Monthly Archives: October 2015

Another dice as resources post: the mathening.

I wanted to test my intuition from last time that the dice expressions I was giving were fair.

So I monte-carlo’d it, results at bottom:

from random import randint as r
import numpy
dice = [20,12,10,8,6,4]

def experiment(i=0, critical=1, failure=3):
  loop = 0
  while True:
  if i>=len(dice):
    return loop
  val = r(1,dice[i])
  if val <= critical: i+=1
  if val <= failure: i+=1
  loop += 1

results = {}

for i in range(len(dice)):
  results[i] = [experiment(i) for j in range(10000)]

for i in range(len(dice)):
  print "d", dice[i], "failed in", \
    numpy.median(results[i]), "median", \
    numpy.mean(results[i]), "mean"

output = """
d 20 failed in 15.0 median 16.3829 mean
d 12 failed in 10.0 median 10.7343 mean
d 10 failed in 7.0 median 7.6144 mean
d 8 failed in 4.0 median 5.0133 mean
d 6 failed in 2.0 median 2.8603 mean
d 4 failed in 1.0 median 1.3304 mean
"""

I kind of like those odds, and I kind of don’t. Mean higher than median means that the winning tail is long — if you go big, you likely go (relatively!) very big — and in all cases, these are numbers that actually come up and matter in play.

The d4 feels very slightly mean — er, I suppose, “cruel” — to me. Its odds of failure are so vast.

I tweaked it and ran the same experiment with 1-2 instead of 1-3 (still “doubling down” on 1s as double failures). Those odds are:

d 20 failed in 20.0 median 22.6337 mean
d 12 failed in 13.0 median 14.7605 mean
d 10 failed in 9.0 median 10.5486 mean
d 8 failed in 6.0 median 7.0481 mean
d 6 failed in 3.0 median 3.9912 mean
d 4 failed in 1.0 median 2.0154 mean

These feel a little fairer to me. There’s a point to rolling, even at the d4 (though, yes, your odds suck). Importantly, the d8 matches the by-the-book 7 charges which a wand should have, so I can use it there. That’s cool.

I’m definitely going to try these out for morale and loyalty — my two hidden NPC stats. I’m also probably going to reintroduce these for wands, scrolls etc.

I could use these for other things, too — death saves feel particularly “fun” to roll on your hit dice, with failures shrinking your hit dice (but not hit points) until such time as you can re-grow them. Wouldn’t work super well for characters whose hit dice are exhausted or multiclass. Might just say “a single instance of your largest hit die, used purely for this tracking mechanism” or something.

Advertisements

One more cut at the NPC

What, you thought I was done? Getting there, not there yet!

I like Morale; it can stay.
I like something in the same vein as Loyalty, but maybe not Loyalty itself.
The whole way I was doing the bond thing is… Working. Clumsy. Let’s try that part again. We’ll just stick it into Loyalty.

We don’t actually need granular bonds. I mean, we need freeform descriptors to create texture for our NPCs, obviously, but we don’t need them to double up as a progress track.

We also don’t need a sliding scale of more- and less- friendly that it implied, not as a piece of general tech.

Using Morale:
Check Morale for each follower upon the first seen death each short rest, and the first time it is bloodied each short rest. The DC for this saving throw is 10.
When an NPC fails a Morale check, it gains the Frightened condition with respect to the cause of the check: the enemy who caused the death or damage.
Indifferent NPCs also check upon rolling for initiative.
Hostile NPCs also make this save with disadvantage.

That’s more or less all that causes Morale checks; additionally saves against fear effects may substitute Morale for the specified statistic.

Remember that you calculate Morale as the better of Wis or Con.

Using Loyalty:
Loyalty are a way of tracking dedication. The default mechanism in the DMG (page 93) is a little heavyweight and requires that I remember things, gross.
As with all hidden data schemes, we can substitute randomness. Staunch allies use a d12; shaky ones use a d4, and Indifferent all the dice between.
Loyalty can be tested, or improved.
Testing Loyalty happens when the party acts counter to Bonds or other traits of the NPC, shrinking the die on a 1-3 — shrinking it two steps on a 1 if the test was egregious!
Roll the test each time the party acts counter to the Bonds.
Improving Loyalty happens when the NPC is treated particularly well — given gifts, etc — using the same mechanism as acquiring the follower. If the resultant die is larger than the current one, the die grows one size.
DC <0: d4 and hostility and betrayal.
DC 0: d4
DC 5: d6
DC 10: d8
DC 15: d10
DC 20: d12
DC 40: d20
The party knows when they've been bad or good; they know the size of the die most of the time. The DM updates the party on the status of the NPC during long rests or other times they might settle up the hidden properties of magic items etc. Because we decrease Loyalty with a die roll, the party knows they're in danger but aren't sure just how much.

