Spells and spellcasters in a campaign world
Recently, I was looking over some old threads, and came across a brief debate I participated in. I'd postulated that there'd be a number of "off-color" and even largely useless (from a combat perspective) spells in a campaign world, simply due to human nature, and how ubiquitous magic is. Another person disagreed, saying that just because something could exist doesn't mean it must. While I agree with that principle, it seems silly to suggest that certain spells won't exist in a campaign world when magic itself operates like a science, and spells are a commodity. To that end, I'm going to crunch some numbers here to try and estimate the number of different spells that exist in a campaign world that holds to the standard 3.5 d20 rules. In several places some things are assumed, but I think the conclusions that can be drawn here are fairly logical, and offer a good guideline for such a campaign.
Before I begin, I want to mention that this is largely my own take on a similar essay found in Distant Horizon Games's superb book, The Practical Enchanter. I got some different numbers than they did (largely because it doesn't seem as though they took the community modifiers into account), and wanted to show my results. I HIGHLY encourage people to go download the book via the link above, as it's not only one of (in my opinion) the best d20 books out there, it's also free to download!
So how does one determine the number of unique spells in a campaign world? Well, the most obvious place to start is with the people casting them. The section on NPCs in communities in the DMG (pg. 138-139), lets us determine the highest levels of NPCs, and in turn calculate their numbers. To establish a baseline, let’s say that the majority of the world's population lives in hamlets (population ranging from 81-400). Just so there’s a fixed number to work from, let's say that the average hamlet has a base population of 200 people.
Now, the next step is to determine the classes and number of people living in the hamlet. Though the DMG already has a sample breakdown of NPC in a hamlet of two hundred people, let’s run the numbers anyway to see what our results are. Using the demographics rules and tables in the DMG (pg. 138-139), lets us start determining everyone who has PC class levels. For the sake of expediency, assume that every die roll on the Highest Level Local Table is average (what happens with the .5 aspect to the average numbers is discussed below), and take the community modifier (-2 in this case) to generate the results. For the PC classes, these are the results:
Barbarian: None.
Bard: One 1st-level bard.
Cleric: One 2nd-level cleric, and two 1st-level clerics.
Druid: One 1st-level druid.
Fighter: One 3rd-level, and two 1st-level fighters.
Monk: One 1st-level monk.
Paladin: None.
Ranger: None.
Rogue: One 2nd-level rogue, and two 1st-level rogues.
Sorcerer: None.
Wizard: One 1st-level wizard.
While the method of determining these levels is explained in the DMG, the results generated here require some explanation. All of the average dice results result in a number that ends with a .5 aspect to it. This slightly skews the data, because it means half the time the number will be 1 greater than it will be the other half of the time, and this does affect the numbers generated for this community. As such, what I’ve done here is take two classes that roll similar dice, and assign the higher part of the average to one class, while the other gets the lower average. For example, to find the highest levels of both barbarians and monks, roll 1d4 (average 2.5) and subtract 2 (the community modifier). The result is 0.5 for each. Hence, I assign one class (monks in this case) to have a result of 1 (thus resulting in the hamlet having a single level 1 monk), and the other class (barbarians) to have a result of 0, meaning that there are no barbarians in this community. The classes that I “paired off” this way to equalize the average dice rolls were: barbarians and monks, clerics and druids, fighters and rogues, paladins and rangers, and sorcerers and wizards. In the case of the bard, the remaining .5 for its average was ported over to the adept.
NPCs with NPC class levels in a community are generated the same way that NPCs with PC class levels are, save for determining 1st-level characters. Thus, determine the higher-level characters first:
Adepts: One 2nd-level adept.
Aristocrat: N/A (generated using rules for level 1 NPCs with NPC class levels).
Commoner: One 8th-level commoner, two 4th-level commoners, and four 2nd-level commoners.
Expert: One 6th-level expert, two 3rd-level experts.
Warrior: One 3rd-level warrior.
