How to fairly add a homebrew skill?

Question

tl;dr

I want to add some extra skills to D&D 5e but can't figure out how to do it fairly in terms of awarding proficiency.

How do I add additional homebrew skills to D&D 5e?

---

Why I want to add skills

I'm currently running a homebrew campaign where exploration and wilderness travel are a major component of the story. The way I run exploration involves a lot of skill challenges, it's how I determine the party's progress toward their goal and/or how safe the journey is.

Going in knowing that the campaign would be exploration heavy I felt that it would place too high an importance on the Wisdom (Survival) skill. It covers:

• Navigation (very common), which is used every time the party travels, typically at least once per day
• Tracking, rarer but still a relatively common task and when it occurs often requires multiple Wisdom (Survival) checks.
• Determining if a campsite is safe, once per day while travelling
• Foraging for food, depends on how desperate they are for rations
• Extracting materials/food from slain creatures, commonly done after killing larger creatures
• Making a mundane fire, mundane traps for hunting, and other basic real-world survival skills. Most of the time these don't require rolls, but if there are adverse conditions or time pressure it is a Wisdom (Survival) check.
• Estimating travel times and determining the safest route.
• As well as many other things I've surely forgotten.

In a normal campaign the rate at which these things occur would be roughly even with other skills such as the Charisma and Intelligence based skills. However in my campaign wilderness exploration represents roughly 1/2 the playtime, making Wisdom (Survival) about 5 times more common than I've experienced in previous campaigns. Therefore I wanted to add some additional skills to split up the exploration proficiencies and potentially allow different PCs to thrive at different elements of it.

The campaign takes place on a continent where most of the locals don't speak 'Common' as the party knows it. Overcoming language barriers and learning to live with the different cultures is an intended part of the campaign. Therefore I wanted to add a skill to reflect the PCs proficiency at learning languages by porting the Linguistics skill from Pathfinder as an Intelligence check.

What I tried

My plan was to add 4 additional skills.

• Intelligence (Linguistics)
• Intelligence (Navigation)
• Constitution (Endurance)
• Constitution (Cooking)

I considered a number of different ways to add them to the game and adjust the granting of proficiency to compensate. Even going so far as to create a complete replacement of the proficiency system styled on the Pathfinder skill rank system. But nothing quite hit the mark.

Here's a summary of the problems I encountered but couldn't solve:

• Proficiency Balance between classes
  • Interactions with Expertise
  • Jack of All Trades Bard Feature
• Feats and features that grant proficiency
• Space on character sheets
  • Custom character sheets are difficult to make/maintain
• Complexity
  • This one was the real killer. No matter what I tried the system felt like the added complexity wasn't worth the benefit to gameplay.

The best option I could find was just adding these to the game and then giving all my players one extra proficiency as a free choice on top those granted by their background and class. I ultimately decided against it and went with the RAW options.

What I'm asking for

15 sessions into the campaign I have confirmed that my gut instinct was correct. Wisdom (Survival) and Wisdom (Perception) have been by far the most used skills at the table. The players choose to use them in basically every skill challenge on top of the incidental use in regular play.

Therefore I'm revisiting this in the hopes that someone has a better solution for adding homebrew skills than anything I have come up with.

Answers should consider the balance impact for features like Expertise and Jack of All Trades as well as address how to distribute proficiency fairly so that the new skills actually get used.

If anyone has tips on a good way to add them to character sheets that's a nice bonus but not the main point of the question.

Answer 1

I agree fully with Thomas Markov's answer, but would like to extend it in a slightly different direction (or: this is too big for a comment, and is meant to compliment their answer).

First, I'll simply state that Wisdom (perception) is often cited as the most-rolled proficiency in the game, and that there's probably nothing that can be done to meaningfully change that. It's one of the few skills that players can specifically invoke at any time ("I take a quick look around the room"), and there's rarely a reason not to use it at least once in a given scene.

So, setting Wisdom (perception) aside and focusing on Wisdom (survival)...

I suspect that this is a bit of an XY problem, where the underlying concern isn't "how do I add skill proficiencies" but "how do I encourage use of proficiencies other than Wisdom (survival)".

I also suspect that two things are happening, to varying extents. First, that the players are responding to "here's the scene; what do you want to do?" with "roll Wisdom (survival)" instead of an in-game action. Second, that the OP is being generous with what Wisdom (survival) is meant to be able to do.

One of the first wilderness encounters I thought of while reading the question is "there's a river in front of you; you need to cross it". Remember that Wisdom (survival) is described (PHB 178) as

The DM might ask you to make a Wisdom (Survival) check to follow tracks, hunt wild game, guide your group through frozen wastelands, identify signs that owlbears live nearby, predict the weather, or avoid quicksand and other natural hazards.

