starts. Because two food resources on the opposite sides of the capital, two spaces away each, is an utter shit.)

Replying to this here because it's an interesting point, but it's more to do with the variance between players due to pic method than the start itself.

The balance between double grain and double animal, without using the mod to change the length of time taken to improve each tile, animal starts will necessarily need to have the food tiles further apart. This has a secondary advantage that it enables double animal starts to dotmap more widely with second cities to share food, something double grain lacks.

The greatest variance is going to be in start techs, which comes down to what civs you are lucky to roll. Some players will be able to go BW first, others will have to go food tech>Mining>BW and be stupidly late to Slavery in comparison. Is this a fault of the start? No, because the map has no control over who rolls Mining. If the starts were all split food, so one animal, one grain, then this would increase the tech cost and increase the profit in wander towards whichever food resource you have the tech to improve, and see if you luck into a second food resource of the same type you can settle for. This is exactly what happened in PB37.

So, how should this problem be solved? It could be, that starts were balanced after the picks, to ensure that each player had double food res unlocked to their civ. This shafts those civs that don't start with any food res. Could give all players additional start techs, which also helps speed up the start. This would be my prefered method, but it was shot down in the sign up thread.

What happens if you give triple food of the same type? Or two of one, one of a different type? In the former case, it can make the early game a little too straightforward as the right move becomes to split off a preimproved food resource, and it speeds up the game a bit more, but it's not necessarily a wrong decision. It does affect the balance of IMP though, because you can go worker>settler>settler, split of two food res and then stack whip anger in multiple cities, but it's not a one right more necessarily.

Source

Do not under any circumstances give double deer as starting food. It needs AH, it's a straight penalty.

I have to add to this, that Krill propably meant as the only starting food.

Source

Tile yields of the first connected resources

My comments on specifics pretty much come down to two themes
- A 6 food, 0 hammer tile is generally considered more valuable than a 5-1 tile or a 4-2 tile, especially for a capital.  Using the   judiciously generally lets you convert 1 food into more than 1 hammer, at least post-Granary - and your capital is the city that usually does most of the whipping.  I would try to modify underlying terrain (hills, plains, etc) to make sure each player has at least one 6 food tile, even though that makes the starts a little more boring.

Source

Thought it through and did some quick mentalsims.  Unsurprising conclusion:  Krill is correct.

In detail: Basically, food > hammers at the start only if it speeds growth onto improved tiles or if a super-early BW allows you to abuse Slavery.  Even in the latter case, 1 food < 2 hammers until you have a granary, which is a virtual eternity away at T0.  Since 5f is good enough to outgrow your worker's tile improvements at game start even if you work max hammers until the first improvement, a pair of 5/1s (or a 6/0 and a 4/2) will beat a 6/0 and a 5/0 over the early turns (e.g. you can get a slightly faster first settler) and then the 1 extra food never catches up because it's competing with a snowball and 2 extra hammers.  It only barely and temporarily beats the hammers alone, even with a granary:  Only until the capital permanently grows past size 4.

Source

[Charriu]

Lastly some meta advise:

Sending out starting screenshots:

(those who have threads; do we PM them to the other three?) along with the first rolled picks.

I'm inclined to say no PMs.  Four reasons, not sure which is more important:
a) Pitboss is a big time investment.  If they can't be bothered to invest enough to create a thread, then perhaps the time to look for replacement players is now, not in 15 turns when they start missing turns.  We've got at least two people who have expressed interest after signups were locked down; should be possible to replace people if need be.
b) Since we're doing this hot-potato style, PMs make it much less clear what the state of play is.  Posts in a thread make it clear to everyone who reads along what's going on.  No risk of someone responding to only me that they want a re-roll, right when I'm too busy to do anything about it, holding things up for a week.  Or worse, responding to everyone that they want a re-roll, then 5 minutes later to just me saying 'actually, never mind, I will keep leader X'.
c) PMs are more work to compile information from.  More chances for things to go wrong.  Also more of our time spent on logistics instead of map-improvement or lurking.
d) If you start acting like a game admin now...people will assume you've volunteered to help sort out the double-move situation that crops up in June .

Source

Designing for the player not the designer.

Second...you don't want to constrain their choices too much.  A nudge in the direction of a more interesting game is one thing, but ultimately we want the story of the game to be the story of the players, not the reveal of the mapmakers' cleverness.