You could use the same mechanism for Morale!
Start with the creature's hit die, adjust 1-2 sizes by wisdom modifier (Con correlates pretty strongly with size). Check Morale in the conditions above, failing on a 1-3.

Finally, Insight. It's just a skill. But players totally lie to their NPCs all the time.

Managing NPCs
Prepare: DC 15 Cha check; success grants allied followers advantage on saving throws for a round.
Incite: DC 15 Cha check; success gives allied followers advantage on attacks for one round.
Rally: Frightened allied followers may re-test morale.


Resources as dice

I’ve spoken before about using dice to track ammunition and rations — roll once per appropriate time period to determine when you’re down to your very last oat-cake or whatever.

I’m not sure I love this, but here it is anyway.

How about gold and hoards? Not magic items, but specifically bulky loot which is just money and weight.

Each character sheet has some tally of “treasure dice” — 4d4+5d8+1d12 or whatever.
Each treasure die weighs, say, 5 lbs of encumbrance.
Copper treasure (and mundane goods like cloth) is worth a d4; silver (and most trade goods) d6, gold (and exotic trade goods like spices) d8, platinum d10, gems a d12, and exotic currencies like souls a d20.
Size-wise, tiny objects weigh and grant 1 die. Small 2, medium 4, large 8, huge 16, modified by density as appropriate.
So a golden statue is 4d8; a solid mithral door (valuing mithral at platinum) 8d10, a sea-chest of silver 2d6, a dragon’s hoard 32d8 (or more!), etc.
Tiny valuable objects (single gems, jewelry) are basically weightless; they can be sold for some (usually small) loot die expression. A large enough haul of gems that makes it no fun to track them is just some high end loot expression again!
Usage example: “You find 3d6 of silver in the chest”, “You find 4d4 of brownweave”, “There’s easily 8d4 of beer barrels here”.

An interesting side effect: if budgeted for as weight, the number of dice a single character can carry is pretty manageable. 10 or 12 dice is pretty much the max per haul, plus gems and art objects.

You can split up dice; splitting 1d(X+2) gets you (X/2)d(X) — 1d12 buys 5d10; 1d6 buys 2d4.
You can partially reverse this process and combine dice; spend (X/2)d(X-2) to get 1dX — 1d6 costs 3d4; 1d12 costs 6d10.
Instead, you can purchase weightless wealth (jewelry, art objects) for a single die of the next larger size.
The d20 doesn’t participate in this system of exchange; it’s too ludicrously valuable. You can exchange it down into 6d12.
You pay a premium for transportable wealth!

Things you want to buy also have a dice expression. I suggest using the same guidelines as above: items bought in hundreds of copper use cost dice d4s; hundreds of gold in d8s; draining the treasury is paid in d10s and d12s for castles and ships.
Costs may aggregate up and do so at the rate by which wealth is broken down; 10 d10 purchases are resolved as a single d12 purchase.
In particular, cheap ~2gp purchases of mundane gear are d4s, common items at 20gp are d6s, uncommon are d8s, rare d10s, very rare d12s, and legendaries d20s.
Within those tiers, feel free to use multiple dice; for instance, for consumables,
Cantrip: 2d4 (or 1d6)
1st level: 2d6
2nd level: 2d8
3rd level: 4d8 (or 1d10)
4th level: 2d10
5th level: 4d10
6th level: 2d12
7th level: 4d12
8th level: 6d12 (or 1d20)
9th level: 2d20
And permanent items should cost +1 die — or more.

To purchase a thing, determine the thing’s dice expression — its cost.
Put forward an equal number of dice (of sides equal to or greater than the cost dice of the thing).
You’ve now bought the thing.

However, it wiped out your ante, which might not be entirely fair.
Roll your purchase dice and roll the cost dice. Arrange both lowest to highest. Retain any purchase dice which scored strictly higher than the cost dice. Shrink any dice which didn’t roll higher one step. D4 are discarded on shrinkage.

