One of the homebrews I am making is a Pokemon homebrew, but the source material uses a 1-100 Level System. As you know, 5e stops leveling at 20. Of course, the obvious solution is to just divide all level-related things from Pokemon by 5 to make the two fit, rounding when necessary, but this creates some issues. Of course, I don't want to make 5e adapt the 1-100 system, either. If I did, I'd be bound to break the Bounded Accuracy of the game. As a DM, house-ruling and minor aspects of home-brewing is pretty basic stuff, but this seems to be more difficult since the choice would have a wider influence than making a pack of Lesser Werewolves or allowing a non-specific Wisdom (Perception) check as an Interaction.

Would my proposed method of making a L100 and 5e's L20 systems fit work optimally or would it cause me serious problems?

Overview of Mechanics

The idea I am toying with is creating a sub-level [Leveling System] exclusively for Pokemon, thus still allowing them to go to Level 100 (following their normal XP/Leveling system from the Pokemon franchise) and having everything else cap at 20 (using 5e's standard XP/Leveling system), while still confining a Pokemon's stat growth to something similar to the 5e system as if their Levels were [Pokemon Level]/5, to maintain balance.

Example: PC Bernard is a Level 3 Paladin. He has a Level 16 Onix. While Onix will display "Level 16" to the other PCs, its stats are the equivalence of a Level 3 PC/NPC within the 5e system. It gains experience following its experience track in the Pokemon games and gains new attacks according to its Pokemon Level, but its proficiencies and gradual stat increases approximately resemble its equivalent 5e level (in this case, "16/5": round down to "3", even if .5 or more). If Onix levels up to Level 20, it will be about the equivalence of a 5e Level 4 PC. (Obvious issue here being wild encounter balancing to make encounters of player party vs. Pokemon an actual challenge.)

I don't know if this would work or if it would break the system in that stats are a little more frequent increases and the their access to a wider array of attacks that the PCs would be able to benefit from early on in a campaign. I've been looking at this too long and need a fresh set of eyes to tell me if this idea is functional or if I need to scrap it.

My intended goals for this change are that it 1) functions, 2) has minimal impact on other aspects of 5e's system, and 3) prevents a Level 5 Pokemon from being a murder machine capable of killing off a Level 1 PC without the system/DM having to enforce an exclusively Pokemon vs. Pokemon battle system that must be strictly adhered to (I want the idea of players vs. Pokemon to be a possible plot that I or others could play around with). These three aspects seem to be mandatory to get a conversion/assimilation to work.

The idea is to make Pokemon be creatures within the world, much like familiars, animals, and monsters. For a "Mystery Dungeon" campaign, where players are Pokemon, I would probably create a different module seeing as there are many stark differences between the two Pokemon systems. This is so you better understand what it is I am trying to do.

  • \$\begingroup\$ Comments are not for extended discussion; this conversation has been moved to chat. \$\endgroup\$
    – mxyzplk
    Commented Jan 18, 2018 at 2:21
  • 1
    \$\begingroup\$ OK, I'm putting this on hold for a moment. I've gotten flags about how we're getting answers that are not tested in any way, as our site rules about homebrew answers require. People with answers below, I recommend you Back Them Up! with how that kind of approach has worked for you. Otherwise, the answers will be deleted and the question reopened after editing to reinforce our Good Subjective, Bad Subjective site rules. \$\endgroup\$
    – mxyzplk
    Commented Jan 19, 2018 at 2:25
  • \$\begingroup\$ @mxyzplk I would like to point out that the question is, essentially, a "yes, it should work without any major issues" or "no, it will not work without major problems" question. The answers have just been... mostly ignoring that. As I said in one comment, there is an implied "What issues would this cause if it won't work without problems." The question itself would basically just boil down to game theory analysis at that point. It, in and of itself, is not subjective. It's just that people are giving subjective or off-topic answers. Is there anything I can change in the OP to reduce that? \$\endgroup\$ Commented Jan 19, 2018 at 2:47
  • 2
    \$\begingroup\$ I believe there's a tendency to answer a question title, rather than the body of the text, which often goes unread when long. Hence most people are answering "how do I model Pokemon in 5e", not "would the homebrew described below cause serious problems". If unclear, my answer was that it would cause serious problems, which I describe in the answer. \$\endgroup\$ Commented Jan 19, 2018 at 3:30
  • 1
    \$\begingroup\$ Regardless, if answers are mostly not answering the question, then usually the question could stand improvement. \$\endgroup\$
    – mxyzplk
    Commented Jan 19, 2018 at 4:16

