Developing an Old School Sandbox for 5E – Part 3

I’ve posted a couple of times about developing a sandbox setting for D&D 5E (Part 1, Part 2), and this week I’m going to expand on the climates, terrain, and monster selection.


As my sandbox is an island, and it’s not so large as to be a full continent, there is a limit to the different climates that I can realistically include. I’ve decided that this island sits in the northern waters of the world, and so climate reflects this.

The island itself is slightly larger than the Northwest Territories in Canada—the surface area is approximately 916,249 square miles (1,474,560 square kilometers). So this gives me some room to work with.

Using the Köppen climate classification types to describe the island, the southern and middle portions of the island are subarctic climate, the northern portion of the island is polar tundra, and the higher elevations are dry-summer subarctic.



This has a direct influence on both the terrain types I will use and the monsters I plan to include.

Terrain Types

As mentioned previously, the 5E DMG provides lists of the monsters divided by terrain type. The terrain types listed in the DMG are Arctic, Coastal, Desert, Forest, Grassland, Hills, Mountains, Swamp, Underdark, Underwater, and Urban.

So applying these terrain types to the island based on the climate I’ve chosen, I get the following:

  • Arctic—The mountains in the northeast of the island use the arctic terrain type. This is due to their elevation in addition to their latitude, increasing the sub-arctic climate to arctic as you climb higher into the range.
  • Coastal—As this is an island, the coastal terrain type is definitely applicable.
  • Desert—Even though the mountains provide a rain shadow for the interior of the island, I don’t want it to be too dry. A sandy desert is out, and I don’t feel the island is far enough north to get a dry snow desert. So I’m not going to use this terrain type on the island.
  • Forest—Most of the middle and southern portions of the island are subarctic and therefore forests are very appropriate. The forests are almost exclusively conifers (needles instead of broad leaves) which remain green throughout the cold months. It’s not unknown for the occasional broadleaf forest to be found within a subarctic zone, and so I’ll probably include one in the southern area of the island.
  • Grassland—The northern tundra can be considered a grassland for the purposes of monster selection by terrain type, though the vegetation is very short and is composed mostly of shrubs, mosses, and lichens. The central area of the island is also covered by a grassland.
  • Hills—Each of the three sets of mountain peaks are surrounded by foothills. In addition, one of the sets of hills extends out into the central part of the island (near the grassland noted above).
  • Mountains—As mentioned, there are three distinct sets of mountain peaks. The mountains in the northeast are fairly low and very cold. The mountains in the southeast are essentially a continuation of that same chain, though the ground between them is low enough that they seem as if they are a separate set of peaks. The mountains on the west side are much larger (cover more area) and have a higher elevation.
  • Swamp—The tundra in the north transforms into swampland during the short summers when the temperatures rise enough to thaw the ground frost. The ice melts and creates many bogs and marshes (as well as lakes and streams).
  • Underdark—This is less a “terrain” type than it is a location that can underlie almost any of the other terrains. On this island, the underdark will be mostly found underneath the hills and mountains. As I plan to have this campaign be mostly about exploration of the island itself, I’m not going to make the underdark too extensive.
  • Underwater—Like the underdark, I don’t want to run an extensive underwater campaign. Therefore, I plan to have a large lake with underwater ruins that can be explored if the PCs are interested, but it won’t be a major part of the campaign.
  • Urban—The point of this campaign is a wilderness hexcrawl focused on exploration. So I’m placing a small town that is the PCs starting point, though I don’t plan to put any adventure hooks that lead to purely urban adventures there. I also have ideas for two other small settlements on the island, but they won’t be sizable urban environments. The one possible exception to this is that there are the ruins of a small city on the island that is entirely abandoned by people after some kind of disaster, and only monsters can be found there (as well as some interesting mysteries and cool set pieces).


So I have my climate, and this affects the terrain types to include on the island. And now I have to select my monsters.