Is a river a natural hazard? A generous interpretation might allow it, but there are much better skills to overcome the challenge of "we need to be on the other side of the river". The question then becomes how the PCs choose to get to the other side. Obvious options include:

• magic (eg., teleport, or even shape water) - there's almost always a spell that can bypass a non-combat challenge
• Strength (athletics) to simply swim across
• Intelligence (investigation) to "look around for clues and make deductions based on those clues" - looking for indications of where the local fauna cross the river
• Intelligence (nature) to "recall lore about terrain" and suggest that a safer crossing might be nearby in "that" direction (eg., "it looks like the terrain flattens out; the river might be flowing less swiftly a mile downstream")
• Dexterity (acrobatics) to walk across some larger stones that come up to the river's surface (possibly in conjunction with an investigation/nature check to find a place where such stones are likely)

Wisdom (survival) can look like a panacea for wilderness encounters due to the "avoid natural hazards" part of the proficiency's description, but "cross the river" isn't avoiding a natural hazard so much as overcoming it.

So, that's 200 words on the "encourage the use of other proficiencies" part; what about the game loop part?

First, in this particular case, I'd recommend a "session 0.5" to tell the players that the GM is frustrated that Wisdom (survival) is resolving too many challenges. Prepare for push-back on this: it could very easily sound like "rolling to attack is resolving too many combat encounters". Personally, I'd start by taking the blame for having made it too easy to use Wisdom (survival) to resolve the challenges, and asking the players to bear with me for a few sessions while I try to encourage other resolutions. And, as Thomas Markov's answer said, that I was going to award inspiration for at least the next 3 sessions for any solution that doesn't rely wholly on Wisdom (survival).

Second, remember the game loop. Make sure that, as the GM, you know how the PCs are attempting to overcome the challenge - what physical (or mental) actions are the PCs going to engage in in furtherance of their goal? In the river example, that might be "swim across" or "try to find a safer place to ford" or "make a makeshift raft to float across". In order, those would be Strength (athletics), Intelligence (nature or investigation), and honestly Wisdom (survival) (to make the raft) followed by probably Strength (athletics) (at a reduced DC from simply swimming) to go straight-ish across while fighting the current.

But: the player shouldn't throw out a proficiency (or even an attribute), they should describe the actions they're taking. That is, the conversation starting with "what do you do?" should go past "I roll Strength (athletics)" until the GM and player are on the same page about what the character is trying to do.

GM: here's a river; you need to be on the other side. What do you do?

Player: I roll Strength (athletics).

GM: Why?

Player: well, because I'm good at it.

GM: But, what is Ragnar doing that would require a Strength (athletics) roll?

Player: Oh, he's swimming across the river.

GM: Okay, the current's pretty strong here; the DC will be 20.

In this particular case, this GM would let the player change their mind if the DC was higher than they thought it'd be - maybe Ragnar tested the water and found it moving much faster than expected, maybe the player and I simply weren't on the same page about that detail of the river.

So, when a player wants to throw Wisdom (survival) against a challenge, it's incumbent upon the GM to ascertain why the player thinks that that particular skill would resolve the challenge. That is: what actions in-universe are the player's character taking that would be represented by a Wisdom (survival) check?

As a side note: in a social setting, this GM encourages a great deal of leniency in letting players pick between, Charisma (intimidation), Charisma (deception), and Charisma (persuasion), so long as the player can point to something they're doing that would trigger one over the other. They player needn't say the exact words that their character would use (any more than a player would need to demonstrate a particular swimming style to swim across a river), just hit the high points of how they're doing it - whether they're using flattery or arguing a greater good scenario (persuasion), suggesting that they have special permission from the king (deception), or suggesting that they have special permission from their swords (intimidation). Similarly, a solution like "fashion a makeshift raft" would be a simple skill check without the players needing to describe how they're choosing which materials to make the raft out of, how they're tying the knots, etc. (unless they'd recently lost their equipment, I suppose).

Answer 2

You’re having problems because your skill challenges are designed around metagaming.

The ability and skill system of 5e is not really built to do the things you’re trying to do with it. In your description of skill challenges, you write:

PCs choose a skill to attempt. Based on the narrative situation the players can choose a skill to assist them in reaching their goal. PCs may choose any skill in which they are proficient, provided they can narratively justify how that skill would help them. Each PC may only attempt each skill once.

This is somewhat backwards from the design of the 5e skill system:

The DM calls for an ability check when a character or monster attempts an action (other than an attack) that has a chance of failure. When the outcome is uncertain, the dice determine the results.

For every ability check, the DM decides which of the six abilities is relevant to the task at hand

Sometimes, the DM might ask for an ability check using a specific skill — for example, “Make a Wisdom (Perception) check.” At other times, a player might ask the DM if proficiency in a particular skill applies to a check.

The design for ability checks as explained above (and drawing from the “How to Play” section of the PHB intro) goes something like:

1. The player describes what the character is attempting to do.
2. The DM determines if it requires an ability check, and if so, which ability.
3. The DM determines if a particular skill proficiency may be applied to the check.
4. The player may then ask if a particular skill proficiency may be applied.

The ability and skill system of the game is built around this sequence of events. Your skill challenge system cuts out two of these steps and does the remaining two in reverse order:

1. The decides which ability and skill proficiency to use for a check.
2. The player describes what the character is attempting to do.

What this does for the players is it makes your skill challenge into “let me figure out how to shove Wisdom (Perception) and Wisdom (Survival) into the scene” instead of “let me read the scene and come up with an appropriate solution”. Because they know the rules of your metagame, they are playing that game instead of playing Dungeons & Dragons. They know that if they can just convince you to let them use what they are good at, they will succeed, so they are doing that instead of thinking about the challenge the character is facing and coming up with a solution appropriate for the narrative.

