Guide for designing MP maps.

Discussion of all aspects of multiplayer development: unit balancing, map development, server development, and so forth.

Moderator: Forum Moderators

Post Reply
bert1
Posts: 240
Joined: December 6th, 2006, 10:39 pm
Location: Morecambe, UK

Guide for designing MP maps.

Post by bert1 »

From time to time people want advice on how to make fun balanced multiplayer maps. There is information scattered through the forums on this, but it's hard to find. This thread is intended to be a dedicated home for such information.

If anyone finds any good gobbets anywhere, perhaps you could copy and paste them in this thread, or link to them. Better still, share your own insights into MP map design.

I will start things off with some simple points:

Broad design considerations:

1. Purpose/design concept. What is the point of the map? Does the map cater for a need which isn't already covered adequately by existing maps? Is there something original/interesting about it?

2. Is it fun to play? Most discussion on maps is about balance. But it's no good having a balanced map if it is ugly, plays boringly and is never played. What makes a fun map? I'm not sure - horses for courses, I guess. Isar's Cross is an example of a flawed map that nevertheless many people find fun to play.

3. Dynamics. Some maps produce slow, bogged down games, others quicker. I don't know much about this. I get the feeling that widening maps and placing villages far to the east and west (assuming the map is north vs south) spreads out the battle front and stops games stagnating by allowing flanking movement and ZOC'ing. A short battle front means that a string of units can block the map, prevent ZOC'ing and allow wounded units to retreat and heal more easily. Depending on what kind of game you want, these are things to consider.

4. Aesthetics. Maps should ideally look nice.

Balancing considerations:

5. Terrain. Avoid large clumps of similar terrain anywhere where the main fighting is likely to happen, as they will advantage the faction that likes that terrain. Mountains especially unbalance a map quickly because Dwarves have high resistances and HP already, and giving them lots of 70% terrain puts them into orbit. Putting mountains next to villages is generally avoided because of the pressure Dwarves can put on the village. Trees are obviously favourable to Elves, but as Elves don't have the hp and resistances of Dwarves, this is less of an issue. Consider not only the defensive values of the terrain, but also its effect on the movement of the various factions. (See below.)

6. Village distribution. Make sure each side can get its villages reasonably easily. Test for early village stealing possibilities with a fast Gryphon (or bat if there are caves to go through). If any villages are stealable, either move them so they are not, or make sure that the defending player has a good chance of trapping and killing the scout unit that stole the village. Proximity of the leader can be useful in this. Essentially, the initial village grab must be fair and yield equal income. For an unusual and dynamic village distribution, see Loris River.

7. Player 1 advantage. (More applicable to 1v1). Player one moves first, and so is potentially ahead of player 2. Therefore, completely symmetrical maps are generally to be avoided, although you will see that some default maps are close to symmetrical (see Den of Onis). This is OK because if the starting keeps are far enough away from each other, and villages are not too out on a limb, then the player 1 advantage is negligible, as by the time combat begins, both sides have had enough time to arrange their units and grab their villages. Very good terrain, keeps and villages are generally not placed in the middle of the map such that player one can get to them all first.

To offset player 1 advantage (if it is significant), you can use village placement to make player 1 run a bit further, or maybe backwards a bit in order to grab all his villages. This is a fine-tuning thing, I suspect, and hard to get right without playtesting. You can also place 'slowing' terrain to inhibit his movement, although you don't want to disadvantage player one for the whole game, just the first couple of moves, so maybe not too much of that.

8. Movement. Mobility is a big advantage. Some factions are more mobile than others. Drakes especially so, as they can mostly fly and Saurians have high movement points. They may need to be slowed down by caves/impenetrable terrain. Don't disable them too much, though, as mobility is what they depend on. You can give them bits of sandy terrain as a boost (40% for drakes and 60% for saurians, and no movement penalty). Too much sand will mean that drakes rule. If you cover the map with hills, northerners might well be advantaged, as they can move well over hills. Again, all this must be playtested and tweaked.

9. Size. I don't know much about this really. I've never tried, but really big maps are apparently hard to balance because of cheap bats pinching all the villages. Similarly, very small maps are difficult to balance because the player 1 advantage and possible village grabbing becomes much harder to sort out. See my my vid on Isar's Cross for an analysis of that map.

--------

Doc Paterson has kindly qualified this last point (9) with the following wisdom:

"The primary difficulties in balancing small maps are not so much playerside oriented as they are faction-oriented: They deemphasize mobility and narrow movement options, often creating situations where Drakes and Undead cannot run far enough when they need to, or where Dwarves are advantaged, etc. (You can probably imagine any number of other factionally unbalanced scenarios that could result from limited mobility and an overemphasis on tight battle lines.)"
Last edited by bert1 on January 29th, 2008, 9:09 pm, edited 1 time in total.
Good is simply that which is willed. - Eugene Halliday
User avatar
Doc Paterson
Drake Cartographer
Posts: 1973
Joined: February 21st, 2005, 9:37 pm
Location: Kazakh
Contact:

Post by Doc Paterson »