  • Humanoids—The first choice I need to make is about how many humanoid races I want to include. D&D contains many different options here, such as the goblin races, orcs, drow, bullywugs, derro, duergar, firenewts, gnolls, grimlocks, grungs, kenku, kobolds, kuo-toa, lizardfolk, merfolk, sahuagin, tabaxi, troglodytes, and yuan-ti.

    Obviously, including all of these would be far too much. Some I can eliminate simply by climate and terrain type (such as yuan-ti), and others just don’t really fit into the setting (grungs).

    Still, that leaves me with many options.

    For now, I expect that I’ll include some form of goblinoid race (probably straight goblins and perhaps bugbears, but likely not hobgoblins). I may also include orcs as a tribal race that inhabits the tundra in the north. As far as the bits that take place in the underdark, I will likely include duergar and one other—most likely either grimlocks or troglodytes, whichever I can make the most interesting.

Of the other monster types, these will be selected on a case-by-case basis:

  • Aberrations—I will definitely include a few aberrations with each as the core monster for a larger encounter area. My plan is to create a few new aberrations to provide something new for the PCs to discover.
  • Beasts—Natural animals will certainly populate most of the wilderness areas, and I will also include some of the giant versions and a few of the larger beasts. Dinosaurs will not be found on the island.
  • Celestials—As celestials are native to the Upper Planes, and generally are of the same (or similar) alignments to the PCs, I don’t have plans to include these creatures (unless as a one-off for a particular encounter area).
  • Constructs—I will certainly include a few constructs on the island, mostly as remnants created by those who lived in the ruined city on the island.
  • Dragons—I do have plans to include at least one dragon, as I have a new race of creatures related to dragons that will play a part in the setting.
  • Elementals—These creatures will appear as appropriate to specific encounter locations only.
  • Fey—Some types of fey will certainly inhabit some of the wild places on the island, though they certainly won’t be common.
  • Fiends—Like elementals, these creatures will appear as appropriate to specific encounter locations only. I do have a couple of ideas already, so there will definitely be a few included.
  • Giants—I do plan for there to be a couple of types of giants on the island. I do not intend to use the Ordning or anything similar to constrain the giants into a hierarchy.
  • Monstrosities—I will certainly include some monstrosities in my list of monsters on the island. They will most often be part of specific encounter locations, but some can be found in the random tables.
  • Oozes—These will be included as appropriate to the climate and terrain type.
  • Plants—I do plan for there to be some plant creatures on the island, and I intend to create a few new ones for PCs to discover.
  • Undead—There will certainly be undead on the island, though they will not be a focus of the campaign.

Island Regions

Now I’m in the process of creating specific regions on the island. A region can be as small as one hex, or as large as I need it to be. A particular forest will usually be a single region, and a region could include an entire mountain range or just a single mountain, depending on its relationship to the surrounding terrain.

From the moment the PCs leave the main town, they will move from one region to another as they explore the island. Each region will usually have a noticeable boundary (such as the edge of a forest into a grassland, or crossing a river into a new area), though some may have large transition areas as regions overlap for some miles.

In some cases, the regions may be defined by the monsters themselves. For example, if I choose to include a colony of ettercaps, they may take over part of a larger forest. While the forest itself could be a single region, it would generally make more sense for the spider-infested area to be a single region, with the regular portion of the forest a neighboring region.

For this reason, the development of regions and the placing of monsters basically goes hand-in-hand.

Random Encounters and Set Pieces

And, of course, once the regions are developed, each one will get its own set of random encounter tables, reflecting the creatures that could be found in that particular region.

Each region will also have at one set piece encounter, and probably a few. These are locations that do not change and are not random. For example, a goblin lair where a particular goblin tribe lives would be a set piece encounter, with a map of the lair and description of the tribe and its members.

Not all set piece encounters will necessarily include monsters, of course. When exploring a hex, there will be interesting things to find that won’t always lead to a fight, or even interaction with living (or undead) creatures.

But this the final, and longest, step in developing the sandbox and will take some time to do.