For this reason, adding more skills will not solve your problem: all you will do is change the skills and abilities your players try to shove into the scene. In fact, I find it likely that splitting Survival up into two or more skills will make it even more dominant. Your Skill Challenge has a rule that does help promote more diversity:

Each PC may only attempt each skill once.

Right now, Survival only gets one attempt per skill challenge. Split it into two different skills, now your homebrewed skill replaces some other attempt using a skill that was nothing like Survival. You mention that the particular adventure is ripe for using the Survival skill, so splitting Survival up into two different skills provides the players with two obvious choices, rather than one. I think this will do the opposite of what you are trying to achieve. Instead of encouraging creative, non-obvious solutions, you're just adding a second obvious solution.

The skill challenge metagame needs to go back under the hood.

The solution to your issue is to get back to the basics, as described in the “How to Play” section of the introduction to the Player’s Handbook.

1. The DM describes the environment.

First, set the stage for your skill challenge. Describe the environment, and present the players with the problem the characters are trying to solve. The emphasis is important here: the characters are trying to solve a problem in the game universe. This is not the player’s trying to beat a number. Another crucial aspect of this step is that the DM is describing the game universe, not the game mechanics. What I mean is that if you say “Alright time for a skill challenge”, you have switched from playing D&D to playing Link’s Skill Challenge, and the players will switch up how they are approaching the scene to succeed on the skill challenge. Instead, the checks for skill challenges need to integrate seamlessly into the game world, just like regular checks do.

2. The players describe what they want to do.

Next, the players describe how their characters are responding to the challenge you have presented, just like they do in every other situation. You’ve presented with a narrative that they are helping to write, rather than a scoresheet they are trying to fill up. Once they have described what they are trying to do, you decide what combinations of skills and abilities are appropriate for determining the outcome of their actions.

And this is where you, the DM, can help bring diversity to the skills and abilities used. Reward creativity and narratively appropriate role playing with advantage and/or Inspiration. When the players realize that engaging with the story, rather than playing a metagame, is the path to success, they will stop shoving the same square pegs into every hole they see, no matter the shape.

3. The DM narrates the results of the adventurers’ actions.

As usual, you narrate the results of their attempts, and you keep score of the skill challenge metagame as a too for determining the ultimate outcome. You don’t have to do away with the skill challenge system. You just keep it behind the DM screen as a tool for determining the direction the narrative takes. You still use it as a tool, you still use it to structure your narration and determination of events, but you allow the players to engage with the story without encouraging them to treat it like points on a board.

I really think just changing how you run skill challenges will fix your problems. Adding more skills definitely won't fix your skill challenges, and I don't see a reason to add Endurance and Cooking when Athletics and Cook's Utensils are already in the system. To directly answer the title question, "How to fairly add a homebrew skill?", don't. At best, you don't fix anything, at worst, you introduce the balance problems noted in your question, and you might end up shooting yourself in the foot by discouraging creativity with additional obvious choices.

Finally, minnmass's answer has some more practical commentary that does a great job of expanding on the ideas I've put forward here.

Answer 3

Define "fairly"

Your first issue is, how do you define "fairly" adding a skill? We'll keep asking that along the way.

"Mechanically" adding a skill

This seems more in line with what you're trying to figure out.

The simple answer, you just add it to the list. But simple doesn't tell the whole story.

What ability score are you tying it too?

Here is a breakdown of the current skills:

Ability	Skills
Strength	Athletics
Dexterity	Acrobatics, Sleight of Hand, Stealth
Intelligence	Arcana, History, Investigation, Nature, Religion
Wisdom	Animal Handling, Insight, Medicine, Perception, Survival
Charisma	Deception, Intimidation, Performance, Persuasion

Key takeaways:

1. There are NO Constitution-based skills. Your example gives two skills that do, so you need to tread lightly with creating those as there may be a developer reason for avoiding it. The reason could be as simple as, no one has Constitution as a dump stat so everyone should at least be +0 at the skill. Whereas everything else has potential for being "bad". Just a guess.
2. Dexterity, Intelligence, Wisdom, and Charisma all have 3-4-5 skills attached to them. That means adding more skills to these abilities will increase the power of characters based on those abilities. In other words, characters weak in these abilities will become more weak as they are not as useful for even more skill checks.

So depending on your definition of "fair", adding skills based on Strength would make steps to even the playing field, but some skills wouldn't make sense as Strength-based. Adding them to Dexterity would be the next lowest score. Adding them to Wisdom would make 6 skills (out of the current 18 total) would mean that one-third of all skills are tied to Wisdom meaning low-Wisdom characters are even more "back of the bus".

How do you gain the skill?

All characters learn skills in at least two ways--by class and by background. Some will also get one or two by race. By race and by class have predefined lists of what skills you do, or can, learn. For instance, Tabaxi automatically specifically learn Perception whereas classes learn 2, 3, or in one case 4 skills, from a short list. Sometimes the character can pick up one more via a sub-class. I'm not sure there is a "fair" way to add new skills to classes; it would be a trial-and-error thing. Same with adding by race; they already are tuned to be balanced. It would also mean that your players would have to choose what they play based on how you allocate the new skills instead of what they want.