Treasure dice above a d6 are rolled as d6 in villages. You can sell larger dice directly into 3d6, skipping intermediate steps.
Small towns, d8 (and sell larger into 4d8).
Large towns, d10 (5d10).
Cities d12 (6d12).
Metropoli d20.

Downtime costs a die expression per week: Wretched is free. Otherwise:
1d4 squalid, 2d4 poor; 1d6 modest, 2d6 comfortable, 3d6 wealthy; 1d8 (and up) aristocratic.

Henchpersons add to this expression their own needs. An unskilled henchperson costs 2d4, a skilled one 2d6, and an expert one 2d8. These are lifestyle costs; a henchperson might also arrange for shares of treasure.


Social Interaction in 5e part II

Part 1 is here.

Okay. We have a specific mechanism: NPCs have (relatively fixed) Loyalty, Morale and Insight scores, and then a laundry list of preconceptions, traits and grudges, their ideals, flaws and bonds (IBFs).

Now let’s talk about attracting henchmen and keeping them loyal.

Your average hired hand has a Bond “The Party” +1, as well as Morale and Loyalty equal to its wisdom save, +/- whatever fudge factor the DM is using to make them more complex (probably a +/- 2 point swing). I’ll be using those stats a lot, so I’ll just be referring to the “party bond” hereafter — it means the strength of the NPC’s relationship to a specific character or to the party as a whole, as is appropriate.

In addition, the kinds of things that can affect bond strength adversely actually operate in similar fashions, so have a handy Bond Damage chart. To use it, roll a check (specific to the type of situation) and compare the total adjusted result below:
<5: Party bond damaged by 1d6/2 (1d6 if doubled, 1d3/2 if halved)
<10: Party bond damaged by 1d3/2 (1d3 if doubled, 1/2 if halved — track fractions!)
<15: Party bond damaged by 1/2 (track fractions! 1d3/2 if doubled, no effect if halved)
<20: No effect (1/2 if doubled — track fractions!)
20+: No effect (even if doubled)
In general, lowering the party bond to -5 or lower as a result of this chart results in the NPC turning on the party (or at least triggering a fight-or-flight response!), and lowering the party bond to -2 or lower as a result of this chart results in the NPC balking at the situation and becoming undependable (but not actively harmful).
One way to do the fractions is to track whole numbers, and whether the bond is “tenuous”. A half-point of damage to a non-tenuous bond makes it tenuous. A half-point of damage to a tenuous bond makes it non-tenuous — but reduces it by 1!
To roll a d3/2 on a d6, treat 1-2 as 1/2, 3-4 as 1, and 5-6 as 3/2.

Risk

When the NPC is exposed to risk, check for Bond Damage via Morale + Party Bond. Sources of risk include:
1) Entering a known dangerous locale, half damage (the dungeon entrance, and then again whenever being asked to scout ahead or put self at a known risk).
2) Being reduced to- or below- 1/2 its maximum hitpoints, or having a party member fall (normal damage).
3) Exposed to friendly fire, double damage.

Temptation, Sloth and Discipline

When the NPC is exposed to an opportunity to better their situation by acting selfishly, check for Bond Damage via Loyalty + Party Bond. Examples of circumstances under which you should check loyalty include:
1) Upon beginning a specific unpleasant or long-duration task, like searching a sewer for something, guarding a prisoner for a week, sorting through paperwork, and so forth. Adjust the damage per the fittingness of the request; someone who knew going in they’d specifically be guarding a latrine takes half damage at that task; someone who thought they would be eating gourmet meals but ends up on latrine duty takes double damage.
2) Upon being exposed to privation: forced marches, low rations, no sleep, or no shore leave. A good proxy is “each time a check for exhaustion is made, whether passed or failed”. Read the “turning on the party” result as fomenting mutiny amongst anyone with a party bond < 0.
3) The Lure of Gold. When treasure (food, drink, friendship, a magic item — bribe, loot or happenstance) is available to the NPC but the PCs prevent it from having the treasure.

Charm effects, Fear effects

The first time each short rest that an NPC is subject by a charm effect or a fear effect, whether or not it makes its save (if any), lower its Party Bond by 1. If the effect originated with the party, defer the change until after the effect ends. This means that recruiting an army to go against the fey or against a dragon or something is going to be a problem: NPCs do not enjoy having their minds messed with.