Source

[Krill]

Next we have some posts around the topic of starting locations:

Resource placement in the initial BFC

(An observation for future mapmakers: if you do not want people to move away from pre-designed capital spots, ruining the map balance, you should give them decent starts. Because two food resources on the opposite sides of the capital, two spaces away each, is an utter shit.)

Replying to this here because it's an interesting point, but it's more to do with the variance between players due to pic method than the start itself.

The balance between double grain and double animal, without using the mod to change the length of time taken to improve each tile, animal starts will necessarily need to have the food tiles further apart. This has a secondary advantage that it enables double animal starts to dotmap more widely with second cities to share food, something double grain lacks.

The greatest variance is going to be in start techs, which comes down to what civs you are lucky to roll. Some players will be able to go BW first, others will have to go food tech>Mining>BW and be stupidly late to Slavery in comparison. Is this a fault of the start? No, because the map has no control over who rolls Mining. If the starts were all split food, so one animal, one grain, then this would increase the tech cost and increase the profit in wander towards whichever food resource you have the tech to improve, and see if you luck into a second food resource of the same type you can settle for. This is exactly what happened in PB37.

So, how should this problem be solved? It could be, that starts were balanced after the picks, to ensure that each player had double food res unlocked to their civ. This shafts those civs that don't start with any food res. Could give all players additional start techs, which also helps speed up the start. This would be my prefered method, but it was shot down in the sign up thread.

What happens if you give triple food of the same type? Or two of one, one of a different type? In the former case, it can make the early game a little too straightforward as the right move becomes to split off a preimproved food resource, and it speeds up the game a bit more, but it's not necessarily a wrong decision. It does affect the balance of IMP though, because you can go worker>settler>settler, split of two food res and then stack whip anger in multiple cities, but it's not a one right more necessarily.

Source

Do not under any circumstances give double deer as starting food. It needs AH, it's a straight penalty.

I have to add to this, that Krill propably meant as the only starting food.

Source

Tile yields of the first connected resources

My comments on specifics pretty much come down to two themes
- A 6 food, 0 hammer tile is generally considered more valuable than a 5-1 tile or a 4-2 tile, especially for a capital.  Using the   judiciously generally lets you convert 1 food into more than 1 hammer, at least post-Granary - and your capital is the city that usually does most of the whipping.  I would try to modify underlying terrain (hills, plains, etc) to make sure each player has at least one 6 food tile, even though that makes the starts a little more boring.

Source

Thought it through and did some quick mentalsims.  Unsurprising conclusion:  Krill is correct.

In detail: Basically, food > hammers at the start only if it speeds growth onto improved tiles or if a super-early BW allows you to abuse Slavery.  Even in the latter case, 1 food < 2 hammers until you have a granary, which is a virtual eternity away at T0.  Since 5f is good enough to outgrow your worker's tile improvements at game start even if you work max hammers until the first improvement, a pair of 5/1s (or a 6/0 and a 4/2) will beat a 6/0 and a 5/0 over the early turns (e.g. you can get a slightly faster first settler) and then the 1 extra food never catches up because it's competing with a snowball and 2 extra hammers.  It only barely and temporarily beats the hammers alone, even with a granary:  Only until the capital permanently grows past size 4.

Source

My original post was following the debacle that was the PB37 start, of triple food, requiring all three different start food techs. That broke the snake pick.

The tldr is: 2 hammer plant, two 6 yield tiles, both tiles give a minimum of 6fpt at size 2 ie pig and cow, corn and wheat, and freshwater fish are the only legitimate starting food if you are tightly balancing the map. Then you have to give a number of grass hills, and same number of forests for each start. Rivers need to be considered, and frankly a river can be worth as much as an early silver tile between starts due to trade route with second city and worked tiles, it's quite important.

If you are going with a random style map and only balancing the starts, not the map, like was requested in both PB41 and PB42, then things change again.

The minimum drops to double 5 yield food resources and a plains hill: however, if you do not give people something more than this it is likely that moving becomes a viable option and the map balance will break on T0.

The options are: river and useful tiles ie grass hills, grass and flood plains to grow and work, an additional food resource, or a pre-calandar happy. The latter can only be ivory or potentially furs: metal happy breaks commerce in early game but AH happy comes later, yet players can move away from this and settle back for this happy.