Which leaves backgrounds... Backgrounds come with two predefined skills assigned to them, but per the rules, you are allowed to work with your DM to mix and match your own background and thus can select one of the new skills. But since backgrounds max out at two skills, dropping an existing skill for a new skill better be important. The skills you have listed are very specialized and players have no idea how relevant they will be to the campaign.

From a practical stand point, linguistics is based on Intelligence so most likely to be picked by Int-based classes (for maximum effectiveness). So that means Artificer and Wizards. But both only have 2 skill slots based on class, and for Wizards, one of which is almost always Arcana, which means they only have one class skill slot left. If available for the class, should they choose it on a skill that may or may not prove useful? Should they drop something from a background in order to squeeze in linguistics?

Or should the Bard, Sorcerer, Warlock, or Wizard just learn the comprehend languages spell?

With an average of only 4 skill slots available, picking an untested/unknown skill will not be a high priority.

To counter this, you can give everyone an extra skill slot. This would be a bit overkill if there is only one new skill but since you're introducing four, it is an option. However, unless you say, "It must be one of the new skills," there is nothing compelling to players to not choose something they know will be useful. And if you do force players to choose one of the new skills, four will pick the one that most closely relates to their modus operandi, while the rest pick at random to give advantage to the others.

Player 1: My Rogue with the high Charisma will take linguistics. That way as I learn their language I can sweet talk them into stuff.
Player 2: My Druid with Goodberry will learn cooking. Maybe I can make a Goodberry pie for even more healing.
Player 3: My Sailor is already proficient with Navigator's tools, so this navigation skill will mean I always get Advantage, right?
Player 4: My half-orc Barbarian will take endurance, because.. yeah.

They could also pick it up via a number of feats, but I'm not going to detail all those out. It's just more options.

In this case, "fairness" will be based on how the characters pick up the skills. As the DM, you better explain at session zero (or in your case session 16) that the characters will need these skills and what allowances you're making so they are not losing out on skills that fit their planned backstory or are vital to the class.

How often is the skill useful?

Are these skill only useful during the first part of the campaign, or will they keep coming up throughout the whole adventure?

Your example was that the party is on a continent where the locals don't speak Common. So, yes, when they first arrive, there is going to be a language barrier. But as time progresses, wouldn't they learn the language enough so that skill checks become pointless? By 5th level, a number of casters can get tongues and solve the whole problem. Couldn't they draw (or buy) maps so navigation is no longer an issue?

There are very few "must have" skills; Arcana, Stealth, Investigation, Perception to name a few. But there are a lot of skills that might never come up (History, Religion, Animal Handling, etc) based on the setting, so there is always a risk of useless choices.

But if you structure the campaign that these new skills are vital in tier 1 and 2 and redundant by tier 3, I would say they are not "fair" to the players as you are railroading them into wasting a resource. With the exception of Bard (talked about next) every character has a finite amount of skills they can learn. By making these extra choices a low-level requirement and a high-level space filler, with no way to swap them out (like spells), then the skills are not "fair", they are a tax.

Expertise and Jack of all Trades and Reliable Talent, oh my

A lot of this already works itself out. You can't say it's unfair to get expertise in Cooking, but not unfair to get expertise in Stealth. The same can be said for Reliable Talent. If you took the time to be proficient in a skill, having an automatic take-10 is not a big deal. A lot of these things have the trade off of investing levels into a class. Sure you are better at skills, but you don't get extra attacks for instance.

Jack of all Trades, is only slightly different in that the bard gets more things, instead improving something they are already skilled at. So the more skills you add, the better the feature becomes. It maxes out at +3 to a skill check at the top levels, so nothing that will kill bounded accuracy.

So is it "fair" with these features? You're giving a slight edge to Bards, but otherwise unremarkable.

Will the skill check be based on an individual or in group checks? Contested against something else?

If they are generally based on individual then players will just spread the love. One character takes it too excel and other takes it to help the first character and grant advantage.

If they are generally based on group checks (I could see endurance as the only practical example of this) then you may see an upswing in this skill.

Also remember that some skill checks are not against a DC but are contests. For instance challenging one creatures deception against another's insight. Will there be opposing skills for the one you're adding?

This consideration is less about "fair" and more about something to think about.

Does the skill bring something useful to the game?

I use this example a lot. I forget where I read it, but early on in my DnD playing there was an article about changing rules. The argument went something like this:

DM: I've got this cool new concept. Instead of air, the world is nothing but carbon dioxide.
Player: But then how will we breathe?
DM: You'll all have special respirators.
Player: What about NPCs? And monsters? How will they breathe?
DM: All the people will have one. And the monsters have adapted.
Player: But what if mine breaks? Seems pretty easy considering all the fighting we do.
DM: They can't be broken. They just always work.
Player: If everyone has a respirator, that never fails, and we can always breathe normally, tell me again why we are getting rid of air?

So does adding these skills add something to the game, or can you use existing rules?

As pointed out above, why get Linguistics when there are spells (comprehend languages, tongues, etc) and features (Monk's Tongue of the Sun and Moon) that will presumably serve just as well. The Navigation skill would also better be served with the Keen Mind feat which gives a +1 to Intelligence and always knowing which way is north, or just proficiency with Navigation tools.