Again, in several places the averages were moved between different classes. The adept, as mentioned previously, received the remaining .5 average from the bard, raising its 1.5 result to a 2. The aristocrat, which had a result of 0.5, gave its average to the expert, lowering its result for this stage of NPC generation to 0. Finally, the commoner and warrior had no 0.5 averages to move, since both had average results that resulted in whole numbers.
Finally, take the remaining number of individuals in the community, and populate them with the percentages given in the DMG. Since our results have generated a grand total of twenty-five individuals thus far, that leaves one hundred seventy-five left to generate. Based on the percentage figures given in the DMG (91% commoners, 5% warriors, 3% experts, 0.5% aristocrats, and 0.5% adepts, all 1st-level), we get (with some rounding) the following figures:
Adept: One 1st-level adept.
Aristocrat: One 1st-level aristocrat.
Commoner: One hundred fifty-nine 1st-level commoners.
Expert: Five 1st-level experts.
Warrior: Nine 1st-level warriors.
Thus, we now have the entire population of an average-size hamlet. This is significant because, as mentioned above, we’re assuming that – since the average campaign world is roughly similar to medieval Europe – this represents the population breakdown of the majority of the world. More specifically, (again using medieval Europe as a rough guideline) we’ll assume a world population of about 70,000,000 people, of which 80 to 90% (we’ll use a baseline of 85%) live in hamlets, and thus are defined by the numbers we generated above. This makes it significant that only 4% of the hamlet can cast spells – eight people out of two hundred – and 75% of those are divine spellcasters. (To be clear, the “world population” is, in this context, limited to creatures of the Humanoid type. If you want, however, you can limit it to the races, including their sub-races, from the PHB. E.g. dwarves, elves, gnomes, halflings, and humans of all types.)
(Just for fun, let’s see what other assumptions we can make about this hamlet based on the data we’ve generated. For example, the 2nd-level cleric is probably the local spiritual counselor, who together with the two 1st-level clerics who most likely serve as his aides, maintains a small church. It’s not too much of a stretch to suggest that this religion has a monastic order as well, which would make the 1st-level monk part of the church order as well. This forms the pillar of the religious part of the community. The 2nd-level adept might be an old wise woman, perhaps thought of as a witch, at the edge of town, together with her 1st-level apprentice. Even further afield, the 1st-level druid might keep watch over the surrounding wilderness, while the 1st-level wizard is conducting his self-taught magical research in relative isolation. The ten warriors in town would probably be the local militia, keeping bandits and goblins away from the community, while the contingent of fighters – having a different fighting style from their fighter bonus feats – would probably be a small group of archers, or perhaps cavalry. And administrating over the town is the mayor, the 1st-level aristocrat.)
What about the other 15% of the world population, then? Well, we can chart their statistics also. However, even using an abbreviated listing for the remaining types of communities is very long and very dry. As such, we’ll just use the results gathered from generating averages for those community levels.
Refining the world population numbers even further than the above becomes very pedantic, but does help us generate a more accurate look at the population breakdown of the average fantasy world. I said before that 85% of the world’s population lives in hamlets, and ran my figures based on the demographic breakdown of the population of the average hamlet. In order to get more accurate totals regarding the remaining 15%, just repeat the initial community demographic figures with each of the other types of communities available. Since it’s best to be conservative, let’s stagger the population towards smaller communities, and assume that the remaining global percentages breakdown like so: 3% live in thorps (population 20-80, average 50 people), 3% live in villages (pop. 401-900, avg. 600 people), 3% live in small towns (pop. 901-2,000, avg. 1,500 people), 2% live in large towns (pop. 2,001-5,000, avg. 3,000 people), 2% live in small cities (pop. 5,001-12,000, avg. 8,000 people), 1% live in large cities (pop. 12,001-25,000, avg. 18,000 people), and 1% live in metropolises (pop. 25,001 or more, avg. 40,000 people).
Since it’d be a bit too tedious to list all the numbers and processes here, I’ll just post the results, using the same calculations that were done to find the demographic breakdown of the population of an average hamlet:
Before I begin, I want to mention that this is largely my own take on a similar essay found in Distant Horizon Games's superb book, The Practical Enchanter. I got some different numbers than they did (largely because it doesn't seem as though they took the community modifiers into account), and wanted to show my results. I HIGHLY encourage people to go download the book via the link above, as it's not only one of (in my opinion) the best d20 books out there, it's also free to download!