4 Answers 4


This answer is probably not going to be exactly what you want, or the answer most useful to you, but I think perhaps this perspective could be valuable. I do some freelance third-party work for RPGs (most recently, Pathfinder; I have not done work with, or really even much played, D&D 5e, so grain of salt, etc), and I think you may be falling into a trap that threatens many attempts to play a video game on the tabletop.

I think you may be trying too hard to emulate the game’s mechanics instead of its narrative

This is a super-common mistake made by a whole lot of “conversion” systems—I have one designer friend who has observed that

you can immediately tell a bad Final Fantasy system by whether or not it has a status condition called Frog or Toad.

That’s kind of a glib summation—there’s nothing especially wrong with a Frog condition in and of itself—but it very well might be symptomatic of a real problem (and in my experience, it turns out to be true in practice). The problem is that the designers of the RPG are failing to recognize that tabletop RPGs are very different from video games.

In many ways, this is obvious. Tabletop RPGs are open-ended and player-controlled (including the GM as a player here), allowing for a far more nuanced and thoughtful game. That’s kind of the draw, after all, that’s why you would want to play a video game as a tabletop RPG instead of just playing the game itself. Everyone who is doing this tends to be quite excited about replacing the computer with a real human being, excited about the things that human beings can do far better than a computer can.

But people seem to forget it when it comes time to draft the mechanics of the system. They are excited about the things that a human being is better at doing than a computer, but forget somehow that there are things a computer is better at doing than a human being. Some of these systems are very close to simply taking the mechanics of the video game, and writing down its rules and playing it like an RPG with a d% replacing the computer’s RNG. That does not make for a good tabletop game. That tends to lead to human players expending a whole lot of effort trying to pretend to be computers and do the kinds of things that computers are good at and humans are not (e.g. dealing with lots and lots of numbers that are all interrelated and can grow quite large in magnitude—in fact, I would modify my friend’s comment about Final Fantasy RPGs and focus instead on damage being capped at 9,999 or more as being a pretty good proxy for the system’s overall quality). In one case, I grew so fed up with one of these systems that I literally wrote a computer program (programming is my day job) to handle my character’s stats because it was just unmanageable otherwise.

So you really want to focus your game’s mechanics on doing things that human beings are good at, rather than what computers are good at. That means your mechanics should be influenced far more strongly by existing tabletop RPGs—like D&D 5e—than by the video games you are seeking to emulate.

Now, you are starting from D&D 5e—great! You’re already ahead of some systems I’ve seen. But you also seem to be starting to fall into the trap of getting too invested in the details of the video game’s mechanics—for example, focusing on level 100 versus level 20. That’s a lot like Final Fantasy players focusing on damage values in the hundreds or thousands (or hundreds of thousands), just because those are the numbers that show up in Final Fantasy games. Those numbers aren’t important to the world of Final Fantasy—and level numbers aren’t important to the world of Pokémon.

Instead, focus on how the existing mechanics of D&D 5e can represent the narrative of Pokémon. Effort and training and love strengthens Pokémon, and that can already be represented well in D&D 5e by using XP. Whether a top-tier Pokémon is “level 20” or “level 100” is inconsequential in comparison. You have made the right choice to leverage an existing system here: leverage it to its full advantage. Don’t go ruining it by chasing after details that aren’t really relevant to the play experience of attempting to actually live in the world of Pokémon.

  • \$\begingroup\$ Comments are not for extended discussion; this conversation has been moved to chat. \$\endgroup\$
    – nitsua60
    Commented Jan 19, 2018 at 2:54

Since you're building Pokémon as monsters rather than as players, it's not level you should be looking at — at least, not the D&D game concept called that. In D&D, level is a concept for player characters (and some NPCs). Traditionally, monsters have hit dice as the rough equivalent of player level, but in 5E, it’s really challenge rating you should be looking at. This is a scale used to assess how well a typical party should be able to cope with that monster as a combat challenge, and it exists exactly to deal with your concern about avoiding accidental “murder machines”.