Once again, this isn't to determine "fair", but to think about usefulness.

Does the skill bring something bad to the game?

Xanathar's introduced tool proficiency and the symbiosis with skills.

By adding new things there may be deeper repercussions; like my example of having a navigator skill and navigator tool proficiency. Is it redundant? Give advantage? Somehow separate?

Would the endurance skill help with exhaustion? With death saving throws? The half-orc's Relentless Endurance?

This is less about being "fair" and understanding how one change can have rippling effects.

Can these be handled better with existing skills/features/proficiencies?

Skills are tied to an ability; but they don't have to be tied to the default ability. The basic rules explain using a different ability for skills. So can you use Constitution (Athletics) instead of creating a new endurance skill? Intelligence (Insight) to learn a language?

And per Xanathar's:

Given enough free time and the services of an instructor, a character can learn a language or pick up proficiency with a tool.

So instead of creating a Navigation skill, give the players time to pick up Navigation Tools as a proficiency.

Once again, this isn't about "fair", this is about not over-complicating the system.

---

So to wrap up...

How to "fairly" add new skills

1. Make sure it is adding something to the game other than complexity
2. Make sure it can't be better handled by using existing or variant rules (skills with different abilities)
3. Balance how the characters will learn the skill with the limited number of slots available for choosing skills
4. Make sure it remains useful throughout the campaign and not just a section
5. Think about all the ways the skill will be used; both in terms of situations and rolling checks
6. Think about the repercussions of how the new skill overlaps spells, features, proficiencies, etc
7. Limit the number of skills added; one doesn't make a big difference, but 3, 4, 5, or more and Bards get a bigger and bigger boost
8. Get good with Excel/Google Sheets/Apple Numbers/other spreadsheets to create a custom character sheet
   • Alternately, D&D Beyond allows you to just add skills and I'm sure other VTT would also

Answer 4

You've already got some good frame challenges to this question--at the risk of being redundant, I'll recap that a bit, but then give you a suggested alternative.

This isn't metagaming

Contra Thomas Markov, the problem isn't metagaming. People in general try to approach problems by using the skills they're best at--it's why we have the expression "when what you have is a hammer, everything looks like a nail"--and there's a very thin line indeed between "I'll have my character use X because of their high proficiency" vs "My character would solve the challenge with X because that's what they're good at doing and how they approach problems generally."

Your adventure has an unusual focus

You've chosen to run an adventure in which an unusually large number of challenges are correctly addressed in RAW by one particular skill. So players use this skill a lot, and may feel obliged to take it.

• This makes in-character sense: Wilderness survival is a focus of this adventure, and it makes in-character sense for a party full of people who are doing this activity to spend some of their expertise budget on getting good at it. If your players signed up for this game with this focus but didn't want to take Survival, they're kind of sandbagging themselves and have no one else to blame.

• Skills With Different Abilities solves most of your problem (PHB 175). Have them roll Int(Survival) for navigating instead of Wis(Survival), and the Survivalist Cleric suddenly lets the Itinerant Wizard take the spotlight when it comes time to choosing a path. Similarly, I'm having a hard time imagining a case for your proposed Con(Endurance) that couldn't be addressed by Con(Athletics) -- a long swim is the specific example given in the PHB for varying the ability check; or Con(Cooking) could just be a Con(Medicine) or whatever--so if you're trying to force more diverse skill proficiency choices, there's already RAW options that are equally good to your proposed changes.

• Survival still doesn't catch everything -- just because it happens outdoors doesn't mean Survival is the right skill, as Minnmass covered. Tell them Survival isn't quite the thing.

Aside: On Language Learning

Your other proposed change was an Int(Linguistics) skill for learning languages. I'm not sure this is the right move. RAW (PHB 187) lists learning a language as a "Training" downtime activity that requires a gold outlay, not a skill check. This acknowledges that learning a language takes a long time, longer than the events of one adventure. I don't think it's plausible to have your characters actually learn a language through a few skill checks. I would refocus this challenge on communicating with people they can't mutually understand (well). Which can even be more diverse/interesting by spreading out the skills to communicate something from the skills to understand the other. This could be Cha(Performance) or Cha(Persuasion) [for conveying a message], or Int(History) [for obscure references or reconstructing word forms], Wis(Perception) or Wis(Insight) [for inferring what someone means from body language or context]--there's lots of existing tools for this, but they won't grant a language proficiency.

So you want to add skills

All that said, you'd like to avoid unbalancing the game by changing the number of skills. I think the key here is going to be to acknowledge your game's unusual focus and be clear on what activities you've dropped to make room for the wilderness exploration stuff. If as you say you're spending 5x as much game time on exploration as normal, that must have come at the expense of something else--what would you normally expect to happen during that time that isn't?

Figure that out. Then cut those skills. When you designed the adventure, you implicitly picked some skills that matter less than they normally do; combine them together. Want to add two novel survival skills? Great: Religion, Arcana, and History are now all the same thing. (It sounds like your PCs weren't doing much of any of them anyway.) Or combine Athletics and Acrobatics and just differentiate based on whether you ask for Str- or Dex-based rolls. Maybe Sleight of Hand is now part of Stealth, if there's never any pockets around for your PCs to pick. If feats or backgrounds give the same proficiency twice, let the player pick a new one for the duplicate.