I’ve been picking away at this setting here and there as I work on other projects, so it’s not moving terribly quickly. I hope, though, that my thoughts here provide some insight into the development of such a sandbox setting.

I’ve already started planning out the regions and marking them on the island map, and I’ve created a couple of the hexes in the first region. The next time I update this project here I will include some of the developed regions and a couple of completed hexes so you can see how I will present the information for use when running the game.

Developing an Old School Sandbox for 5E – Part 1

I’ve recently talked about using the D&D 5E rules for old school play (part 1, part 2, part 3, part 4, part 5). As I mentioned last week, I’ve been really interested in putting together an old school sandbox (also known as a hexcrawl) campaign.

I’ve been a DM since I bought the original Tom Moldvay red box Basic Dungeons & Dragons boxed set back in 1981. And a big part of being a DM back in the day was designing your own setting that did what you wanted it to do.

So I’m going to develop my own lands for the PCs to explore (probably for some future campaign), and I figured it might be interesting to others, so I will blog about it here. And, who knows, it might eventually end up as something publishable, or at least downloadable here.

The Foundation

When looking at this project, I think it’s important to outline some of my core objectives in doing this. This will help to keep me on track and ensure that I don’t waste time focusing on things that might temporarily grab my interest, but won’t actually contribute anything useful to the project.

I’ve thought about this a bit, and here are my initial goals for this setting:

  1. Develop a setting for use in actual D&D play.
  2. Focus on multiple regions, each with a distinct feel and look.
  3. Each region should have a reason to explore it, aside from specific “adventure locations” that the characters may find. For example, there might be mining opportunities in certain regions, or characters might be hired to explore certain areas because the empire is considering expansion.
  4. Provide opportunities to place individual adventure locations scattered across the locations. These will be unique geographic features, dungeons, settlements, monster lairs, localized magical effects, and anything else I can come up with that will provide direct character interaction beyond just exploring.
  5. Ensure the regions have a “frontier” feel to them—the empire has never before settled or even explored in this direction in any official capacity.
  6. Keep the home base as a safe location for the characters. Ensure adventure is “out there” rather than inside the actual home settlement.

I expect that, as I go forward, I will likely add another objective or two, but I think this is a good start. It will help me develop the foundation of the setting at the very least and ensure that it is strong and consistent.

The Regions

The core elements of the campaign setting will be the regions to explore. As I mentioned, each region should have a distinct feel and look. It should have a “character” all its own, and the players should be able to tell when their player characters leave one region and enter another.

One of the easiest ways to do this is, of course, with the specific environment. The 5E Dungeon Master’s Guide provides a list of environments in which the monsters of the game are mostly commonly encountered. These environments are:

  • Arctic
  • Coastal
  • Desert
  • Forest
  • Grassland
  • Hill
  • Mountain
  • Swamp
  • Underdark
  • Underwater
  • Urban

At this early point in development of the campaign setting, I’m not entirely sure I’m going to include all of those environments. I want my setting to be realistic enough that the players can understand the world, and so it’s difficult to have all of those environments close enough together in a way that makes sense. Further, some environments, like forests, may show up multiple times in different directions.

The Scale

Which brings me to the scale of the setting. I want the players to have the opportunity to travel and explore, and this means that there needs to be a fairly large area in which to place each region. Plus, each region needs to have enough space to house multiple adventure locations without adventurers stumbling upon one location after another during each day of travel.

So the key is to determine how fast the party will be able to travel in a single day while exploring, and then determine how much travel it will take to reach the next region.

According to the 5E Player’s Handbook, the normal travel pace is 24 miles in a day (assuming a movement speed of 30 feet per round). However, different characters are going to have different movement rates (e.g. a wood elf has a speed of 45 feet, but a dwarf is only 25 feet), I plan to incorporate the variant encumbrance rules on page 176 of the PHB, and armor could also modify a character’s speed.