Making Friends and Influencing People

It’s not all bad news! The good news is that bonds can also be strengthened, at great cost of effort and gold. Remember that a +5 is the sort of loyalty one should expect of a parent/child or spouse, a +2 the sort of loyalty one should expect from a well-paid mercenary eager to do a job, +1 from a paid mercenary willing to do a job, and a +0 from someone with no active reason to be distrustful.

Downtime activity: Recruit.
You can amass followers and hirelings — or at least bodies — depending on your stated goals, reputation, timeline and price point. What you can gather depends on where you are and what’s available. This is going to work a lot like rolling for treasure: You’re determining who’s in the labor pool and whether they’re willing to sign on. There can always be one-off special laborers (rare as magic items) in the pool, additionally. When recruiting for a specific set of skills, roll based on the CR for the number of hirelings available.
CR 2 (or costing 10gp/day — may request an additional full share of treasure)-> d4
CR 1 (or costing 5gp/day — will generally request an additional half-share of treasure)-> d6
CR 1/2 (or costing 2gp/day — skilled base pay, may request an additional half-share of treasure)-> d8
CR 1/4 (or costing 1gp/day)-> d10
CR 1/8 (or costing .5gp/day)-> d12
CR 0 (or costing .2gp/day — unskilled base pay)-> d20

Up to 1% of a given settlement’s population is available for hire, and they are generally available from least skilled to most skilled. As a result:

Village: Up to 10 hirelings are available — usually 10 CR0.
Town: Up to 60 hirelings are available — usually 20 CR0, 12 CR1/8, 10 CR1/4, 8 CR1/2, 6 CR1, 4 CR2, 1 CR 3
City: Up to 250 hirelings are available — usually 80 CR0, 48 CR1/8, 40 CR1/4, 32 CR1/2, 24 CR1, 16 CR2, 4 CR 3

Each downtime day spent recruiting results in the given die size in recruited hirelings of the requested quality at the given price. Recruitment rates can be greatly improved by spending more — allow an additional roll if players are doubling base pay and providing other incentives, or have access to an actual craft guild to connect hirelings with work.

The number of hirelings in a settlement returns to the listed number after a season (Say: 90 downtime days).

Perhaps just recruiting isn’t enough.

Downtime activity: Bond. You can devote time and treasure to improving the bond of the worker with the party.
Spend 1d10 days. Spend 10 times the base pay of the hireling (+ number of days). Make a Charisma (Persuasion) check, opposed by the hirelings Insight + Party Bond (the more they already trust you, the harder it becomes to improve their connection).

If you succeed, improve the party bond by (1d6)/2. If you fail by fewer than 5, improve it by 1/2. If you fail by more than 5, reduce it by 1/2.

This activity may be combined with carousing, and may be repeated indefinitely.

Downtime activity: Train. You can devote time and treasure to improving your hirelings, outfitting them with new gear and improving their skills.

Spend 1 day and 10 times the base pay of the hireling. Make an Intelligence check. You can improve the Morale, Leadership or Insight of your hireling, or give them a proficiency which you have, by the following schedule.
>20: 1.5 points, to a maximum of +5 (or your proficiency modifier, whichever is lower), taking 1d4 additional days.
>15: 1 point, to a maximum of +4 (or your proficiency modifier, whichever is lower), taking 1d6 additional days.
>10: .5 points, to a maximum of +3 (or your proficiency modifier, whichever is lower), taking 1d8 additional days.
<10: Failure, taking 1d10 additional days.

You may train half your check result in hirelings simultaneously, each additional hireling costing one additional downtime day overall and another charge of 10 times their base pay.

Combat Action: Rally. Make an Intimidate check. Half your result in hirelings who have taken bond damage since their last short rest may immediately roll Morale + Bond. If they beat 10, they gain +1/2 bond. If they beat 15, they get +1 bond. There is no penalty for failure.

Combat Action: Plead. Make a Persuasion check, and agree to a bribe. Half of your result in hirelings who have taken bond damage since their last short rest may immediately roll Loyalty + Bond. If they beat 10, they gain +1/2 bond. If they beat 15, they get +1 bond. There is no penalty for failure.


Dominance: A boardgame

I really like boardgames where the board includes random potential: Settlers of Catan with its random tiles.