You can view this either as staying within the limits of the skill budget; or if it makes you feel better, you can think of it as "sweetening the pot" for the skills that don't come up as much, to make them more competitive options for the skills that are obviously used all the time.

The balance problem here isn't how your new skills impacts the proficiency budget. The problem is that (possibly without telling your players!) you decided that some subset of the RAW skills are going to be unexpectedly unimportant. Hopefully they've responded by focusing on the ones they do expect to use--rather than just taking something that never comes up and feeling frustrated--but that's already impacting the skill economy. And that's probably where your biggest frustration is coming in--you've hit a point where the set of skills that characters reasonably expect to use is narrow enough that everybody has all of them.

Fortunately, the answer is the same as the cause: refocus the game around what you are doing by blurring out the things you aren't doing.

Answer 5

Stick to 17-19 Skills

• Proficiency Balance between classes
  • Interactions with Expertise
  • Jack of All Trades Bard Feature
• Feats and features that grant proficiency
• Space on character sheets

All these problems are dependent on the total number of skills and you can side step them entirely by sticking to a total of 18 skills. Just remove or combine some existing skills for every new skill you add. However, having one more or less skill won't make a significant difference to balance.

Consider the Abilities

Ability	Skills
Strength	Athletics
Dexterity	Acrobatics, Sleight of Hand, Stealth
Intelligence	Arcana, History, Investigation, Nature, Religion
Wisdom	Animal Handling, Insight, Medicine, Perception, Survival
Charisma	Deception, Intimidation, Performance, Persuasion

That's the spread of skills for each ability and you should try not to deviate too far from it, but the variant rule 'Skills with Different Abilities' is a strong indication that you don't have to be too exact and if you adopt this variant rule you can ignore this point entirely.
That said, be wary of adding Constitution Skills, I won't guess at the designer intent, but the absence of such skills is conspicuous in and of itself.

Define the Custom Skill List

Start with the existing list and make some changes from there: add, remove, split, and combine skills as you see fit. Consider that simply adding examples to the description of an existing skill can expand or restrict its scope.

You've identified aspects of your campaign that are dominant enough to warrant new skills. I posit that there must also be aspects of your campaign that have been reduced, enough to warrant fewer skills. Be wary of outright removing a skill instead of combining it with another, as this could lead to a large 'hole' in the activities covered by the skill list, but even such a hole wouldn't be a big problem, because you can simply call for an ability check instead.
For instance, if the social encounters are few and revolve primarily around the language barrier (and therefore the new Linguistics skills), then you could combine Persuasion and Deception into Talk. Maybe Arcana, History, and Religion can't compare to Nature because the latter overshadows them in the wilderness, so you could combine the former three into Scholastics.

how to distribute proficiency fairly so that the new skills actually get used

Exactly how to split Survival, and which skills to combine is up to you and mostly a thematic and stylistic choice. The default skill list is by no means fair or evenly used, not even close. So whichever custom skill list you create has a low bar to clear.
And I really mean that last point, so lets look at some numbers to nail it in. Critical Role can't stand in for every table, but they make for a well documented case study. At the time of writing, these are their skill checks:

Skill	Rolls	Percent
Perception	2869	28%
Stealth	2097	20%
Investigation	1294	13%
Athletics	645	6.3%
Insight	543	5.3%
Persuasion	517	5.1%
Acrobatics	371	3.6%
Deception	367	3.6%
Arcana	257	2.5%
Survival	202	2.0%
History	196	1.9%
Nature	193	1.9%
Religion	169	1.7%
Intimidation	141	1.4%
Sleight of Hand	133	1.3%
Medicine	102	1.0%
Animal Handling	71	0.7%
Performance	66	0.6%
All	10233	100%

That's right, 61% of all their skill checks are covered by only three skills: Perception, Stealth, and Investigation. The least and most used skills differ by two orders of magnitude. If Critical Role had used Survival five times as often it would still only be their 4th most used skill...

On one hand, this should make you wonder whether your perceived dominance of the Survival skill is actually that big of a problem. On the other hand, if you understand your campaign enough to perceive a problem with the skills and want to tailor a new list for that campaign, then you'd be hard pressed to do worse than default skill list.

Complexity and Rules

Now that you have your custom skills, their associated abilities, and thier associated descriptions, it's time to open your digital copy of the PHB and other books your players reference for rules, and use Ctrl+F to find skills that you have removed, split, or significantly altered. Skim through each rule that references these skills and make a call about which skill(s) these rules should use instead, or whether these rules are even relevant to your campaign at all.

For example, I decide to split Survival into Travel (planning a route, avoiding or overcoming minor natural obstacles, setting up camps) and Forage (hunting and gathering, including tracking), so I Ctrl+F for 'Survival' in the PHB and found only 25 matches. Of these, only four could be considered standalone rules: the Favored Enemy class feature, a sidebar about self-sufficiency, a paragraph about other activities to do while traveling, and the hunter's mark spell. The sidebar is truly unremarkable and the paragraph about other activities is informative, but overridden by your use of skill challenges, so I only note down that Favored Enemy and hunter's mark use Forage instead of Survival.
Every other match is either a proficiency lists which we can handle with a broader rule (see #2 later), definitions of Survival which are obviously overridden by my introduction of Travel and Forage, or false positives.