So, I don’t think you should try to modify the 5E level system for PCs to fit your Pokémon creatures. Instead, you should have a parallel system of "Pokémon Levels" (or "P-Levels").

Take a look at the DMG’s “Creating a Monster” section, and particularly the Monster Statistics by Challenge Rating on page 274 (or on D&D Beyond). This gives proficiency bonuses, AC, HP, save DCs for attacks, and attack bonuses and expected damage per round.

Create a mapping of P-Levels to values derived from this chart. The CR goes up to 30, and dividing by 3.3 gives a close enough approximation, although you could fiddle with that. P-Level 41, 42, and 43 could all be CR 12 creatures, with the P-L 41 one having HP closer to 236 (the lower value on the chart) and the P-L 43 having HP closer to 250 (the top value).

You might want to pick a higher divisor to keep top-level Pokémon from rivaling the Tarrasque — but I guess that's a world-building choice. You can decide if you want your top-tier critters to be CR30 — or CR15, or whatever. If you want to have different kinds of Pokémon actually have different strengths at the same level — something very different from the balance implied by 5E's level system — you could assign a different divisor to each kind. (Or, you could do "divisor and offset", with some types starting at CR 0 but others at CR 1 or CR 2 or higher.)

That gives you the basic idea of what the combat power of a Pokémon of a given level should look like in D&D. And then, just fill in everything else around that framework.

The chart doesn't mention ability scores, but those should also increase — you can basically do this however you want, as long as you don’t let the impact of increased modifiers exceed the values appropriate for the target CR. Reverse engineering suggests that the average ability score for a monster of a given CR should be about half of that CR plus 10.5. (This holds true all the way up to Tiamat.)

This actually works out nicely with approximately 3 P-Levels per CR. Since there are six ability scores, three +1 increases (whether to one ability or spread out) makes the overall average go up by 0.5, right in line with the formula.

So, if we start with a system where P-Level 4 is CR 1 (with fractional CRs below that), you can start at P-Level 1 with average ability scores of 10 (maybe a "standard pokémon array" of 13,12,11,10,8,6) and allow any ability score to go up by 1 for every P-Level (perhaps with a caveat that no single score be more than 20 until P-Level 60 or something like that, and with an overall cap of 30). As long as you keep the CR chart stats in line, having a crazy-high Wisdom isn't really going to break anything.

One approach which gives a lot of flexibility and reduces the need to chart out everything advance would be to make increasing any one ability score by 1 be the primary action at each P-Level increase.

  • Increase Proficiency Bonus by CR, and increase HP, but don't raise any of the other chart values. Instead, calculate those directly from ability scores:

    • Compute AC as normal — 10 + Dex to start, possibly factoring in armor
    • Have a base HP per P-Level, and modify by Con modifier
    • Use either Dex or Str for Attack Bonus, adding in proficiency
    • Start with 1d4 for damage dice and a single attack — see below for increasing. (I guess if it's thematic for a particular Pokémon to start with a multiattack, use just "1" for damage rather than a die)
    • For any special attacks that are spell-like or cause conditions instead of damage, set the DC at 8 + proficiency + an appropriate ability modifier
  • If the increase would take AC, Attack Bonus, or Save DC more than 2 points over the value for the target CR on the chart, another ability must be chosen instead. (See "Final Challenge rating" in the DMG section for rationale for "2 points".)

Also at each P-Level increase, allow one other improvement:

  • a damage die size
  • additional armor
  • some other bonus to a skill or action
  • or even an additional attack or special ability (I guess a "move" in Pokémon terms) as long as that does not increase anything beyond what's appropriate for the CR which maps to that level.

This would allow a lot of flexibility without needing to figure out everything in advance. You might need to space out the "one other increase" thing to every other level or every third level — that'd need to be play-tested, or at least invented by someone with a lot more knowledge of how Pokémon gain power as the level up than I have.

Have I ever done anything like this? No. But I've made monsters both on the fly and in advance using the DMG guidance, and it seems like it's well-suited to systematic application like this.