A previous post on map identity/character/aesthetic:
Doc Paterson wrote:
Hans wrote: I liked the "castle in the clouds" effect. Bot ok, the keep is a bit much. I'll remove it.
I like the "castle in the clouds" concept as well, and I hope that you are able to preserve it through the phases of balancing. One thing that I believe in strongly (in terms of map design) is that a map should really try to have its own character, independant of its balance; that it should suggest a place that is distinct from other places. There are hundreds of grass/water/mountain/forest maps out there, most of which use the same old basic aliases: the green grass for grassland, the pine forest for forest, the brown or sandy dirt for roads, etc, etc. I like it when someone tries to convey a place that existed first in their mind, and filters that through the map editor with distinctiveness as the first priority. I'm not saying that I think a map is good just because it is different....

http://s153.photobucket.com/albums/s230 ... t=dood.jpg

....but rather that, in my opinion a truly great map will have, as a defining quality, a unique and creative character. I'm always excited to look at maps that use the editor pallete in interesting new ways- For example, a map that uses the "stone" GL alias as the dominant terrain type (over the green grass), or a wildly asymetrical map that delves into the more complicated and/or intuitive qualities of factional balance.

It may be a common misconception that I add only perfectly balanced maps to mainline, that the balance is more heavily considered than the artistic identity (forgive me if "artistic identity" sounds too bogus or pretentious when applied to wesnoth maps; these things are somewhat difficult to descibe. :) )That's not really true- I'm looking for maps that are interesting, aesthetically and functionally, that suggest strange, wild, unique or evocative places. (We have plenty of green grass/green tree/green hill/ stone mountain maps, and these maps are good to have. They form a foundation, a base to push off from when trying to create new things.) I'm reminded of certain user maps that I found to be particularly inspired; F8's "Chilled Caves" and "Underworld", Velensk's "Titan Hall," Nickname's "Dying World," Sapient's "Tomb of Kings," and of various users who've tried to make maps around the ideas of volcanos or ruins or deserts. These things are very very hard to balance, but as I said, it can be done; the sense of place does not have to be destroyed by the overall restrictions of factional and playerside balancing. In summary, a map that makes it into mainline does need to be balanced for all factional combinations and playersides, but it also needs to have an equally strong level of distinctiveness.

Balancing this character with the "play" qualities of the map can be a very trying process. Other players can help, to a point, but really, playing the game intensely for a few months is a much better education. When you experience the factional interactions for yourself, map balancing will become more and more intuitive.

At any rate, what I mean to say in regards to this map is that I think that it has some of that creative spirit, in spite of its many factional imbalances. You should keep at it. Keep tinkering, and play as many games with good players as you can. Maybe try to make that central concept of the castle in the clouds even more defined. Study the default 1v1s and 2v2s...Note the general terrain patterns, and ask yourself why they may have been done in this way. You'll probably get a lot of advice, some of it rigid, some it it vague, and it's easy to interpret people's advice in terms of absolutes. The important thing, in my view, is to take that advice in without letting it discourage you or dull the creative impulse.
I will not tell you my corner / where threads don't get locked because of mostly no reason /
because I don't want your hostile disease / to spread all over the world.
I prefer that corner to remain hidden /
without your noses.
-Nosebane, Sorcerer Supreme
User avatar
Doc Paterson
Drake Cartographer
Posts: 1973
Joined: February 21st, 2005, 9:37 pm
Location: Kazakh
Contact:

Post by Doc Paterson »

This discussion developed in response to the original version of a map posted by Salafar (now greatly altered, and appearing in the first post of page one).

http://www.wesnoth.org/forum/viewtopic. ... sc&start=0

I'm posting this because it touches on the issue of pressure imbalances. (Probably one of the most common problems that mapmakers have, as it's significantly more complicated than "Let's count the hexes to see if a quick scout can steal a village....") Many of the issues referred to are no longer present in the map on page one, but I think they're described well enough to give the reader a decent mental image of the original version. The details/coordinates etc. aren't particularly important to the overall message.
Doc Paterson wrote:
Jami wrote:On the original map a 8 move unit from P2 could capture the central villages in the north and south on turn 2. While P1 was 11 hexes away from them. That would be a P2 advantage wouldn't it? Look at the reverted version, as a knarlgan with gryphons, a UD bat, or a Glider you can still cap the south-center village on turn 2 as player 2 before P1 can.
It's not nearly so simple. Only 4 units can be recruited on turn 1. Send those ideal P2 units toward the central south target and you'll find that they'll always lose 13,20 on P1's second turn, even if they take 11,17. (We're assuming two scouts from each side heading south here.) P1 can leave 6,16 open and take care of any P2 scout trying to take it with their leader (the P1 leader need only move one hex from their starting location on the previous turn to set up this trap), moving southern scout number 2 to cover 9,20. 3-2 village edge in the south for Player 1. This is made even worse by the fact that the P1 leader is a mere 2 moves away from any P2 scout that would try to take 11,17, strongly favoring a quick 4-1 swing in the south. Not a P2 advantage at all.

Even if P2 scouts could flat out steal them, with none of these complications whatsoever, they'd still probably die because P1 is going to hit them with more units than they (P2 that is) have, close to a 2:1 ratio, a turn before an equal ammount of P2 units arrive. The ratio is amplified because we're talking so early in the game here. It's not as extreme in the North, but P2 scouts would be likewise pressured, and the southern situation would far outweigh it, regardless.

If you play a map that "forces" P2 to get a lot of scouts, just to prevent the possibility of theft, P1 can just play it safe from the start, get a lot of standard combat units, march towards the villages in question and kill or drive off the scouts, who will be far less effective later in the game (loads of wasted money for P2).