While some class abilities provide even faster movement as they go up in level, like a monk’s increased speed and even the gaining of a fly speed at 20th level, I’m going to use the base speed to determine the size of the regions in this setting. That way, as characters gain levels, they will find it easier to explore farther away from their home base.

So here are the possible movement rates and the per day travel maximums.

Movement Rate
Per Round Per Day
35 feet 27 miles
30 feet 24 miles
25 feet 21 miles
20 feet 18 miles
15 feet 15 miles
10 feet 12 miles
5 feet 9 miles

Unless the characters are hauling a ton of equipment (or treasure!), however, most the time the party—assuming they travel at the speed of the slowest member—will have a daily speed of either 21 miles (if the party has a dwarf or halfling) or 24 miles (if no dwarf or halfling is present).

It’s also important to keep in mind, though, that these travel times “assume relatively simple terrain: roads, open plains, or clear dungeon corridors.” Difficult terrain halves movement speed, and includes “dense forests, deep swamps, rubble-filled ruins, steep mountains, and ice-covered ground.” So a lot of the exploring that the characters will undertake will be done at a much slower pace.

Note: I’m based in Canada, and many of the visitors to my blog also live outside the United States (or Burma or Liberia). So why am I using such an arbitrary and outdated measurement system? In this case, it is purely because these are the distances used in the core rulebooks for D&D. It would be great if they switched over to the International System of Units (i.e. the metric system) like the rest of the world, but as of the current printing of the D&D books, they have not. So, in order to save myself a ton of work converting everything, I’m going to use imperial units in these posts.

Now, for mapping wilderness areas, a common recommendation is to use a 6-mile hex. There’s a great post on The Hydra’s Grotto from back in 2009 about the use the 6-mile hex, and I fully agree with his arguments. This means that a party will likely be able to cross 4 hexes—or 3.5 hexes if they have dwarves or halflings in the party—in a single day.

Obviously, some regions are going to be larger than others, and the size may vary quite a bit. I’m going to start with a rule of thumb that the largest regions will take about a week to cross. The smallest regions should be crossable in about a day. By planning this out by time, rather than distance, it will give me an idea of how large the region should be.

For example, a dense forest that takes a week to cross isn’t actually (24 miles x 7 days = ) 168 miles in width. Since dense forest means half-speed travel, it’s actually only (12 miles x 7 days = ) 84 miles in width. Still a great deal of area to explore, but it won’t need to be so large on the map due to the speed restrictions.


So that’s it for this week. I have the basic idea down for the campaign wilderness, and my objectives are set. Next time (not necessarily next week), I’m going to talk a bit about the various monsters that I might want to include in the setting.

D&D 5E – Old School and Encounter Balance

Recently, I’ve been talking about using D&D 5E for an “old school” campaign style.

Sandbox Gaming

One of the reasons I’ve been thinking so much about old school play lately is that I really want to run a full sandbox game.

If you’re not familiar with the term, I’m going to quote Ben Robbins from his ars lundi blog, who inspired a lot of gamers with his posts about his West Marches Campaign.

“There was no regular plot: The players decided where to go and what to do. It was a sandbox game in the sense that’s now used to describe video games like Grand Theft Auto, minus the missions. There was no mysterious old man sending them on quests. No overarching plot, just an overarching environment.”


“The game was set in a frontier region on the edge of civilization (the eponymous West Marches). There’s a convenient fortified town that marked the farthest outpost of civilization and law, but beyond that is sketchy wilderness. All the PCs are would-be adventurers based in this town…The whole territory is (by necessity) very detailed. The landscape is broken up into a variety of regions (Frog Marshes, Cradle Wood, Pike Hollow, etc.) each with its own particular tone, ecology and hazards. There are dungeons, ruins, and caves all over the place, some big and many small. Some are known landmarks (everbody knows where the Sunken Fort is), some are rumored but their exact location is unknown (the Hall of Kings is said to be somewhere in Cradle Wood) and others are completely unknown and only discovered by exploring (search the spider-infested woods and you find the Spider Mound nest).”