So how does one determine the number of unique spells in a campaign world? Well, the most obvious place to start is with the people casting them. The section on NPCs in communities in the DMG (pg. 138-139), lets us determine the highest levels of NPCs, and in turn calculate their numbers. To establish a baseline, let’s say that the majority of the world's population lives in hamlets (population ranging from 81-400). Just so there’s a fixed number to work from, let's say that the average hamlet has a base population of 200 people.
Now, the next step is to determine the classes and number of people living in the hamlet. Though the DMG already has a sample breakdown of NPC in a hamlet of two hundred people, let’s run the numbers anyway to see what our results are. Using the demographics rules and tables in the DMG (pg. 138-139), lets us start determining everyone who has PC class levels. For the sake of expediency, assume that every die roll on the Highest Level Local Table is average (what happens with the .5 aspect to the average numbers is discussed below), and take the community modifier (-2 in this case) to generate the results. For the PC classes, these are the results:
Barbarian: None.
Bard: One 1st-level bard.
Cleric: One 2nd-level cleric, and two 1st-level clerics.
Druid: One 1st-level druid.
Fighter: One 3rd-level, and two 1st-level fighters.
Monk: One 1st-level monk.
Paladin: None.
Ranger: None.
Rogue: One 2nd-level rogue, and two 1st-level rogues.
Sorcerer: None.
Wizard: One 1st-level wizard.
While the method of determining these levels is explained in the DMG, the results generated here require some explanation. All of the average dice results result in a number that ends with a .5 aspect to it. This slightly skews the data, because it means half the time the number will be 1 greater than it will be the other half of the time, and this does affect the numbers generated for this community. As such, what I’ve done here is take two classes that roll similar dice, and assign the higher part of the average to one class, while the other gets the lower average. For example, to find the highest levels of both barbarians and monks, roll 1d4 (average 2.5) and subtract 2 (the community modifier). The result is 0.5 for each. Hence, I assign one class (monks in this case) to have a result of 1 (thus resulting in the hamlet having a single level 1 monk), and the other class (barbarians) to have a result of 0, meaning that there are no barbarians in this community. The classes that I “paired off” this way to equalize the average dice rolls were: barbarians and monks, clerics and druids, fighters and rogues, paladins and rangers, and sorcerers and wizards. In the case of the bard, the remaining .5 for its average was ported over to the adept.
NPCs with NPC class levels in a community are generated the same way that NPCs with PC class levels are, save for determining 1st-level characters. Thus, determine the higher-level characters first:
Adepts: One 2nd-level adept.
Aristocrat: N/A (generated using rules for level 1 NPCs with NPC class levels).
Commoner: One 8th-level commoner, two 4th-level commoners, and four 2nd-level commoners.
Expert: One 6th-level expert, two 3rd-level experts.
Warrior: One 3rd-level warrior.
Again, in several places the averages were moved between different classes. The adept, as mentioned previously, received the remaining .5 average from the bard, raising its 1.5 result to a 2. The aristocrat, which had a result of 0.5, gave its average to the expert, lowering its result for this stage of NPC generation to 0. Finally, the commoner and warrior had no 0.5 averages to move, since both had average results that resulted in whole numbers.
Finally, take the remaining number of individuals in the community, and populate them with the percentages given in the DMG. Since our results have generated a grand total of twenty-five individuals thus far, that leaves one hundred seventy-five left to generate. Based on the percentage figures given in the DMG (91% commoners, 5% warriors, 3% experts, 0.5% aristocrats, and 0.5% adepts, all 1st-level), we get (with some rounding) the following figures:
Adept: One 1st-level adept.
Aristocrat: One 1st-level aristocrat.
Commoner: One hundred fifty-nine 1st-level commoners.
Expert: Five 1st-level experts.
Warrior: Nine 1st-level warriors.