There's little point in talking about the old version of the map, but I wanted to explain that P1 advantage is a lot more than P2 scout units being able to make it to X or Y location.
Jami wrote: as it stands right now, P2 has the ability to get more villages than P1 on turn 2 if they recruit the right units. Advantage player 2. Look at the map with the water in the centre and count hexes if you don't belive me.
(I take it we're talking about the new version here.) Everything that I said above applies here. P1 will still take a 3-2, possible 4-1 village lead in the south. P2 has a chance for equal village distribution (considering all villages on the map) if they get a quick scout in one particular hex. There is a lot of pressure for P2 to get 4 scouts. If they do that, and P1 suspects as much and moves towards those villages with stronger, more standard units (which will arrive very quickly relative to when the P2 scouts arrived, as P1 is a full turn ahead), the P2 scouts will be driven from the area or killed, their villages taken by durable units, and the game will be all but over. So no, there definitely was never a version of this map in which P2 had any sort of advantage.

* * * * *


Below is a segment from a brief discussion with Glowing Fish, regarding a map he'd made. It addresses some of the same points as above.
Glowing Fish wrote: I did spot the two villages at 12,19 and 15,18 being too close, and have modified the map so that the two keeps are equal distances from the ford.
Doc Paterson wrote: People thinking that things ought to be "equal distances" is the primary cause of player-1 advantage. Do you forget that player one is an entire turn ahead, and will practically always win a race to a central area that is equidistant from both players' starting points?

I haven't seen how you changed the map, but the only way you're going to keep those two villages from being stolen is to make it so that P2 can take both on their first turn of movement (this is only because this map is so small, and wouldn't be true for larger maps). This could create another problem:

If you make it so that P2 has to recruit scouts to keep gryphons, drakes, ghosts, etc from stealing their villages, you force P2 into an unequal guessing game. If they play it safe and get scouts, they'll be significantly disadvantaged when combat occurs later in the game.
Last edited by Doc Paterson on February 1st, 2008, 8:45 pm, edited 4 times in total.
I will not tell you my corner / where threads don't get locked because of mostly no reason /
because I don't want your hostile disease / to spread all over the world.
I prefer that corner to remain hidden /
without your noses.
-Nosebane, Sorcerer Supreme
User avatar
db0
Posts: 400
Joined: January 3rd, 2006, 8:39 pm
Location: Somewhere Far Beyond...
Contact:

Post by db0 »

I suggest you also start a wiki page on this and link it from "Create"
User avatar
Doc Paterson
Drake Cartographer
Posts: 1973
Joined: February 21st, 2005, 9:37 pm
Location: Kazakh
Contact:

Post by Doc Paterson »

A discussion that started in reference to some loose guidelines suggested by eyu100:
Doc Paterson wrote:
eyu100 wrote:Here is an attempt:
Loyalists underpowered/overpowered: add/remove grassland
Northerners underpowered/overpowered: strengthen/weaken all other factions, add/remove hills
Drakes underpowered/overpowered: add swamp, add water in important locations, add chasm, make map bigger, add snow, add sand; make map smaller, more grassland, remove swamp, less snow, less sand, less chasm, less significant water
Rebels underpowered/overpowered: add/remove forest
Knalgans underpowered/overpowered: add mountains (especially next to villages), add hills, make map smaller, add sand, add swamp; remove mountains, remove hills, make map larger if small, remove sand, remove swamp
Undead underpowered/overpowered: add chasm, add grassland, add cave; remove chasm, remove grassland add cave

P1 advantage: give P2 another village (or two if the problem is extreme)

Try the faction balancing methods first. Apply the terrain and changes relatively evenly.
A lot of this is really bad, particularly the whole premise of "do this to increase/reduce X faction's power." You can't just say "I'm going to add forest to power up only the elves"- There will be complications and consequences in all of the other factional matchups- though they may be vague and/or hard to describe. "Add grassland to power up Loyalists" will have massive influence on every other factional matchup, an influence that is cumulatively much, much larger than any sort of general edge that you attempt to give to only Loyalists. Ken Oh's example with the ghost is too simplistic as well- adding "good terrain" around vilages has a much wider-reaching impact than any that it would have as an anti-ghost measure. I'm not, by the way, saying that some villages shouldn't have adjacent terrain- just that thinking in the way that you described tends to create more problems than it solves. There should really be little trouble removing a ghost from a village, given the right units.

Back to eyu100-What you said about fixing P1 advantage is pretty off as well- The method you advocate generally leads to P1 having to defend villages that are unreasonably far from their keep. Winning a race to a village, while very significant, is not the same thing as holding that village for the majority of a game. It is true that winning the race to a supposedly "contestable" village, placed equidistant from both keeps is 99.99 percent of the time abuseable by P1, given an otherwise even distribution. You do not want a village that P1 "has" to hold to protect their economy that gets harder and harder for them to hold as turns accumulate (because it is so far from their keep). If you want to mitigate P1 advantage, bear this last point in mind, and begin by designating which villages are intended for which player, and measure the distances, making sure that P1, with gryphons for example, can't steal any of them without being blocked from or beaten to the village. What you need to be particularly wary of is a surrounding terrain setup that could fascilitate an easy rotation of a wounded P1 thieving scout to healthy P1 village holding unit.