The reality is that if you give players a low yield start on a random map, and by low yield I mean two 5 yield tiles requiring different tech, players will likely wander in the hope they find another food resource requiring the same tech as the food resource they move towards. So then the option is to limit the ability to move with forests and hills, or provide increased vision as originally requested in PB37 which was decided was impossible to facilitate: I disagreed then and I still disagree, but then I think the PB37 map was an abomination.

Personally, I don't see why starts should be any worse than double 6 yield food, river, some forests and a 2 hammer capital. The starts are capable of being solved before we start the game. The capitals are just where we start and they facilitate the expansion, the growth that leads to players interacting. Make them good, don't be stingy and let players get on with playing the game.

Also, if anyone gives triple food, triple food type again, I will hunt you down, spear you, and use your entrails to fertalize a crop field. Double food type is bad enough on slowing a start down compared to a single food type start as it delays BW by up to 20 turns.

[Krill]

Also, double seafood starts are difficult, best to not make them either. Single 6 yield fish is fine as the food can go straight into a worker, and the path to the second food resource being hooked finishes in a similar time frame, civ and trait dependent. Double seafood is slower as the food can't be dumped into the second workboat, and in best case scenario with a 3 hammer tile, the capital hits size 3 before you can even start the worker and it doesn't finish until around t20 given none exp, and double inner ring freshwater lake fish.

This is about the only scenario that double seafood is comparable to other starts. And then it gives a massive commerce advantage: none fin gives up to 40 commerce above a double land food, fin start gives up to 60 none river tile starts to turn 20. So basically, it's not worth doing.

[superdeath]

Giving everyone silver at the start is also bad i assume? or is it ok to do something like that, so long as everyone is equal?

[Krill]

We play with relatively stable settings: monarch difficulty, and this is for a single reason: it gives a stable set of conditions to weigh up in game decisions. Units and buildings cost the same, tile improvements cost the same, city maintenance varies by map size as do tech costs, but this is acknowledged by some as a problem for games on huge maps.

The point being, everything else is kept constant and the variable is tech costs and the turns to unlock different things needed to progress through the game, giving windows of opportunity to make certain plays, or be taken advantage of by opponents. When we play civ, we are constantly assessing and making decisions within this framework.

Giving people a commerce tile at the start drastically alters that balance and creates a new framework: most of the time it creates one right choices and this is what detracts from the game: one right choices mean we have less to choose, fewer options with which to create strategies.

Also, define equal when players have different traits civs and start techs. Players can never be equal, only roughly equivalent: give everyone a silver at the capital and everyone grows to work it and doesn't whip off it, but some players will now be forced to grab 2 different food techs and then Mining, and will be delayed in working the commerce tile compared to another player that started with one right food tech and Mining: it affects civ balance. In a snake pick it limits reasonable choices, and in a random pick process makes a major part of the game be decided by RNG.

[Charriu]

Lately I've been thinking about stone vs marble. I know that in the past there were many games, where one half of the players has only stone and the other half only has marble. But is that really fair, to have?

[RefSteel]

Lately I've been thinking about stone vs marble. I know that in the past there were many games, where one half of the players has only stone and the other half only has marble. But is that really fair, to have?

I would say no, not in the recent versions of RtR - and depending on what you mean by "really fair."*  Castles are currently valuable economic buildings for players with stone (which as I recall provides +100% production speed for both walls and castles, the only buildings for which a resource provides more than +50%) and can be built everywhere the defensive bonus and/or trade route, culture, and espionage would be of value, unlike wonders, which can only be built once apiece.  Moreover, marble's value has dropped significantly with the changes to MoM:  Originally one of Marble's strongest wonders, MoM now gets no bonus from marble at all, while its expiration at Nationalism - meaning Taj can't be combined with it - reduces the potential strength of another strong marble-boosted wonder as well.

*- I think there's something to be said for assymetrical balance, even with the understanding that it might never really be as balanced as giving everybody sufficiently-similar areas would be, and recognizing that everyone will complain about a hand-balanced map where some players' advantages are attempted to be balanced against others' different advantages:  There is no objective measure of balance (on which everyone can agree) that compares number of food resources to number of unique luxuries to number of high-yield tiles to availability of trade route islands to sheer accessible territory to availability of good food-sharing second cities to conveniently-placed strategic resources to [etc ad nauseum].  But insofar as the assymetry is acceptable, it can make for much greater map variety, without resorting to plain map scripts and [spoilers from PB41 and PB42 redacted-in-advance].