• Complexity
  • This one was the real killer. No matter what I tried the system felt like the added complexity wasn't worth the benefit to gameplay.

Unfortunately, some extra complexity is unavoidable. The best you can do is write a clear and succinct document for the players to understand the new skill list and which rules are affected. It should contain:

1. [A table, listing all the skills, the ability associated to each skill, the description of what each skill does, the corresponding defunct skills (if any).]
2. If a rule would grant you proficiency in a defunct skill, it grants proficiency in one of the corresponding skills instead, per the table. If you're already proficient in all the corresponding skills, it grants proficiency in a skill of your choosing instead.
3. If a rule alters or calls for a check with a defunct skill, then it alters / calls for the check with the corresponding skill instead, per the table. If there are multiple corresponding skills the DM chooses which, or can give you the choice. Here's some that are already figured out:
   • [Favored Enemy and hunter's mark uses Forage instead of the defunct Survival.]
   • [etc.]

On the DM side, you don't need to change the skills of NPCs (including monsters), most don't even get the opportunity to use their skills, and when they do it's fine if you roll, say, Survival, even though that skill doesn't exist for the players. If NPC abilities, traps, or what-have-you call for a defunct skill from the player, just make a ruling on the fly, you'll have the practice from the previous PHB Ctrl+F search.
The DM side and player side need not be symmetrical: you are altering the skill list because the players face a lot of skill challenges during exploration, but NPCs are not direct participants in these skill challenges.

Character Sheets

• Custom character sheets are difficult to make/maintain

The solution here depends entirely on which sheets your table uses or is willing to use. The answers vary from 'use white out tape to erase defunct skills from your paper sheet' or 'that virtual sheet has a built-in option' to 'it's impractical with the sheet you're using now, you'll have to switch'. It might be best to open a question dedicated to this point.

Conclusion

The default skill list is neither fair nor even in its use of skills, so a custom skill list has a lot of room for error and is likely to be better in these respects given that you know the specifics of your table and of your campaign better than WotC ever could. What the default skill list does best is cover a wide breath of activities, so as long as the custom skill list covers a similar breath with a total of 17-19 skills, it will work well within the existing system.

Answer 6

Add just one skill, or add one free pick and 3 skills

The other answers focus on how to handle game situations that call for skill checks and essentially frame challenge the question, advising you to not add skills. I do not neccesarily disagree (I would not add skills, myself). Still this answer is trying to provide some consideration for how to actually add skills if you want to do it, without affecting game balance in a major way.

Just one skill: The game by default has a total of 14 skills. So adding just one more, without adding additional skill picks for characters, effectively increases the number of skills to pick by 7%. This is probably not so much that you need to worry about it, especially if you consider that there are also 17 different artisan tools that can to some extend stand in as skills. You could just add Intelligence (Linguistics), if cultural languages are a major part of your campaign, without needing any other changes, and it should be fine.

However, adding four different skills starts to seriously dilute the PCs capabilities. Now there are suddenly nearly 29% more skills than before that they need to cover with their picks.

A typical race has on average 1 skill, a background has 2, and a typical class has on average a bit more than 2. Thas means a character has maybe 5 skill picks to work with.

Three skills and one pick: Based on these numbers, the "fairest" way to introduce more skills would be to introduce 3 additional skills (about 21% more skills), and give the characters one extra skill pick (about 20% more pick capacity) to cover them.

Possibly, you could add one more pick and three more skills multiple times for even more skills. However, this depends on how evenly distributed the power and value of the different skills is. In a typiccal campaign Wisdom (Perception) is a lot more valuable than Wisdom (Medicine), so unless you balance it really well, by adding more free picks you allow the characters to pick up all the high-impact skills, and ignore the bulk of what you add. To address this, you could limit the new picks to just the new skills, for example by assigning a defined extra skill as part of backgrounds in your world, but this would quickly become more complicated. I therefore think, if your goal is to just tweak the system to support what you want, going down this route is not worthwhile.

If you only add three skills and one pick, and want to avoid this cherry-picking effect, you can limit spending it on the new skills.

Expertise eventually will give you four picks (for both bards and rogues), so you could likewise add one additional skill to pick for the three skills you add, probably for the option at first level, optionally at sixth/tenth level instead. For feats like Skill Expert, that grant expertise in a single skill, you normally will pick one specific skill you need for your character build, so having more skills should not materially influence that.

Jack-of-All-Trades I think there is nothing you need to do about that. The function is that the Bard has natural aptitude in any check he tries, and that does not change with the number of skills. Before you got a bonus on any conceivable skill check, and after you get a bonus on any conceivable skill check. Whatever set of skills you use will covers all possible situations and the bard should get the bonus in all those situations. For a simple example, if you split Wisdon (Survival) into several new skills, they would have gotten the bonus on survival, and now they get it on whatever new skill they use. Nothing really changes. And if you balance the added skills with an additional proficiency, the other characters also do not become relatively weaker in doing skill checks on the wider range of skills in a meaningful way.