What you said about putting mountains next to villages to power up Knalgans is way off- There's more on this issue elsewhere in the forums.
eyu100 wrote:I get a lot of this. I guess you would have to try to adjust each faction many, many times until you get it right. Another way to get rid of P1 advantage is to get rid of all important things near the center (you could make the center one big buffer of grassland and keep all the villages easily uncontested. )
Doc Paterson wrote:Eyu100: I wouldn't recommend a central/dividing expanse of grassland- You'd often end up with WW1-type scenarios, where a side would have to charge into a well defended position and take heavy losses, just to try to make any sort of progress in the game. You need to have a carefully thought out distribution of defensive terrain in areas that divide combat zones, which allow sides ample opportunity to retreat/advance without retreating/advancing too far. In some cases, where the neutral ground is larger, castle hexes can help move things along- providing good cover without slowing land units as much as heavier terrain. Sulla's Ruins comes to mind- Take a look at that one.
I will not tell you my corner / where threads don't get locked because of mostly no reason /
because I don't want your hostile disease / to spread all over the world.
I prefer that corner to remain hidden /
without your noses.
-Nosebane, Sorcerer Supreme
User avatar
Doc Paterson
Drake Cartographer
Posts: 1973
Joined: February 21st, 2005, 9:37 pm
Location: Kazakh
Contact:

Post by Doc Paterson »

Many users seemed a bit surprised that the vampire bat, after having its village defense cut to 40, was still getting the full 60 on a majority of village types. I realize that this point is rather basic, but I think that it bears repeating, for the benefit of newer players and/or more experienced designers who may have briefly overlooked it. ;)

Villages with dual terrain status often "break" the precedent of other dual-terrain types, in that some deviate from the standard "best defense, worst movement" quality. For example, Cavalry, mermen and the like are able to move on mountain villages (an instance of "best movement, best defense"). The two types of swamp villages have interesting qualities as well. One is "swamp-movement, swamp-defense," while the other is "best movement, best defense," among the swamp and village types.) (The first is the newer of the two types and looks similar to the original water village.)

Below, I will list the dual-terrain types. This isn't particularly difficult information to gather (by playing the game itself, looking at the files, etc.), but I'm guessing that such a list will serve as a quick and useful reference for newer users.

For those unfamiliar with the terminology, "worst movement, best defense," and the like refer to a unit's movement penalty and defense over a given terrain. The desert village, for example, which is a dual terrain of the sand and village types, gives a unit "best movement, best defense." A spearman moving into a desert village hex would therefore receive no movement penalty (spearman movement into a standard village, triggered by the "best movement" quality, which overrides the -2 penalty for sand) and would receive 60 defense (spearman's defense in a standard village, overriding the 30 sand defense).

Some of the dual terrain listed may also serve as reminders of the many fine-tuning components available to MP map designers. For example, the sand-hill terrain is an almost "surgical" piece, very narrow in its scope of factional interaction modification- Imagine, for example, a map in which a higher than standard hill count was required, or employed for the sake of map distinctiveness (resulting, either way, in balanced matches for all combinations of non-Northerner games). The sand-hill could, proportional to its use/integration, power down Northerners by inhibiting their movement, with zero influence over other factional interactions. There are other terrain-"tools" with similarly narrow fields of influence, but, to greater and lesser extents, the ripple effect of factional influence must be considered/estimated/adjusted for appropriately.


Dual-Terrain

Bridge-over-shallow-water, bridge-over-swamp, and bridge-over-deep-water: (grassland and shallow water, grassland and swamp, grassland and deepwater) Best movement, best defense.

Swamp-ruin and sunken-ruin: (castle and swamp, castle and shallow water) Best movement, best defense.

River Ford: (grassland and shallow water) Best movement, best defense.

Snow Forest" (snow and forest) Worst movement, best defense.

Snow Hill: (snow and hill) Worst movement, best defense.

Desert Hill: (sand and hill) Worst movement, best defense.

Cave Hill: (cave and hill) Worst movement, best defense.

Desert Village: (sand and village) Best movement, best defense.

Snow Village: (snow and village) Best movement, best defense.

Hill Village: (hill and village) Best movement, best defense.

Mountain Village: (mountain and village) Best movement, best defense.

Cave Village: (cave and village) Best movement, best defense.

Water Village: (shallow water, with village attributes) Shallow water movement, shallow water defense.

Swamp Village (thatched roof): Best movement, best defense.

Swamp Village (partially submerged): (swamp, with village attributes) swamp movement, swamp defense.


* * * * *

Special Note: The Cave/Mushroom Grove is often thought of as a cave/mushroom dual terrain, but it is actually a straight-forward alias of the mushroom grove. Even the "illuminated-cave mushroom grove" is not considered to be part cave.

The "rubble" terrain is likewise often thought of as a sand/hill alias, when it is, in fact, simply a hill alias.

The "Bridge-over-lava" terrain is not a grassland or cave/lava dual. It is only a cave alias, and does not illuminate.

The "Snow Hill Village" is not a triple terrain, and is merely a Hill-Village alias. In other words, don't plan on your Yeti getting increased defense there. ;)
I will not tell you my corner / where threads don't get locked because of mostly no reason /
because I don't want your hostile disease / to spread all over the world.
I prefer that corner to remain hidden /
without your noses.
-Nosebane, Sorcerer Supreme
bert1
Posts: 240
Joined: December 6th, 2006, 10:39 pm
Location: Morecambe, UK

Re: Guide for designing MP maps.

