Occultist's Coven Familiar Bond in a Barovia FoundryVTT Game

 One of my players in my Curse of Strahd campaign just started a character based on KibblesTasty's homebrew Occultist v1.1.  We've moved it from Wisdom-based to Intelligence as the group has zero intelligence-centered characters and it feels right for someone to be smart in a group trying to survive in Barovia, and as a largely arcane caster it seems appropriate for the character to push intelligence. 

A more challenging portion of the character has been the Familiar Bond options available to the three coven's defined in the post of the class. Specifically, how to implement one of them in a FoundryVTT based campaign without making them annoying.  

In this post, I'm going to briefly consider each of the three and then dive deeper into what seems like the best-fit ability. 

Familiar Bond

At 3rd level, the Occultist can acquire a familiar that comes powered up with one of three special abilities. They are as follows:

• Green Coven - When you complete a long rest, your familiar gains illusory duplicates equal to your Wisdom modifier. These persist until destroyed by taking damage or you complete a long rest. The illusory duplicates cannot take any actions, but move and act like your familiar during your turn, staying within 30 feet of you or it. At the end of your familiar's turn or when it takes damage, it can swap places with an illusory familiar. If it does this as a result of taking damage, the illusory familiar is destroyed but the familiar takes no damage. Illusory duplicates are immune to damage while within 5 feet of you. You can assume illusory familiars are in the witch's space unless otherwise specified.
• White Coven - When you complete a long rest, your familiar gains temporary hit points equal to your Wisdom modifier + twice your Occultist level. While your familiar has any temporary hit points, as a reaction to a creature within 30 feet of it being attacked, you can have your familiar teleport to that creature and give it some of its temporary hit points (minimum one). If the familiar has no temporary hit points remaining, you cannot use this reaction.
• Black Coven - As a reaction to a creature within 30 feet of your familiar hitting with an attack, you can have your familiar subtract 1d4 + half your Occultist level (rounded down) from the attack roll. You can choose to use this feature after the creature makes its roll, but before the DM determines whether the attack roll succeeds or fails, or before the creature deals its damage. It can do this a number of times equal to your Wisdom modifier, regaining all uses after you complete a long rest.

Green Coven 

The green coven ability creates a cloud of copies of the familiar.  Adding potentially five more tokens to encounter maps that all need to stay with 30 feet of the owner. Managing that many tokens is not something that anyone should relish which means they are likely. to be assumed to be hanging around the owner unless one is going to be used to provide mobility to the familiar with its teleport ability. 

With most/all of the copies handled as an assumed presence (immune from damage) near the ower they really just become a pool of instant reincarnations of the familiar. This is easy enough to manage in FoundryVTT as a resource pool, but having familiar with 6 lives may not sit well with the GM or other players at the table.  It's just odd to have a familiar that just can't die. 

We have been playing this ability and have taken to mostly annoying it as the extra tokens are just a lot and what seems like infinite familiar lives feels annoying (to all but the owner of the familiar).

White Coven 

The white coven ability amounts to a pool of temp hit points that can make the familiar much more sturdy and allow it to act as essentially a life-transference emergency healer.  At 8th level with a capped primary stat the familiar would be packing 18 temp hit points, significant, better than it might appear on its face as they are applied as a reaction which with fast-forwarding would likely be after damage application and not after an attack / before damage is applied.  This would likely amount to a familiar able to save some number of characters from fatal wounds on a very timely basis. 

Implementing this in FoundryVTT wouldn't be very hard, though it might be implemented as reactionary healing as the damage will have already been applied in a fast-forward game setting. It becomes more complicated when damage is truncated at zero on attacks that drop a creature below zero.  This opens up a can of mathematical or manual worms that might best be avoided. 

Overall, this is a very useful ability as it is likely to keep several characters standing in tough fights and it has some implementation challenges. 

Black Coven 

This ability is quite similar to a Lore Bard's Cutting Words.  It shares a number of features that make it challenging in a fast-forward FoundryVTT setting.