This is pretty rough, but… Let me know how you feel about this one.
You will need:
One set of 6-dot dominoes (so 28 tiles)
4 “white” and 4 “black” stones (pawns, tokens)
4 six-sided dice

Set the seven dominoes with blanks aside, in order (one, two, three, …, six, blank). This is The Underworld. For purposes of game mechanics, tiles in The Underworld can hold an infinite number of pawns.
Lay the others out randomly in a 3×7 grid. This is The Battlefield. Each domino tile is divided in halves into two spaces, the two numbers on the domino. This means we’ve just made a 6×7 board of spaces.

Taking turns, roll pairs of dice, and place two stones on the corresponding tile (one in each space) — for instance, if you roll a 4 and a 3, place your stones on the domino 3,4 tile.
If there are already stones there (you’ve duplicated a roll), return them to the owner.
Continue doing this until all stones are placed.

Once all stones are placed, the game begins. Players take turns. Within a turn, the player will cast the dice, move their stones, capture stones, and return stones.

In more detail: roll all four dice; spaces bearing those numbers are “active”.
Stones of your color on active spaces* may move to other orthogonally adjacent spaces, and may repeat this movement until they move to a non-active space. Stones cannot move into or through spaces already containing a stone. This phase includes movement in The Underworld, which works in the same fashion (save that spaces in the Underworld are infinitely large and so can be stepped into or through even if already occupied).
Now each of your stones still on an “active” space may capture: step into an adjacent space containing a stone of either color; place the stone in The Underworld tile with the same value as The Battlefield space it just vacated. Each stone may only capture once per turn. You may capture your own stones.
Finally, if any of your stones are on the blank Underworld space (having moved into it from the 6 space during the move phase), they return to the board. Similar to the beginning of the game, one at a time roll two six sided dice. You may place the reincarnating stone in either of the two matching spaces, and if it was already occupied, capture the offending stone (placing it in The Underworld as normal).

If none of your stones are on active spaces during your turn, I feel badly for you but you receive no other recompense.

The game ends when all of one side’s stones are in The Underworld — the other player wins.

Strategies: On The Battlefield, well-connected runs of numbers are powerful, since they give you a lot of mobility. Pieces that are on 6s have a big advantage in The Underworld. Having a variety of different numbered spaces gives you better odds of being able to activate on your turn, but will probably somewhat limit mobility.


* This is my big dissatisfaction: to determine if a space is active you have to peek under your stones. Might work best with translucent pawns?

Open questions: allow additional capture of pawns in The Underworld? A sop for having no tokens on active spaces? It’s a bit swingy, but I think I still feel like I’m making decisions that matter — am I? Do they?


Social Interaction in 5e

I’ve been nibbling around the edges of Courtney Campbell’s On the Non-Player Character, but have thus far been dissuaded by price tag (a whole other topic: he can of course charge whatever he likes!)

What do we do when we see something shiny but unavailable (unpalatable, just out of reach way over there, scuffed up, it looked at me funny)?
We make one of our own!

Lay of the Land

5e gives us NPCs with loyalty scores that are a sort of “passive reaction check” (dmg 93). It gives NPC followers a “cost”, lifestyle expenses and a share of experience (the latter of which I don’t use).
They begin Friendly, Hostile or Indifferent (dmg 244), and can be temporarily or permanently shifted a step via interactions in a loose, story-based way. They expose their traits via insight checks and suggests waiting until the point is reached and then making a single charisma-based check, outcome determined by current friendly-level.

Good, but (modulo specific numbers), I have summarized the whole if the rules system as provided. The rest are left to squishy DM adjudication.

Let’s see what we can do about that! Throw the above system out. We’ll be rebuilding it; we’ll wind up pretty close to that, but not identical.

Let’s give the NPC some social stats.

One wisdom save isn’t enough to represent willpower; let’s split it in thirds.
Loyalty: by default, this is the creature’s Wisdom (charm) save. For loyal or trusting but dimwitted creatures, you might substitute Charisma as the base statistic. This statistic generally opposes Persuasion.
Morale: by default, this is the creature’s Wisdom (fear) save. For brave or disciplined but dimwitted creatures, you may substitute Constitution as the base statistic. This statistic generally opposes Intimidate.
Insight: by default, this is the creature’s Wisdom (insight) skill, but one might usefully substitute an Intelligence-based score or a perception based score. This statistic generally opposes Deceit.