Post by bert1 »

wintermute wrote:One very simple tip is to ask yourself "what would I recruit on this map?", and if the answer is "pretty much one type of unit", then you might want to make some changes. After all, how much fun would it be to play every game on the map only recruiting merman? This also holds for things like "pretty much just elves" or "pretty much just dwarves".
Good is simply that which is willed. - Eugene Halliday
Becephalus
Inactive Developer
Posts: 521
Joined: October 27th, 2005, 5:30 am
Location: St. Paul, Minnesota, USA, Earth

Re: Guide for designing MP maps.

Post by Becephalus »

Nice guide Bert1 I like it a lot. I also love the video and would love to hear more.
There are three roads to ruin: by gambling, which is the quickest; through women, which is the most pleasurable; and through taking the advice of experts, which is the most certain. -de Gaulle
bert1
Posts: 240
Joined: December 6th, 2006, 10:39 pm
Location: Morecambe, UK

Re: Guide for designing MP maps.

Post by bert1 »

Thanks Becephalus.

Here is an insight or two from Doc in this thread. I bolded the generalisable bit.
Doc Paterson wrote:I gave a quick overview of some of the issues that a 1v1 based on a 2v2 concept will face in my PM to you, but here are a few more specific things:

-A P1 gryphon or bat can steal P2's lower water village.

-P2's northern swamp village can be stolen by any P1 faction.

Both of these issues are amplified by the fact that gold is twice the normal amount, and replacement units will certainly be "in the area" right away to relieve (replace/rotate in for) a wounded village thief. The map is not so large that that unit density is not going to be very significantly higher than in any other 1v1.

-Even if P1 could not get guaranteed steals on any of P2s village, the advantage is still going to be very large for several reasons. Primarily: P2 can not use their leader very well to take the heat off of any of those highly pressurable north and south villages. They're just so far away, and not recruiting for that long in a 1v1 makes the journey kind of pointless. In a 2v2, your opponent could at least pick up the slack a little by continuing to recruit.

The map as it is creates another sort of dilemma. The terrain density is kind of low, which will give Loys a noticeable advantage in a lot of matchups. If you increase it more though, it'd be very easy to inflate the Drake advantage over most every faction, and, as I said in the PM, the Drake issue is already there because of the nature of the front (For other readers though I'll reiterate: 1v1s are often chopped into compartments for a reason, much of which is the limitation of the mobility of certain units. In a 2v2, mobility advantages are very often "watered down" by the fact that your team tends to have more answers to certain situations, being able to draw from a bigger pool of recruit options {and of course mirror matches are accordingly much more likely, leveling the playing field all the more}. In a 1v1, a wide unrestricted front like this greatly amplifies the Drakes' ability to crush a weak village area. Sometimes that extra turn that a defender gets from Drakes having to move around an obstacle is enough to give them a significantly better regional defense.....Imagine Freelands, Den of Onis, Caves of the Basilisk, Silverhead Crossing etc. etc. without impassible regions.) That's the challenge that this map faces, having a long, vertical, mostly undifferentiated front (an example of a long central vertical differentiated front would be Fallenstar Lake.)

Sulla's Ruins kind of walks that line, so you may want to study that one. A vaguely similar approach might do some good in this case; carefully adding some castle hexes to certain areas, letting other factions, particularly Undead, advance, pursue and retreat more quickly and effectively. As you do this, I'd recommend increasing the forest and hill count very slightly. This being in response to the Loy-vs.-non-Drake factions issue.

There's probably much more to say here, but I'll leave it at this for now, and we can discuss more if and after you make some adjustments.

:)
Good is simply that which is willed. - Eugene Halliday
User avatar
Doc Paterson
Drake Cartographer
Posts: 1973
Joined: February 21st, 2005, 9:37 pm
Location: Kazakh
Contact:

Re: Guide for designing MP maps.

Post by Doc Paterson »

An excellent addendum written by Velensk:
Velensk wrote:
This is a rough draft for a guide on map making . The purpose of this guide is to extend the guide to designing multiplayer maps with more advanced concepts and after this is beyond the draft phase I hope to have it transplanted into the stickied guide. Comments/feedback/suggestions are welcome.

Many people enjoy making maps but most new players underestimate how difficult it is to create a balanced and interesting map. This guide is intended to be a few remarks on some of the harder/more abstract ideas behind map design.

Strategic terrain design:

This section is about how the map as a whole is shaped. This defines in general how armies will move, what areas are easier to defend, which areas are susceptible to attack, how easy it is to form a ZoC wall across a front, how easy it is to feint ect. It also describes the placement of villages as targets, the number of villages (as a general indication of the number of forces), and the locations of keeps and castles.

When thinking about this perspective don’t worry very much about specific terrain. Think more about shapes and how they affect movement for different factions/units. Once you have a shape in your mind think about how it would cause units to move and which units would be able to change the typical flow (with qualities such as flying, swimming, dwarven terrain walk, ect.). Also look at the way the terrain will naturally draw up defensive lines and how narrow various channels are/how hard it is to scale the channel (by channel I mean an area of easy to transverse ground between two hard to transverse/impassable areas.