But most important is the following:

“PCs get to explore anywhere they want, the only rule being that going back east is off-limits — there are no adventures in the civilized lands, just peaceful retirement.”

Now, none of this absolutely forbids the use of pre-published adventure modules in the campaign. But the key element is that the players decide where their characters are going to go, what rumors inspire them to investigate further, what job offers they want to accept, and which direction they want to explore.

So if the DM has an adventure that he feels is interesting and would like to run, then he or she drops it into an appropriate area on the campaign map and feeds the characters the hooks (rumors, job offer, requests for aid, etc.). They players may or may not show any interest. If they decide to go in another direction, then the DM doesn’t move the adventure over to where the PCs are headed, because that’s not how the real world works.

What this means is that the DM needs—at the very least—a campaign map with major terrain features on it, and a bunch of potential adventure locations (which might be settlements, ruins, dungeons, monster lairs, or interesting geographical features). The DM then sprinkles a few adventure possibilities in the town where they start, and the game begins.

I have to admit, I’ve never run a fully sandbox game before (often referred to as a hexcrawl for the fact that most large overland maps tend to use hex grids instead of square grids). And I’ve been thinking about it a lot lately and I’m really interested in starting a brand new campaign at first level and exploring the possibilities in such a campaign.

Now, early editions of D&D were certainly well-suited to such a campaign. But I’ve been exploring how the current edition of D&D, fifth edition, would work when combined with old school play. And so the idea of combining the general precepts of old school play with a pure sandbox campaign has grabbed me and won’t let go.

Encounter Balance

Since one of the core elements of a sandbox/hexcrawl is that the world exists independent of the player characters, it means that not everything will be balanced against the current level of the characters at the time they encounter it.

Ben Robbins had this to say about the West Marches campaign:

“The environment is dangerous. Very dangerous…PCs have to work together or they are going to get creamed. They also have to think and pick their battles — since they can go anywhere, there is nothing stopping them from strolling into areas that will wipe them out. If they just strap on their swords and charge everything they see they are going to be rolling up new characters. Players learn to observe their environment and adapt — when they find owlbear tracks in the woods they give the area a wide berth (at least until they gain a few levels). When they stumble into the lair of a terrifying hydra they retreat and round up a huge posse to hunt it down. The PCs are weak but central: they are small fish in a dangerous world that they have to explore with caution, but because they are the only adventurers they never play second fiddle. Overshadowed by looming peaks and foreboding forests yes. Overshadowed by other characters, no.”

In Ben’s campaign, he set up his map into specific areas (a particular set of woods, a mountain, a swamp, an old battlefield, etc.) and then gave each area a particular “level” of challenge. Areas nearest the character’s home base were generally of low level, and the challenge increased as the characters proceeded farther away from “civilization.” This maintained a certain level of verisimilitude in the game, as the players were on the edge of a settled empire, so areas closest to it would be the least dangerous.

It also had a bit of a balancing effect, as the players’ first-level characters would likely explore closest to their home base, and then branch out farther as they gained in level.

However, Ben also put individual locations within those areas that might be more dangerous than the surrounding area would indicate. A crypt with a bunch of wights inside could be in the middle of the goblin-infested forest. As long as the wights were trapped inside the sealed crypt, the surrounding forest was only as dangerous as the goblins. But the characters could find the crypt and decide to break into it…

One of the keys to making this work is that the DM provides some clues to the players that they are entering a more dangerous area. The tracks of an owlbear are mentioned above. The entrance to the wight-filled crypt might have (cryptic) warnings on the outside. If the players are paying attention, they should have an idea that they are about to enter a more dangerous area, and then they are making an informed decision if they choose to do so. Or perhaps they aren’t paying attention and rush their way into an area without scouting it out or otherwise trying to find out some information about it. And then, when they encounter monsters that are too tough for them to beat, they can learn how important it is to run away when you’re outmatched.