  • \$\begingroup\$ This system seems like a great start, but it doesn't actually seem to account for the fact that Pokémon are not made equal, a level 100 Caterpie is still going to die against a level 10 Mewtwo. One would have to decide what CR a level 1 and a level 100 Pokémon of a type should be, and then calculate the other levels as an inbetween, I'd imagine. \$\endgroup\$
    – Theik
    Commented Jan 18, 2018 at 12:02
  • \$\begingroup\$ @Theik Made edit. How's that? \$\endgroup\$
    – mattdm
    Commented Jan 18, 2018 at 17:03
  • \$\begingroup\$ I think you understood me halfway and missed me halfway. I don’t mind this though, because I still gained some knowledge from this answer. First, for future reference I am a DM. I know about CR. Still, while the Pokemon are being built to fill a role of “monsters”, their desired functionality is more a kin to that of a PC than it is a dire wolf. That said, that is merely a comparison, not a statement of what they will do. Pokemon will be Pokemon. The Level System I was talking about would be, in essence, a self-mapped level system. Not just a mere reflection of PC leveling. \$\endgroup\$ Commented Jan 18, 2018 at 18:50
  • \$\begingroup\$ However, I also planned on mapping CRs based off of their cumulative stats as well. I appreciate the new information about average ability scores. While I don’t think all of what you said is applicable, it certainly is all useful. Thank you! \$\endgroup\$ Commented Jan 18, 2018 at 18:50
  • 1
    \$\begingroup\$ Theik, actually, bad example. A Lv 100 Caterpie with abysmal stats, no item, and a disadvantageous Nature will always OHKO a Level27 and lower Mewtwo that is perfect in every way without fail. Even if Mewtwo manages to attack first, the most powerful attack it can do to Caterpie (Fire Blast) can only be a guaranteed 2HKO. Anyhow, what you pointed out was one of my points in the original discussion in the OP comments back when it was [On Hold], CR and Pokemon Levels wouldn't be perfectly comparable, and whatnot. \$\endgroup\$ Commented Jan 18, 2018 at 19:04

Merging Pokemon with D&D 5th edition is possible, but you have to overcome several inconsistencies between two very different rule systems.

Pokemon's 100-level system represents the higher numbers possible when a computer handles all the math. A Pokemon's stats might go from 5 to 250 within 100 levels, meaning there's opportunity for meaningful increase at each level.

Compare this to D&D 5th edition. If you started at 10 Strength and even gained even 1 Strength at each of 20 levels, you would break game balance. That's still not enough to actually give you something at each of 100 levels.

In other words, you have a dilemma. If you actually get something at each of 100 levels, your creature radically out-powers any standard D&D character of his level. But if you have a 100-level scale full of empty levels, you're really just playing a 20-level scale while keeping the big numbers for thematic purpose.

In either case, because Pokemon level up five times as often as PCs, you now have to stop and level up your Pokemon every two encounters or so. That's hassle because unlike in the video game you have to stop and do it manually.

Certainly, you can have a game where a player controls both a traditional character and that character's Pokemon ally. You're essentially just giving one player two characters, which is easily done.

A bigger problem, however, is this: If you want to balance PCs against Pokemon, you have to build the Pokemon on approximately to the same math as PCs, and that makes it difficult to authentically represent a game where different species of Pokemon have radically different statistics, strengths and weaknesses, even among creatures of the same level.

We also have to consider that D&D and Pokemon have always worked on fundamentally different paradigms when it comes to ability scores. Pokemon follows the JRPG tradition that characters progress by increasing ability scores with level, but in D&D your ability scores remain largely static and your power increase is largely from class abilities.

If you look at the actual mechanics of Pokemon, the stats you see in game are not actually equivalent to D&D's ability scores. In fact, they are derived from five sources: IVs, EVs, species, nature and level. IVs are equivalent to D&D's base ability scores, EVs are bonuses based on what enemy types you have fought (D&D has no equivalent), species determines stats far more radically than D&D's +2 racial bonuses, nature is a +/-10% bonus that closely matches D&D race in its magnitude, and level increases the final stats by a large degree.

Conversely, D&D 5th edition assumes that no single race is radically stronger than another, level bonus is fixed by level regardless of race, ability score modifier is a relatively large component, and ability scores don't increase much with level. In converting from one to another, you have to solve these basic differences as to what stats mean and how they work.