Answer 7

You list a number of potential items that you might want to have different skill checks for:

• Navigation (very common), which is used every time the party travels, typically at least once per day
• Tracking, rarer but still a relatively common task and when it occurs often requires multiple Wisdom (Survival) checks.
• Determining if a campsite is safe, once per day while travelling
• Foraging for food, depends on how desperate they are for rations
• Extracting materials/food from slain creatures, commonly done after killing larger creatures
• Making a mundane fire, mundane traps for hunting, and other basic real-world survival skills. Most of the time these don't require rolls, but if there are adverse conditions or time pressure it is a Wisdom (Survival) check.
• Estimating travel times and determining the safest route.
• As well as many other things I've surely forgotten.

Your stated option is to add a number of additional skills that move the reliance away from Wisdom (Survival), however you faced a number of different issues that you would like to resolve:

With a number of problems with adding those skills that you encountered, with complexity of the solution being the biggest problem.

Thankfully D&D 5e already has a way to handle things like this, by utilising alterative abilities paired with the skill itself.

The PHB provides for this variant rule explicitly (in Using Ability Scores > Ability Checks):

Variant: Skills with Different Abilities Normally, your proficiency in a skill applies only to a specific kind of ability check. Proficiency in Athletics, for example, usually applies to Strength checks. In some situations, though, your proficiency might reasonably apply to a different kind of check. In such cases, the DM might ask for a check using an unusual combination of ability and skill, or you might ask your DM if you can apply a proficiency to a different check. For example, if you have to swim from an offshore island to the mainland, your DM might call for a Constitution check to see if you have the stamina to make it that far. In this case, your DM might allow you to apply your proficiency in Athletics and ask for a Constitution (Athletics) check. So if you’re proficient in Athletics, you apply your proficiency bonus to the Constitution check just as you would normally do for a Strength (Athletics) check. Similarly, when your half-orc barbarian uses a display of raw strength to intimidate an enemy, your DM might ask for a Strength (Intimidation) check, even though Intimidation is normally associated with Charisma.

The DMG states (in Running the Game > Proficiency > Skills):

Under certain circumstances, you can decide a character’s proficiency in a skill can be applied to a different ability check. For example, you might decide that a character forced to swim from an island to the mainland must succeed on a Constitution check (as opposed to a Strength check) because of the distance involved. The character is proficient in the Athletics skill, which covers swimming, so you allow the character’s proficiency bonus to apply to this ability check. In effect, you’re asking for a Constitution (Athletics) check, instead of a Strength (Athletics) check.

Often, players ask whether they can apply a skill proficiency to an ability check. If a player can provide a good justification for why a character’s training and aptitude in a skill should apply to the check, go ahead and allow it, rewarding the player’s creative thinking.

So, both the PHB and DMG give us this freedom, in appropriate situations, to apply alternative abilities to skills. We can use this rule to achieve the aims you want, without requiring additional skills.

Examining what the game tells us about some of the existing skills and abilities will help here:

Strength: Strength measures bodily power, athletic training, and the extent to which you can exert raw physical force.

Dexterity: Dexterity measures agility, reflexes, and balance.

Constitution: Constitution measures health, stamina, and vital force.

Intelligence: Intelligence measures mental acuity, accuracy of recall, and the ability to reason.

Wisdom: Wisdom reflects how attuned you are to the world around you and represents perceptiveness and intuition.

Charisma: Charisma measures your ability to interact effectively with others. It includes such factors as confidence and eloquence, and it can represent a charming or commanding personality.

Skills:

Nature: Your Intelligence (Nature) check measures your ability to recall lore about terrain, plants and animals, the weather, and natural cycles.

Survival: The DM might ask you to make a Wisdom (Survival) check to follow tracks, hunt wild game, guide your group through frozen wastelands, identify signs that owlbears live nearby, predict the weather, or avoid quicksand and other natural hazards.

We can use the above to come up with appropriate checks for each of the checks you describe depending on the situation.

Taking your examples from above, I would suggest the following, depending on how the players intend to meet the requirement:

• Navigation: Wisdom (Survival) or Intelligence (Survival)
• Tracking: Wisdom (Survival) or Constitution (Survival) or Dexterity (Nature)
• Determining if a campsite is safe: Strength (Survival) or Dexterity (Nature) or Wisdom (Nature) or Wisdom (Survival)
• Foraging for food: Intelligence (Nature) or Intelligence (Survival) or Wisdom (Nature) or Wisdom (Survival)
• Extracting materials/food from slain creatures: Dexterity (Survival) or Dexterity (Nature) or Intelligence (Nature) or Intelligence (Survival)
• Making a mundane fire, mundane traps for hunting, and other basic real-world survival skills: Dexterity (Survival) or Dexterity (Nature) depending on what they are attempting to do
• Estimating travel times and determining the safest route: Intelligence (Survival) or Intelligence (Nature) or Wisdom (Survival) or Wisdom (Investigation) depending on the situation

This removes the need for complexity and is something that can be communicated in advance to your players that you will be using alternative skill/ability combinations depending on what they are trying to do. It also avoids any of the balance concerns you have since the game was designed with this in mind.