Each map should have the units moving in different ways and have different considerations to take into effect for strategic variety.

A couple general notes about design:

Force Flow: Can be abstracted to a line that goes the fastest possible route between recruiting castles and the various fronts. On some maps (like hornshark) you also need to consider the number of forces that will be/can be recruited an sent each way. On hornshark chances are each player will probably recruit 2-4 units to send directly up or down to gather/hold villages then their leader and a flying/water unit will head the other direction to grab villages/move to the other keep and from that point on most forces will come from the second keep.

Fronts: Generally in maps combat will tend to take place along certain lines that will move with the flow of time of day and as leaders move to alternate keeps. The areas where they tend to form are by pieces of cover, villages, and (if close to combat) recruiting keeps. Determining where fronts will be will be helpful for determining where you should pay the most attention to tactical terrain design (covered later). The fronts will (or should) move around a bit. Access between alternate fronts is something that should be monitored.

Terrain position relative to force flow and fronts: Terrain is important in defining where the force flows and fronts form but after this is determined, the position of the terrain along the fronts determines much of how the map will play.

(for the purpose of this paragraph, most usage of the word terrain refer to defensive terrain that slows people down)Clumped terrain tends to be a massing point for a force who is preparing to attack but does not have the strength to stand in the open yet or for a place to use as a corner for a formation. Clumps can also be hid behind/retreated over or can serve as the location for the final stand of a trapped unit. Terrain that strings out to forms lines perpendicular the the force flow tend to be points easy to defend leading to more defensive games. Terrain strung out parallel to the force flow tends to lead to battles where both players have access to terrain but the majority of units are in the open. A caution about the parallel terrain is that it can cause imbalance if one side can move over these obstacles much faster than the other for the reason of feinting covered next. Terrain where the corners are exposed tend to be easier to attack than terrain where the corners are covered in some way.

Feint potential: Consider a situation where X is several units for an attacker, Y is several units for a defender, # is terrain that is impassable (or slows sufficiently) and the 0s are villages that are by default the defenders.

...Y...
...#...
.0#0.
...#...
...#...
...X...

The defender units can either go to defend one village or the other or they can split their forces between the two villages and try to hold them the full might of the attacker (or of course they can simply surrender both but in that case it doesn't matter). If defender splits his forces and the attacker commits all his forces to one village it will be a couple turns before the defenders other forces can get to the fight. Now In a similar situation.

...Y...
...#...
.0#0.
...#...
...#...
X......

X can still reach either village in two turns but most players (even some of the more experienced ones) will still rush all their forces to the village on the left village. The attacker can then move all his forces around the corner and have a large number of units threatening the village (so that a single defender can’t hold it) and the defender could not rush enough forces there to prevent this before the attacker could commit. The speed of the attacker determines how far away he can do this kind of thing from.

How much of this you want on your maps is up to you however I would argue that at least some of it is good for most maps but too much of it will kill any map.

Village Exposure: This is simply how easy each village is to attack or defend. Generally villages that are easy to attack will not be in a narrow channel and will not have defendable terrain around them (though a single piece of terrain directly in-front of them can offer cover for attackers under certain conditions). In my opinion it is a good idea to have at least one village that is fairly exposed to help avoid stalemate situations. When calculating village exposure using the distance of the village from the opponents keep is also a consideration as it determines the rate at which reinforcements can arrive (and if it’s close enough the leader can defend it).

Distance between villages: In general the greater the distance between the villages controlled by the two players the harder it is to attack. There are two reasons for this, the first is that enemy villages are generally a primary target and the further away from your safety zone you have to go before your enemy cannot just simply retreat to avoid loses and wait for a better time of day. The second consideration is that villages are healing stations and being further away from the ones you control means that you will be able to reinforce less quickly. This is especially the case when fighting an opponent who uses poison. The later problem is can be avoided in the fashion of Quetzacoatl’s map Astral Port where the map is speckled with wml implemented ‘healing runes’ that act like villages for the purpose of healing but do not generate income.

Multiple keeps: These are a good way of expanding the players reach on maps that are too big for a single keep to recruit from to cover defensively or agressively, they are also actually a way of slowing down the expansion of the player while keeping the leader close to all the critical villages. Some maps use many of them in clusters to encourage players to have their leader out on the battlefield and constantly skirmishing around it. Use with prudence as they can cause other flow problems.

Number of villages: Should be proportional to the number of units you want on the map and probably the number of players as well.

Examples of analysis of the strategic terrain design of a few mainline maps.

In the freelands there are three main flanks. From the center you can reach either flank though the terrain will probably slow you down considerably. The center flank is also the hardest to attack in because it naturally forms narrow straits and has a concave curve of terrain and the village in the center is also within easy reach of the leader. There are narrow channels on either side of the center flank where troops can transfer to the opponents side of one flank freely, but a single enemy unit can defend this point standing on a hill and barring flying/deepwater units it can only be attacked from two hexes. Because of this, general flows of force go around the center in a loop going around the center area through the non-recruiting castles and the villages closest to the enemy. Fronts tend to form in the corners of the center, near the castles on the side, and by the village closest to the enemy.
There is an instance of feint potential very much like my example on either side however it is balanced by the fact that it is usually fairly easy for the defender to assemble a quick defense in the center leaving him available to focus on the sides.
On the sides there are a few clumps of terrain that generally act as staging grounds for attacks.
The player on the north side has a rich field of villages headed by a fairly exposed village on the left and his opponent mirrors this trend on the right. The village that heads the field is far enough away from the north players castle that he cannot reinforce with most infantry in a single turn. However after a certain point the north player can form a solid wall of units completely ZoC walling the channel with the village off in a very solid defense. Once this forward village is taken it is generally easy to retreat and if the enemy cannot regroup there are many other villages that can be claimed.
On the opposite side the north player has two villages close to a lump of castle terrain which enemies can stage attacks from or which he can hop to when staging attacks of his own. One of these villages is fairly exposed but is deep enough into north player territory that the south player risks getting cut off if he attacks it and the other is close to the leader/reinforcements and can be defended to give minimum exposure.