Now, give the NPC Ideals, Bonds and Flaws like a PC would have.
These get a call-out, because they come with a rating, sort of like an attribute
The rating of a bond varies between -5 (extremely weak) through +5 (extremely strong), and applies to Loyalty, Morale and Insight checks when appropriate.

Flaws, in particular, might be used to “cut against” Bonds and Ideals; if a guard has the following stats:
Loyalty +3
Morale +1
Insight +1
IBF: My King +2, Alcoholism +5
then we’re in trouble. If a spy catches wind of the guard’s Alcoholism, they might attempt a bribe; they can roll (the spy’s) Persuasion + (the guard’s) Alcoholism vs (the guard’s) Loyalty + (the guard’s) My King . I’d rather take a +5 to my roll than a +2 — wouldn’t you?

Only ever roll one bond per die roll, like proficiency. Both IBFs should always come from the same character — the passive one.

This gives us a good mechanism for individual interactions — does the NPC betray the party? Roll Loyalty + Bond against a DC set by the temptation and find out! Does the NPC break under pressure? Morale + Bond!

Naturally, we’ll want everyone to have hugely positive bond towards us. We can do this by inducing or creating one, a skill check which is opposed by their insight (if no more appropriate stat) and to which the bond strength itself harms us! (The more friendly we already are, the harder it is to get friendlier-still).
No matter how high you get a bond, it returns after a short rest, usually.

How about a framing scene? Now that we have a way to track an NPC’s state and loyalty over time, we need a way to tell when they get fed up and leave.
The answer: depends on scene. Pick a number of “exchanges”, and freeze things after that many. I suggest a charisma check — 1 + 1 per 5 points on the check.


The Missing Fey

Still on my fey kick. My old post on Illyrians is all well and good, but those guys are hella complex. Let’s think about something a little lower power. Our theme is “curses” — these fey control fate at least a little, and bestow curses and geases, instead of straight up charms.

And, because the monsters I need all got placed elsewhere, we’ll be stealing monsters from other types. A-here we go!

For inspiration: the Lamia (which I shall consider a fey) and the Rakshasa (ditto). Those anchor our lineup at CR 4 and 13.

For another sample point, we can steal the Fomorian Giant, at CR 8. Another creature we shall drag bodily into feyhood.

That gives us a good set of curse-y monsters. But 3 monsters a subtype doesn’t make yet. We’ll need one or two at the bottom — CR 1/4, 1, around there — and at the top, around 15 or 16 — and then we’ll be good.

Curse Sprite
This is mostly just the monster manual “sprite”, but several of its mechanics are “cursier”.
Tiny fey, unaligned
Armor Class 15 (leather armor)
Hit Points 2 (1d4)
Speed 10ft., fly 40ft.
STR -4 DEX +4 CON +0 INT +2 WIS +1 CHA +0
Skills Perception +3, Stealth +8
Senses darkvision 60ft., passive Perception 13
Languages Common, Elvish, Sylvan
Challenge 1/4 (50 XP)
TRAITS
Sneak attack. This creature’s attacks deal 1d6 extra damage against targets of its curse.
ACTIONS
Multiattack. Bite, then Sprite’s Curse the target of the attack.
Bite. +2, 1 piercing.
Sprite’s Curse. One creature within 30ft. That creature is magically cursed to be unable to perceive creatures of the fey type while they hold their breath, treating such creatures as invisible and hidden. A DC 10 Wisdom save negates, and may be repeated each time the victim’s eyes are washed (an action) or the creature suffers damage from a fey.