Thus, we now have the entire population of an average-size hamlet. This is significant because, as mentioned above, we’re assuming that – since the average campaign world is roughly similar to medieval Europe – this represents the population breakdown of the majority of the world. More specifically, (again using medieval Europe as a rough guideline) we’ll assume a world population of about 70,000,000 people, of which 80 to 90% (we’ll use a baseline of 85%) live in hamlets, and thus are defined by the numbers we generated above. This makes it significant that only 4% of the hamlet can cast spells – eight people out of two hundred – and 75% of those are divine spellcasters. (To be clear, the “world population” is, in this context, limited to creatures of the Humanoid type. If you want, however, you can limit it to the races, including their sub-races, from the PHB. E.g. dwarves, elves, gnomes, halflings, and humans of all types.)
(Just for fun, let’s see what other assumptions we can make about this hamlet based on the data we’ve generated. For example, the 2nd-level cleric is probably the local spiritual counselor, who together with the two 1st-level clerics who most likely serve as his aides, maintains a small church. It’s not too much of a stretch to suggest that this religion has a monastic order as well, which would make the 1st-level monk part of the church order as well. This forms the pillar of the religious part of the community. The 2nd-level adept might be an old wise woman, perhaps thought of as a witch, at the edge of town, together with her 1st-level apprentice. Even further afield, the 1st-level druid might keep watch over the surrounding wilderness, while the 1st-level wizard is conducting his self-taught magical research in relative isolation. The ten warriors in town would probably be the local militia, keeping bandits and goblins away from the community, while the contingent of fighters – having a different fighting style from their fighter bonus feats – would probably be a small group of archers, or perhaps cavalry. And administrating over the town is the mayor, the 1st-level aristocrat.)
What about the other 15% of the world population, then? Well, we can chart their statistics also. However, even using an abbreviated listing for the remaining types of communities is very long and very dry. As such, we’ll just use the results gathered from generating averages for those community levels.
Refining the world population numbers even further than the above becomes very pedantic, but does help us generate a more accurate look at the population breakdown of the average fantasy world. I said before that 85% of the world’s population lives in hamlets, and ran my figures based on the demographic breakdown of the population of the average hamlet. In order to get more accurate totals regarding the remaining 15%, just repeat the initial community demographic figures with each of the other types of communities available. Since it’s best to be conservative, let’s stagger the population towards smaller communities, and assume that the remaining global percentages breakdown like so: 3% live in thorps (population 20-80, average 50 people), 3% live in villages (pop. 401-900, avg. 600 people), 3% live in small towns (pop. 901-2,000, avg. 1,500 people), 2% live in large towns (pop. 2,001-5,000, avg. 3,000 people), 2% live in small cities (pop. 5,001-12,000, avg. 8,000 people), 1% live in large cities (pop. 12,001-25,000, avg. 18,000 people), and 1% live in metropolises (pop. 25,001 or more, avg. 40,000 people).
Since it’d be a bit too tedious to list all the numbers and processes here, I’ll just post the results, using the same calculations that were done to find the demographic breakdown of the population of an average hamlet:
Tags: demographics, population
Total Comments 5
Comments
-
Part 2
A thorp of fifty people has just one 1st-level cleric and one 1st-level adept for their spellcasting population. This keeps with what I got for hamlets in that 4% of the community population are spellcasters, save that here this is entirely divine spellcasters (perhaps further reinforcing a bias against arcane spellcasters). The remaining population is one 2nd-level and two 1st-level fighters; one 1st-level rogue; one 7th-level, two 3rd-level, and thirty-four 1st-level commoners; one 4th-level, two 2nd-level, and one 1st-level experts (the last one being perhaps an apprentice); and one 2nd-level and two 1st-level warriors.