[Charriu]

Distance between players

A small thing you have to be wary of if you want it to be super super balanced : 13 tiles in diagonal and 13 tiles straight are not exactly the same gameplay wise. In the first case, if your opponent is say 13 tiles directly to your NE, you can plant a city 4N of your cap, and it will still be 13 tiles from your enemy. Same for 4E. Only if you settle 4 tiles NE are you 9 tiles away from your enemy's capital. If your enemy is 13 tiles N of your cap, and you plant a city 4NE of your cap, that city is 9 tiles away from the enemy cap. Same with a city 4 tiles N and 4 tiles NW of your cap. Which means that you'll expand towards your enemy more rapidly.

On distances to neighbours: The way the game treats distances is a little counter-intuitive. I *think* that a site 14 tiles N of your capital and 7W + 7S of an opponents won't say "Liberate" for either of you - ie the game considers that to be in the middle even though it looks a little odd to our eyes.

For most of the other purposes in the game diagonal border pops give the best example of how distance should be calculated, so for a distance of 4 tiles east the equivalent diagonal is 3Ex3N.


For 5 tiles east the equivalent is also 3Ex3N, but with a slightly different pattern.


So I'd use the border pops version when figuring out diagonal distances - is it Pythagoras' theorem? I can't remember. Good luck!

I guess it is Pythagoras. 3t NE would be sqrt(18) which is between 4 and 5 in distance, so it would make sense that both 4 and 5 tiles E would be characterized as 3 tiles NE (4 tiles NE is sqrt(32) which is higher than 5, 2 tiles NE is sqrt(8) which is lower than 3)

I guess it is Pythagoras. 3t NE would be sqrt(18) which is between 4 and 5 in distance, so it would make sense that both 4 and 5 tiles E would be characterized as 3 tiles NE (4 tiles NE is sqrt(32) which is higher than 5, 2 tiles NE is sqrt(8) which is lower than 3)

Almost pythag, but not quite:  It actually checks the distance between tiles on both the X axis and the Y axis (i.e. East-West and North-South).  Then the "total distance" it calculates is equal to the longer of those two distances plus half of the shorter distance, rounded down.

Creating border regions

• You can already see that I placed gems and silver between players to raise the tension. I plan to make the area around those very fertile, but hard to defend.
  
• Place 1 gold resource in the middle of the island chains between player to encourage some competition.
  

The following quote, from your other thread, is a bit out of context, but I think it applies to some extent even to this, and the reason is something I've been trying to think of a way to express for some time:

All judgement on game balance is from player experience though. Don't do things because you think it would be interesting to play.

My feeling is that expressly placing important resources between players to raise tension is unnecessary and at this point kind of trite:  There will be tension in the regions between players anyway, just in the nature of competition for cities and territory, as long as the land isn't complete trash.  With potential super-cities and unique resources in the in-between zones, you need to be especially careful to make the capitals themselves very strong so that players aren't tempted to play the meta and try to move their starting settler toward another player for the expected super-site.  Then if you try to discourage people from moving their starting settlers with a ring of barren and/or jungled land around the capital, you get this weird pattern of amazingly attractive land and amazingly unattractive land like a sine wave, or if you try to keep things reasonably even, you end up with an incredibly lush map.  Not that those map types can't be fun to play, but the recent resumption of interest in script-based maps is partly due to the predictability of these forms (and the over-lushness of both, and especially the second) - also due to other types of predictability of course.  So doing this isn't wrong - it's kind of normal for ~hand-drawn maps, I think - but I'd kind of prefer to see less of it.  (In fact, arguably, it's most useful as a mapmaker's crutch:  You don't want to "force" players toward one opponent or the other, and making sure that there are attractive prospects toward all neighbors is an easy way to ~equalize things.)

In particular, there's an issue with making the lush or resource-rich area hard to defend:  It tends to encourage more extreme play like settling past the contested resources with a border fortress city.  A map where extreme plays and risky gambits (including a starting settler move) are possible and potentially viable can be cool, but one which encourages the same type for each player is much less so.

Erm.  Hopefully that long-winded explanation of my thinking is of some use to somebody (or will spark debate from someone who disagrees).