What this all amounts to is this: the campaign world map is designed as if it is a real place with real adventure locations on it. Some of those locations will be somewhat dangerous, and some will be downright deadly. It will be up to the players to determine how much of a challenge they want to face, based on the information they can determine about the world through exploring, scouting, and otherwise seeking knowledge. But the campaign world never throws “level-appropriate” challenges at the characters. The decision as to what encounters to face is ultimately on the players’ side, not the DM’s side.

5E and Challenge Ratings

So how does the use of the D&D 5E rules support or prevent such an approach to a campaign. Well, 5E has a couple of tools in the Dungeon Master’s Guide and Monster Manual that helps DMs judge the lethality of a particular monster or set of monsters.

In the Monster Manual, all monsters have a Challenge Rating. The MM defines the Challenge thusly:

“A monster’s challenge rating tells you how great a threat the monster is. An appropriately equipped and well-rested party of four adventurers should be able to defeat a monster that has a challenge rating equal to its level without suffering any deaths.”

This is a nice tool. It tells me that I probably shouldn’t place the lair of an ancient red dragon in the region next to the character’s home base, as that is going to be far too difficult a challenge for low-level characters (and the in-world explanation is that the empire from which the characters hail would probably have already sent in soldiers to deal with a powerful red dragon right on their doorstep).

The second tool is in the DMG and shows the DM how to combine multiple monsters into encounters, and figure out how dangerous those encounters will be to the party of characters.

Again, this is a great tool that provides good information to the DM.

Now, at this point I have to point out that some people look at the information under “Creating a Combat Encounter” in the DMG and think that they are somehow required to always abide by the charts and advice in this section. But that is silly—no one from Wizards of the Coast cares in the slightest whether or not you use the information in the DMG to develop the encounters in your campaign. It’s there for two reasons:

  1. Advice for those who have never played the game before and are looking to balance the challenge to the party of player characters.
  2. Information for DMs so that they fully understand how challenging a particular of monsters will be to a party of a certain level.

But just because that information is there, doesn’t make it a rule that a DM is required to slavishly obey.

Personally, I prefer to have more information than less. If I’m creating a sandbox campaign, I want to have a rough idea of how challenging the various encounters I build and drop onto the map are going to be in actual play. This way, I can ensure that I’m not throwing truly deadly challenges into an area when I’ve intended the area to only be mildly challenging…and vice versa.

These tools are just like the pile of actual tools I have in my toolbox at home. Just because I own about fifty different tools doesn’t mean I have to use them all whenever I need to do some work on my home. Sometimes, I just need a screwdriver. Sometimes I just need a hammer. And sometimes (on a big project), I need most of the tools in the toolbox and I’m glad that I already have them.

The Challenge Ratings in the MM and the tables in the DMG are like that. I can use them when I want them, and they stay in storage when I don’t need them.

The Sandbox Campaign

So how does this all tie together?

For this sandbox campaign, my plan is to develop a campaign map, just like Ben Robbins did for the West Marches Campaign. I will have a series of regions on the map, most likely divided by appropriate terrain features, and will populate them with potential adventure locations. Those regions nearest to the characters’ home base will tend to be less dangerous (with individual locations perhaps being far more dangerous than the surrounding area might indicate), and regions farther away will be more dangerous.

Further, I will develop rumors, legends, and other clues that observant players will be able to pick up on in order to gain a better understanding of the different regions and what kinds of things they might find as they go exploring in a particular direction.

I will use the tools in the MM and the DMG to make sure I understand how challenging particular areas are. But it will still be up to the PCs to decide if (and when) they approach those areas and face those challenges.

And then the campaign will be ready to start.


While the phrase “balanced encounters” gives many gamers—especially old school gamers—the willies, it is important to keep in mind that the tools in the 5E MM and DMG are there to provide information to the DM, not to put a straightjacket on them. A sandbox/hexcrawl campaign is just as viable in 5E as it is in previous editions of the game.