A village of six hundred people has a spellcasting populace of one 3rd-level and two 1st-level clerics; one 2nd-level and two 1st-level druids; one 1st-level sorcerer; one 2nd-level and two 1st-level wizards; one 2nd-level and two 1st-level bards; and one 3rd-level and two 1st-level adepts. This gives us a total of sixteen people out of six hundred who can cast spells, or 2.67%. Interestingly, this is a one-third drop in percentage from thorps and hamlets. However, the strength of the magic available is actually stronger than before; this is the first time that 2nd-level spells are available, albeit divine ones only (from the 3rd-level cleric). The remaining population is one 4th-level, two 2nd-level, and four 1st-level fighters; one 3rd-level and two 1st-level rogues; one 2nd-level and two 1st-level monks; one 1st-level barbarian; one 1st-level ranger; three 1st-level aristocrats; one 7th-level, two 3rd-level, and seventeen 1st-level experts; one 4th-level, two 2nd-level, and twenty-eight 1st-level warriors; and one 9th-level, two 4th-level, four 2nd-level, and five hundred-eight 1st-level commoners.
A small town of fifteen hundred people has a spellcasting populace of one 4th-level, two 2nd-level, and four 1st-level clerics; one 3rd-level and two 1st-level druids; one 2nd-level and two 1st-level sorcerers; one 3rd-level and two 1st-level wizards; one 4th-level, two 2nd-level, and four 1st-level bards; and one 3rd-level and seven 1st-level adepts. This total of thirty-one spellcasters is almost double what could be found in a village, but again the overall percentage is less, this time being only about 2.07%. However, the other trend continues as well, in that for the first time, 2nd-level arcane spells are available (via the 3rd-level wizard and 4th-level bard). The remaining population is as follows: one 5th-level, two 2nd-level, and four 1st-level fighters; one 4th-level, two 2nd-level, and four 1st-level rogues; one 3rd-level and two 1st-level monks; one 2nd-level and two 1st-level barbarians; one 2nd-level and two 1st-level rangers; one 1st-level paladin; one 2nd-level and seven 1st-level aristocrats; one 8th-level, two 4th-level, four 2nd-level, and forty-three 1st-level experts; one 5th-level, two 2nd-level, and seventy-two 1st-level warriors; and one 10th-level, two 5th-level, four 2nd-level, and one thousand three hundred-five 1st-level commoners.
In a large town of three thousand people, the spellcasting population again shrinks somewhat in terms of overall percentage, despite the fact that some paladins and rangers are now strong enough to be counted among them. The spellcasting populace is as follows: one 7th-level, two 3rd-level, and four 1st-level clerics; one 6th-level, two 3rd-level, and four 1st-level druids; one 6th-level, two 3rd-level, and four 1st-level wizards; one 5th-level, two 2nd-level, and four 1st-level sorcerers; one 4th-level paladin; one 5th-level ranger; One 7th-level, two 3rd-level, and four 1st-level bards; and one 6th-level, two 3rd-level, and fourteen 1st-level adepts. At only fifty-four people, this brings the spellcasting percentage of the populace to a mere 1.80%. The remaining population is one 8th-level, two 4th-level, four 2nd-level, and eight 1st-level fighters; one 7th-level, two 3rd-level, and four 1st-level rogues; two 2nd-level and four 1st-level paladins; two 2nd-level and four 1st-level rangers; one 5th-level, two 2nd-level, and four 1st-level barbarians; one 6th-level, two 3rd-level, and four 1st-level monks; one 5th-level, two 2nd-level, and fourteen 1st-level aristocrats; one 11th-level, two 5th-level, four 2nd-level, and eighty-seven 1st-level experts; one 8th-level, two 4th-level, four 2nd-level, and one hundred forty-five 1st-level warriors; and one 13th-level, two 6th-level, four 3rd-level, and two thousand six hundred twenty-eight 1st-level commoners.