The terrain along the fronts tends to be parallel to the flow of forces with a few clumps off to the side. With 8 villages per side, a map its size tends to fill up at a decent speed.

Hornshark Island

Both players start at a keep too far away from some of their villages to effectively protect them but their leaders can reach this village by turn 3 for p1 and turn 2 for p2. The map has two distinct sides with a somewhat neutral area with no villages in-between (other than a water village with highly limited exposure). The first castle for each player is on the opposite side from the second and the access to the villages on one side from the other is poor for both players. There is a tendency for each player to recruit what it needs to grab the villages on the side of the first keep at the first keep and then after moving their leader to not be able to reinforce it easily. As a result one side starts stronger than second but the second side soon gains more strength leading to a kind of imbalance in forces that hopefully spurs action. After the leaders move they actually have an easier time accessing some of the enemies villages (by enemies village I mean villages the enemy can reach first) than they do their own. There is practically no feint potential because enemy villages tend to be too close together. The main fronts tend to form between the the leaders on one side and the close enemy villages on the same side and in the neutral area as both players try to reinforce their weak side.
Most of the village exposure is cause more by distance from keep than by the positioning of terrain around them. Terrain tends to take the form of clumps. As the lines of force are less easily defined than on some maps it is harder to define it in the terms I did in my example however there are a few lines of both types. At eight villages per side a map its size would fill up fairly quickly if it weren't for the fact that it tends to be very aggressive and many units tend to die early.
The outside edge of the map is deep water making it possible for flying or swimming units to get around the edge to steal villages or preform back attacks. However these require several turns to pull off and can be defeated by water/flying units of the defender.

Ruins of Terra-Dwelve

Large four player map that is 2vs2 (1221) only. Because it is large each player starts with 25 extra gold so that they can deploy the number of units required to contain a bat spammer. It uses a 3 keep set-up to slow down the village grabbing while still allowing leaders to be fairly close to the front and cover more area. Each player has 3 general lanes of attack including one that is very close to their ally and his/her enemy. Inbetween the three fronts are areas that are generally hard to transverse. In between the center (of the map) path and the middle (of the players front) path is an area with water/cave/mushrooms and other things that slow however as it is clumped it is easy to go around either way. Inbetween the middle (of the players front) and the outside (of the map) path is an area with lots of defensive terrain arrayed in a somewhat clump-like and somewhat string-like manner. This area is much harder to go around but can be gotten through without too much trouble but it will slow down most factions. In the outermost region of the outermost area is a small path that is very hard for anyone non-flying to use that can be blocked easily by one units.
There is one village which is incredibly close to the enemies matching village that is exposed to four hexes of attack. To get at all four hexes one must control both the center path (of the map) and the middle path (of the players flank).

Other villages could be exposed due to the players leader being out of place or to far away or quite defendable if enough units are in the area. There is a clump of terrain in the center (of the map) path for staging attacks from or fighting over. At 9 villages per player the map gains units at a high rate but it is large enough that it takes awhile to reach saturation.

There is loads and loads of feint potential due to the way the various terrains are positioned to slow players down in moving between fronts. The map tends to give players 1 and 2 the initiative however hopefully they cancel each other out.

Teamwork in the center can be critical to victory or the center could be almost completely unused.

Tactical Terrain Placement

In the last section I told you to not worry to much about specific terrain and more about its general effect (open, slows most units, provides defense the most units, slows fliers, water ect.) Obviously what terrain you use and the more specific shape you put them in are very important to both how the map actually plays and how enjoyable it is.

When designing a clump of terrain always be aware of what its purpose is from your analysis of strategic terrain design. Be aware of what directions forces will typically be approaching it from and things of that nature. Now for most clumps this will be as simple as ‘provide a defendable point in this region’ however it might be more complex. The other major concern that should be considered is factional balance. In most cases it is fine to have clumps of terrain more friendly to some factions than others however, you should not have areas that simply cannot be attacked by one faction if defended by another and you should try to balance the overall feel of map so that no faction has a net advantage over others.

Generally in a clump of terrain the most important terrains are the terrain on the corners and at the narrow points. The critical elements to determining what hexes require the most consideration for what type of terrain they hold are that ability to deny access by controlling this hex, the number of hexes that an enemy can be expected to attack this hex from, and if this hex provides cover to attack another important hex from. Consider if in the freelands the hills close to the narrow side channels were forests instead. It would suddenly play quite differently because now dwarves and trolls could not hold those positions so easily and deny access through the channel and elves would become much better at it.

On the different types of terrains and their uses.