Curse Walker
A strange warrior fey, filling a role sort of like a giant spider (‘pon which it’s based). Giant spiders tie you up and do poison damage, right? Right. Well, the Curse Walker kinda fills a similar niche; it steals your spells (a little) and uses them against you (a little) and ties up your ranged characters (the jerks).
Medium Fey, unaligned
Armor Class 14 (leather armor)
Hit Points 27 (6d8)
Speed 30ft..
STR +1 DEX +3 CON +0 INT +2 WIS +1 CHA +2
Skills Stealth +7
Senses darkvision 60ft., passive Perception 10
Languages Common, Elvish, Sylvan
Challenge 1 (200 XP)
ACTIONS
Rapier. Melee Weapon: +6; hit: 7 (1d8+3) piercing.
Fling Stone. Ranged Weapon (30/120): +6; hit: 10 (2d6+3) piercing. The attack may originate from any point within 30 ft. of the Curse Walker.
Lodestone Curse. Ranged spell attack (one target, 60 ft.): The target is magically cursed; they are vulnerable to piercing damage, have disadvantage on ranged attacks, and disadvantage on checks to avoid becoming lost. Each time such a target casts a spell of 1st level or lower, they must make a DC 13 spellcasting attribute save. If they fail this save, the most irritating possible legal target within 10 ft. of the intended target becomes the actual target of the spell (the DM may decide). When first placed this curse may be negated with a DC 13 Wisdom save; thereafter it may be removed with a DC 13 Wisdom save at the conclusion of a long rest in which the victim slept on iron.

Lamia
As written in the Monster Manual, but is considered to be of type Fey; it’s more lizard than lion, and so its “touch of idiocy” is a “caress of idiocy” (it may also use its claws as, well, claws). I really liked the 4e turns-to-bugs lamia-as-archwitch creature. Some other time, I guess.

Fomorian Giant
I stole the name Fomor for my own purposes, so these will have to shift. I suggest Haggard Giant, for the pun on “Hag”, and so. Still: they gaze, they warp with their gaze, they’re jerks.

Rakshasa
Minor changes. They’re fey, not fiends; their “piercing from magic weapons wielded by good creatures” is cute and all but actually becomes “piercing from magical iron weapons lowers their Limited Magic Immunity to 2nd level spells and lower until the start of their next turn”.

Fata Morgana
An aspect of some incomprehensibly magical fey being, the Fata Morgana represents a hard to pin down semilegendary fey ruler. Its shadowy nature is reflected in its elusiveness and mastery of curses [and cursed items!]
Large fey, unaligned
Armor Class 18 (natural armor)
Hit Points 130 (20d10 + 20)
Speed Fly 40ft. (hover)
STR +2 DEX +3 CON +1 INT +3 WIS +2 CHA +4
Saving Throws Dex +8, Con +6, Wis +7, Cha +9
Skills Perception +7, Stealth +8
Damage Resistances acid, cold, fire, lightning, thunder, poison, nonmagical weapons
Condition Immunities grappled, restrained, forced movement.
Senses blindsight 60ft., extending into the ethereal plane and around obstacles, passive Perception 17
Languages Common, Elven, Sylvan, telepathy 30ft.
Challenge 14 (11,500 XP)
Shadow Existence. The Fata Morgana can move through other creatures and objects as if they were difficult terrain. She takes 5 (1d10) force damage if she ends its turn inside an object. The first time she enters a creature’s space on a turn, when a creature enters or starts its turn in her space, or if a creature touches or hits her with a melee attack while within 5 ft., that creature is subject to the Lesser Curse.
Curse Striker. The Fata Morgana deals 1d10 extra force damage to targets per curse they are affected by, and per cursed item they carry.
REACTIONS
Puppetry (Recharge 5-6). One subject to the Greater Curse begins its turn within 1000 ft. The subject acts as the Fata Morgana wishes and may not take bonus actions, reactions, or cast spells above 6th level for one round. This is not a charm effect, but it is blocked by any effect which prevents paralysis.
ACTIONS
Caress. Melee Spell Attack: One target, or one ethereal target, within 15 ft. is subject to the Lesser Curse. If it is affected, it takes 11 (2d10) force damage (and Curse Striker).
Cursed Visage (Recharge 5-6). Ranged Spell Attack: One target, or one ethereal target, within 120 ft. who can see the Fata Morgana and is already affected by the Lesser Curse becomes subject to the Greater Curse. If affected, it takes 22 (4d10) force damage (and Curse Striker).
EFFECTS
Lesser Curse. A target subjected to the Lesser Curse must make a DC 17 Charisma saving throw or be affected. An affected target has disadvantage to all checks of an attribute chosen by the Fata Morgana, and suffers 1d10 + Curse Striker damage each time they cast a spell. This curse lasts until removed with remove curse.
Greater Curse. A target subjected to the Greater Curse must make a DC 17 Charisma saving throw or be affected. An affected target has disadvantage on all attacks and saves of an attribute chosen by the Fata Morgana to which they already had disadvantage on skill checks.