In a small city of eight thousand people, we start to have a difference that skews the data slightly. Here, we roll twice to determine the highest level of NPCs, with each such roll generating more lower-level NPCs. Hence, the spellcasting population increases slightly than it would otherwise, but still falls as a percentage of the total population. A small city gives us one two 10th-level, four 5th-level, eight 2nd-level, and sixteen 1st-level clerics; two 9th-level, four 4th-level, eight 2nd-level, and sixteen 1st-level druids; two 9th-level, four 4th-level, eight 2nd-level, and sixteen 1st-level wizards; two 8th-level, four 4th-level, eight 2nd-level, and sixteen 1st-level sorcerers; two 8th-level and four 4th-level paladins; One 7th-level ranger; Two 10th-level, four 5th-level, eight 2nd-level, and sixteen 1st-level bards; and two 9th-level, four 4th-level, eight 2nd-level, and thirty-eight 1st-level adepts. This spellcasting population of two hundred-nine people is 2.60%, almost equal to the results for the village. The rest of the population breakdown is as follows: Two 11th-level, four 5th-level, eight 2nd-level, and sixteen 1st-level fighters; Two 10th-level, four 5th-level, eight 2nd-level, and sixteen 1st-level rogues; Eight 2nd-level and sixteen 1st-level paladins; Four 3rd-level and eight 1st-level rangers; Two 8th-level, four 4th-level, eight 2nd-level, and sixteen 1st-level barbarians; Two 9th-level, four 4th-level, eight 2nd-level, and sixteen 1st-level monks; Two 8th-level, four 4th-level, eight 2nd-level, and thirty-eight 1st-level aristocrats; Two 14th-level, four 7th-level, eight 3rd-level, and two hundred twenty-eight 1st-level experts; Two 11th-level, four 5th-level, eight 2nd-level, and three hundred eighty 1st-level warriors; and two 16th-level, four 8th-level, eight 4th-level, sixteen 2nd-level, and six thousand nine hundred sixteen 1st-level commoners.
A large city has an average population of eighteen thousand. Here, you roll three times, tripling the usual number of high-level NPCs, which continues to increase the spellcasting population. With three 13th-level, six 6th-level, twelve 3rd-level, and twenty-four 1st-level clerics; Three 12th-level, six 6th-level, twelve 3rd-level, and twenty-four 1st-level druids; Three 12th-level, six 6th-level, twelve 3rd-level, and twenty-four 1st-level wizards; Three 11th-level, six 5th-level, twelve 2nd-level, and twenty-four 1st-level sorcerers; Three 11th-level and six 5th-level paladins; Three 10th-level and six 5th-level rangers; Three 13th-level, six 6th-level, twelve 3rd-level, and twenty-four 1st-level bards; and three 12th-level, six 6th-level, twelve 3rd-level, and eighty-seven 1st-level adepts, there is a total of three hundred fifty-one spellcasters in the large city, for 1.95% of the local population. The remainder of the population breaks down as follows: Three 14th-level, six 7th-level, twelve 3rd-level, and twenty-four 1st-level fighters; Three 13th-level, six 6th-level, twelve 3rd-level, and twenty-four 1st-level rogues; Twelve 2nd-level and twenty-four 1st-level paladins; Twelve 2nd-level and twenty-four 1st-level rangers; Three 11th-level, six 5th-level, twelve 2nd-level, and twenty-four 1st-level barbarians; Three 12th-level, six 6th-level, twelve 3rd-level, and twenty-four 1st-level monks; Three 11th-level, six 5th-level, twelve 2nd-level, and eighty-seven 1st-level aristocrats; Three 17th-level, six 8th-level, twelve 4th-level, twenty-four 2nd-level, and five hundred twenty 1st-level experts; Three 14th-level, six 7th-level, twelve 3rd-level, and eight hundred sixty-eight 1st-level warriors; and three 19th-level, six 9th-level, twelve 4th-level, twenty-four 2nd-level, and fifteen thousand seven hundred ninety 1st-level commoners.Posted 28th August 2008 at 03:15 AM by Alzrius
Updated 4th October 2008 at 05:09 AM by Alzrius -
Part 3