My suggestions, then, are as follows:

  • Pokemon in D&D 5e can work, but you're essentially creating a small character class for each Pokemon. That's going to be way more work if you have to make those 100-level classes, and you're going to struggle to find enough content to give something at each level, unless you give out too many bonuses, which will ruin balance relative to PCs of their level.
  • You must balance Pokemon and their progressions by equivalence to player characters of equivalent level, otherwise they will never be balanced relative to each other. If a paladin has +6 to hit while his Onix has +12, even if that matches the Pokemon lore or the numbers in the Pokemon game, it's not balanced in D&D rules, and your human PCs will soon play the part of Krillin watching superbeings fight it out.
  • \$\begingroup\$ I love your answer. Not because I agree with it, but because you comment a lot of the things I thought of when coming up with this system leading me to where I am at now. This may take a few comments... 1. Effects of Pokemon Level Up can be made during rests/evolution/[PC Leveling] as to prevent stuttering gameplay. I considered that problem. 2. Pokemon would be functionally a blend of monster and PC factors. They wouldn't be distinctly one or the other by nature of what they are. 3. Empty levels are more issues for Pokemon that don’t gain anything after like PkMn Level 15 \$\endgroup\$ Commented Jan 19, 2018 at 2:02
  • \$\begingroup\$ (i.e. Caterpie, basically only good for evolving.) That said, I agree, this can be an issue in certain cases, and I don’t have a solution for it yet. However, creating a L20 system for Pokemon presented me with more issues than it solved. 4. Pokemon would still serve a role like monsters. They need to be balanced against PCs like monsters are, while still being able to grow like PCs. This is the core issue at the root of my question. I need to know if my system would maintain this balance. 5. I know the stat systems are fundamentally different… This is an adaptation not a translation. \$\endgroup\$ Commented Jan 19, 2018 at 2:06
  • \$\begingroup\$ I have my own solution to the discrepancies. 6. I’m not trying to make Pokemon into PC races. I am trying to make them into something a kin to a PC, while still being in nature filling the roles of “monsters”. Under normal circumstances, you’d be right with your Krillin analogy, but remember: part of being a trainer is being on the sidelines. That said, I am wanting PCs to play just as active a role in a campaign as their Pokemon, hence my concern about the balance of things. \$\endgroup\$ Commented Jan 19, 2018 at 2:10
  • \$\begingroup\$ I don't say all of this to make an extended discussion in the comments about my mechanics. I mention it so you can use this to adjust your answer to be more immediately relevant to my question. If a mod takes issue with this use of the comments for correction, feel free to give me punishment, but I wanted to correct any misunderstandings Quad had. Due to the nature of his answer, it's hard for me to suggest improvements without pointing out his misgivings. Still, overall, one of my favorite answers. Combined with @mattdm's answer, you two gave me the info I needed. +1. \$\endgroup\$ Commented Jan 19, 2018 at 2:15

If you are balancing the pokemon around 20 break points like player characters are balanced around, there should be no problems. This would involve comparable AC, damage output, spell (or spell-like abilities) access, ability scores, and movement options to what are available to PCs of the adjusted level to avoid the level 5 vs level 1 imbalance. The hardest part will be porting over pokemons 6 stats to D&Ds many critical statistics which are more detailed and variable than Pokemon. For example, Defense and Speed at least would have to be factored in to AC since pokemon doesn't have Dexterity modified defenses as pokemon just take the hits (without boosted evasion).

  • \$\begingroup\$ What do you mean by "20 break points"? I want to make sure I understand your answer before I respond to it. \$\endgroup\$ Commented Jan 17, 2018 at 21:04
  • 2
    \$\begingroup\$ Break points adjust the scale of a scheme. Pokemon has 100 break points called levels. 5e has 20. Shifting to 20 means Pokemon levels 1-5 will be the same. \$\endgroup\$ Commented Jan 17, 2018 at 21:22
  • \$\begingroup\$ Ok, I thought that might have been what you meant, but the term sounded... odd to me. And, I appreciate the input. \$\endgroup\$ Commented Jan 18, 2018 at 6:28

Not the answer you're looking for? Browse other questions tagged .