Open: The majority of the map should be this. You will notice that even in games between good players most of the fighting is done in the open (though close to terrain usually). Do not simply place clumps of terrain in any. In terms of factional balance, loyalists tend to do better than most factions in the open due to the resilience of their cavalry in the open.

Forest/Hills: These fill roughly the same role. They are general ‘rough terrain’ that most factions can defend decently but not well and that slow down most factions. Hills tend to be slightly more ‘generally friendly’ than forests however excesses of hills will give an advantage to knalgans and to a lesser extent northerners and drakes. Forests are of course the favored terrain of rebels.

Mountains: The upgraded version of the hill for most purposes. Slows units down even more and gives even higher defense. The major notes about this is that dwarves are not slowed and enjoy ridicules defense in addition to their generic toughness. As a result many mountains in critical areas or even a single mountain next to a village is generally a bad idea. Even in a non-critical areas mountains can provide an excellent place for fliers to retreat to that others can’t follow. Any area blocked off entirely by mountains is completely inaccessible to any loyalist cavalry.

Castles: Castles are useful because they provide high defense without slowing anybody down. Castles are a good way of making undead stronger as most undead can defend castles as well as most other people but have very little problem attacking castles due to the frequency of dark adepts.

Mushrooms: This peculiar terrain fills a critical role in slowing down flying units. It is the only terrain in existence that slows bats. It is also a good refuge for relatively slow undead. It slows down most types of units and it provides medium defense to most so it has use other than

Caves: Slows down most fast units considerably. The exceptions are both the undead scouts and saurians (and to a lessor extent wolf riders and footpads). The exceptions make it so that you cannot reliably use it as a way to slow down units for the initial village grabbing however you can still use it to make slow areas on the front and restrict general access. An excess of caves will seriously harm rebels and loyalists.

Snow: Another good terrain for slowing everyone. The problem is that unlike caves it doesn't slow down one of the units you occasionally really need to slow down (griffons). Snow is a good way to deal with drake mobility and lock down saurian mobility almost entirely. Loyalists also have allot of trouble with snow. The other factions generally suffer to a lesser extent.

Sand: This is a good terrain for slowing everyone except drakes. Drakes will not only not be slowed but will get max defense in sand. As drake mobility tends to be a problem anyway be very careful with it. Knalgans suffer less than most other factions too.

Normal Village vs Mixed Village: In general the two things you need to keep in mind here are bats and water units. On any village other than a standard one a bat is a nuisance to kill and having a large number of them around can make bats even more annoying. On the other hand, mixed villages may be more aesthetically appealing or it may be that you want village stealing to be a serious threat. Villages that water units can hold and attack can add an incentive to use water units or merely make for an interesting choice for your land units (do I leave that water village open to be taken or do I step into 20% defense to hold it). Keep in mind what units can hold what mixed villages better than standard villages when using them. For this reason I generally suggest not using mountain villages as it will make it a pain and a half to get that dwarf off that mountain without using magic.

Water: Water is generally defined in the strategic shape of the map and should be considered there. However there are a few notes about designing ‘clumps’ of water in the tactical sense.

Deep Water vs Shallow Water: Generally it’s best to use mostly shallow water. Water units are at an inherent disadvantage to flying units in that they generally have much more limited ability to threaten villages than flying units do. In deep water flying units get the same defense as water units thus hurting the practice of the ‘flying units may be able to fight both in land and water, but water units can fight in water better’ idea. However, deep water does have a few uses. First of all it does add a low-defense/high-defense terrain dynamic to sea vs sea battles. Second of all, by no amount of trudging can land units (except skeletons) get through deep water so you can use it in rivers as an effectively impassable region. The final one is related the the second: Use of deep water provides potential to use the submerge special. It will probably never be a huge thing even if you give the opportunity but it can be very cool when it happens.

Swamps are a good piece of water to use all by themselves in strings or masses of land terrain. They are still generally bad but there are a couple units that can make use of it making it a little more interesting. I would generally advice not using large clumps of swamp to form rivers or lakes unless you want areas that saurians will be able to navigate incredibly easily and other land units cannot.

Fords can allow you to have areas in your open ground where water units can join the fray or pass over open ground quicker.

The coastal reef terrain acts like the hills/forest of the water for sea vs sea battles. It is also easier for land units to cross so it can be used like a slightly more realistic ford (i.e. it slows both land and sea units slightly but both can cross it without too much difficulty).
I will not tell you my corner / where threads don't get locked because of mostly no reason /
because I don't want your hostile disease / to spread all over the world.
I prefer that corner to remain hidden /
without your noses.
-Nosebane, Sorcerer Supreme
User avatar
Doc Paterson
Drake Cartographer
Posts: 1973
Joined: February 21st, 2005, 9:37 pm
Location: Kazakh
Contact:

Re: Guide for designing MP maps.

Post by Doc Paterson »

I did some general cleanup here and deleted several posts that weren't really related to this thread.

I might as well link to this as well: http://forums.wesnoth.org/viewtopic.php ... 78#p438778

as it contains some of my thoughts on map design that I always seem to end up repeating. :)
I will not tell you my corner / where threads don't get locked because of mostly no reason /
because I don't want your hostile disease / to spread all over the world.
I prefer that corner to remain hidden /
without your noses.
-Nosebane, Sorcerer Supreme
Post Reply