Last is the largest type of community, a metropolis, with an average of forty thousand people living in it. The spellcasters living in an average metropolis is as follows: four 16th-level, eight 8th-level, sixteen 4th-level, thirty-two 2nd-level, and sixty-four 1st-level clerics; Four 15th-level, eight 7th-level, sixteen 3rd-level, and thirty-two 1st-level druids; Four 15th-level, eight 7th-level, sixteen 3rd-level, and thirty-two 1st-level wizards; Four 14th-level, eight 7th-level, sixteen 3rd-level, and thirty-two 1st-level sorcerers; Four 14th-level and eight 7th-level paladins; Four 13th-level and eight 6th-level rangers; Four 16th-level, eight 8th-level, sixteen 4th-level, thirty-two 2nd-level, and sixty-four 1st-level bards; and four 15th-level, eight 7th-level, sixteen 3rd-level, and one hundred ninety-five 1st-level adepts. Out of the total population, six hundred seventy-five can cast spells, or 1.69%. The remaining population is as follows: Four 17th-level, eight 8th-level, sixteen 4th-level, thirty-two 2nd-level, and sixty-four 1st-level fighters; Four 16th-level, eight 8th-level, sixteen 4th-level, thirty-two 2nd-level, and sixty-four 1st-level rogues; Sixteen 3rd-level and thirty-two 1st-level paladins; Sixteen 3rd-level and thirty-two 1st-level rangers; Four 14th-level, eight 7th-level, sixteen 3rd-level, and thirty-two 1st-level barbarians; Four 15th-level, eight 7th-level, sixteen 3rd-level, and thirty-two 1st-level monks; Four 14th-level, eight 7th-level, sixteen 3rd-level, and one hundred ninety-five 1st-level aristocrats; Four 20th-level, eight 10th-level, sixteen 5th-level, thirty-two 2nd-level, and one thousand one hundred sixty-seven 1st-level experts; Four 17th-level, eight 8th-level, sixteen 4th-level, thirty-two 2nd-level, and one thousand nine hundred forty-four 1st-level warriors; and four 20th-level, eight 10th-level, sixteen 5th-level, thirty-two 2nd-level, and thirty-five thousand three hundred ninety-three 1st-level commoners.
It’s worth pointing out that these numbers are, from a very strict standpoint, somewhat unreliable. Even overlooking the averaging that was done, there’s also some ambiguity over rounding up or down when generating NPC levels. For example, if the highest-level Fighter in a hamlet is 3rd-level, are there also two 2nd-level Fighters and four 1st-level Fighters (rounding up), or just the 3rd-level Fighter and two 1st-level Fighters (rounding down)? Unfortunately, the DMG gives examples that use both methods. To try and maintain uniformity with the d20 System, NPC levels were always rounded down, rather than up. It’s also worth noting that it was always assumed that all rangers and paladins of 4th or 5th level have high enough Wisdom scores to be able to cast spells. I’m also overlooking the 5% chance that a thorp or hamlet has to add +10 to the level of their highest-level druid or ranger.
Of course, all of these averages are just that: averages. There are a lot of NPCs who’ll have higher and lower levels than the ones generated here. For example, by these results, even a metropolis will never have an NPC of high-enough level to create 9th-level spells. Of course, this is true for arcane spellcasters anyway. The highest level NPC wizard or sorcerer that can be created on the Highest-Level Locals table in the DMG is level 16! Perhaps all 9th-level arcane spells in the campaign world are special, unique creations by specific NPCs.
So, what does this mean for spells in the game world? Well, by dividing the percentage of spellcasters in each type of community into the percentage of the world population that lives in each type of community, and then adding the results together and dividing them into the total global population, gives us a percentage of how many people in the campaign world are spellcasters. This figure comes out to 3.7866%, but let’s round it to 3.8% to account for the smattering of NPC spellcasters who aren’t merely products of the community generation tables. Thus, we know that 3.8%, or a little less than one in every twenty-five people in the world, has some ability to cast spells. Moreover, the arcane/divine breakdown from the hamlet holds true on the global scale as well – 75% of those who can cast spells are divine spellcasters, with the remaining 25% being arcane. Taking that into account, 96.20% of the game world can’t cast spells at all, 2.85% can cast divine spells, and only 0.95% of the populace can cast arcane spells! No wonder arcane spellcasters have a reputation for being persecuted and misunderstood by the populace!
So, when only 3.8% of the world’s 70,000,000 people is a spellcaster, that gives us a general spellcasting populace of 2,660,000. Now, in order to gauge the number of distinct spells in existence, and being created over time, let’s make some assumptions. First, let’s say that in the span of a given year, only a fairly small number, about one in every thousand spellcasters, invents a new spell. Hence, at any given time, only 2,660 spellcasters will engage in successful spell research in a year (and it’s assumed they only do so once in that year). Bear in mind that since the average human lifespan (according to the age tables in the PHB) is 91 years, that means that only 9% (rounding away the last 0.1%) of human spellcasters ever have even one idea that they’re able to successfully bring to fruition as a new spell over the course of their entire lives – the other 91% never engage in (successful) spell research; only one out of every eleven human spellcasters is innovative enough to create something new. Of course, this percentage goes up among longer-lived races (just over half of all elven spellcasters, for example, create a new spell over the course of their lives, which might explain why elves are typically thought of as the first race to master magic).
Having established that 2,660 new spells are invented each year, let’s keep things conservative by making another large assumption: fully 50% of all new spells that are invented are never passed into any sort of record, dying with their creator. Whether due to negligent records, something happening before they could distribute copies, or simply not having a means or desire of passing their knowledge on, half of all new spells created each year are lost to the world. Hence, only 1,330 new spells ever enter any form of circulation. If that doesn’t sound like too much, remember that this number is annual.
The next limiting provision is to say that this has only been going on for the last two thousand years. All magical research from before that period is completely lost, destroyed, or otherwise unavailable (most worlds seem to have some sort of magical cataclysm that makes this sort of thing entirely believable). Hence, there have been 1,330 new spells being created every year for the last two thousand years, for a grand total of 2,660,000 spells.
In order to cut down on this total, another large assumption is made - that while all of the spells created are distinct (in that they were independently thought of, researched, and created, with no form of collaboration), there is still a lot of overlap; say 99% (“overlap” here being understood to mean that the spells are functionally identical, save for very small differences between them, such as their name, their material components, etc). In other words, different spellcasters all independently had the same idea at various times, and created what is essentially the same spell (this is even more true when you remember that there’s a 50% rate of loss for spells made). Eliminating the overlap means that there are really only 26,600 spells that unique in the entire game world.
For our next assumed figure, let’s say that only one out of every five spells is one that could be of any conceivable use to adventurers. All the rest of these are spells that have no real practical use outside of mundane life. That may sound a bit implausible, considering how many sourcebooks detail spells that deal with attacking, defending, and combat utility, but that is because those are the spells that fall into that one-fifth category. When the majority of the world’s spellcasters live in hamlets, they’re going to invent spells that are practical in regards to themselves and their communities. Hence, the game world will only have 5,320 spells in it that will be of any interest to the PCs. In contrast, the PHB only has a little over six hundred spells in it, meaning that even a character who adventures from level 1 to 20 will probably never come close to seeing all the (adventurer-oriented) spells out there. (Don’t forget that that 75% of the spellcasting population is composed of divine spellcasters; hence, it’d make sense that 75% of the game world’s spells were divine spells, leaving only 1,330 arcane spells to be mastered.)
Now, a lot of assumptions were made in generating these figures. Canny readers will have picked up on a common theme for all of these assumptions (in fact, it was even mentioned a few times): they’re all relatively conservative in the numbers they estimate. Every such guess, from the population (with a low global population; primarily in rural areas) to the spell creation process (only 0.1% invents a spell in a year; only half of them pass their work on; this has only been going on for the last two thousand years; the 99% overlap in spells created; and only one-fifth are worthwhile to adventurers – that is, the PCs) was meant to act as a barrier to spell creation. This was done so that this essay would act as a relatively realistic baseline for determining the number of spells in your campaign world; raise any of these figures, and the number of spells that exist will skyrocket dramatically.
Your d20 campaign world is a magical place; the data proves it.
Posted 28th August 2008 at 03:18 AM by Alzrius
Updated 26th September 2008 at 04:03 AM by Alzrius (fixed grammatical error) -
Wow... I must really read this when I can really focus on this.Posted 22nd September 2008 at 06:15 AM by MichaelSomething -
dude....wow...thanks for the link in response to my post, man. much appreciated.Posted 24th October 2008 at 01:33 AM by joethelawyer - Nice work!
Posted 25th October 2008 at 02:51 AM by Roman
Total Trackbacks 0
And yet another word from our sponsors