1. Reaction - Always a bit fun to deal with in FoundryVTT as automation approaches leed to a pop-up for the player on every attack, until used each round and a delay for everyone else. 
2. Use after attack roll before success/failure or damage roll - This time is pretty literally non-existent in FoundryVTT with fast-forwarding forcing a change to the ability or elimination of fast-forwarding. 

The idea behind two is likely to force a bit of a guessing game on the Occultist player.  They can't be certain if their ability will make a difference and they don't know the stakes; was it a high damage roll? This creates an interesting gameplay element that FoundryVTT steps on hard if it is applied after the attack is resolved. 

From my experience with Lore Bards, cutting words often is handled after the attack in question is completed.  There is still a game tension as the variability in cutting words is high.  That isn't the case with this ability, which I think is the core of the problem.  

Since this ability is so close to Cutting Words, I think a direct comparison of them might be helpful.

Black Coven vs Cutting Words

First, a recitation of RAW Cutting Words:

Also at 3rd level...When a creature that you can see within 60 feet of you makes an attack roll, an ability check, or a damage roll, you can use your reaction to expend one of your uses of Bardic Inspiration, rolling a Bardic Inspiration die and subtracting the number rolled from the creature’s roll. You can choose to use this feature after the creature makes its roll, but before the DM determines whether the attack roll or ability check succeeds or fails, or before the creature deals its damage.

Cutting Words has a longer range, 60 vs 30 feet, but it is from the Bard, not the familiar making this seem like a wash. Effectively the same from a power perspective. 

Cutting Words can be applied to ability checks and damage rolls in addition to attack rolls.  Flexibility is a good thing, though the ability is most often used against attack rolls giving a slight edge to the Lore Bard.

Both abilities are to be applied after the attack roll (presumably with an announced result on the d20) and before the result is calculated (not that big a deal as players typically have a good idea of what's going to be a hit or miss after a round or two) or damage rolled (this is more important).  Implementing either in a fast-forward environment is going to force the ability to be applied after the fact, boosting the power of each.  

The boost is dependent on the predictability of the ability's usage.  The bard will be rolling a single inspiration die (d6 to d12) which has a huge variability while the occultist will be adding a d4 to half of their level making the effect highly predictable for mid to high-level characters. 

The range of results and impact of each ability is vastly different, with the Occultist's ability at most levels being far, far more powerful.  The accompanying table shows the minimum, average, and maximum results for both of these abilities at various levels. 

The bard wins this comparison only at levels 1 & 2, with lower, significantly lower impacts at higher levels. A 10th level bard when choosing to use cutting words can anticipate something between a 1 and a 10.  The equivalent occultist is looking at values ranging from 6 to 9.  

This gives the occultist a huge boost in applying the ability only when it matters if it is done after the attack is resolved.  Heck, it is a huge benefit if it is handled RAW with some uncertainty on the front end.  

I see two problems with the Black Coven's ability:

1. Minimal variability in the roll,
2. Very large numeric impact.

The lack of variability plays poorly if attack results are announced before the application of this ability.  This is exacerbated by using Fast Forwarding in FoundryVTT. 

The numeric impact is out of line with the bounded accuracy of 5E.  Mid to high-level Occultists are rocking minimum/average modifiers much higher than anything else in the game.  

Combining these two results in nigh on absurd power five times a day...the Occultist can make the five most significant (non-critical) attacks of a day become misses.  This puts the Divination Wizards near godly Portent ability to shame.

Looking at the numbers, believe the Occultist's and Bard's abilities are similar mechanically, but the Occultist is vastly more powerful and problematic. All of this ignores the fact that the Bard has other uses for inspiration dice which reduces the number of cutting words in some situations below the uses the Occultist has. 

How to Fix the Occultist?

I don't much like any of those abilities in my game environment.  Green is annoying and the other two are super powerful in a fast-forward world.  

My player has suggested giving the occultist a passive debuff effect on attacks or hits that imposes a small penalty on her target's next attack.  That is easy to implement in FoundryVTT and if small enough not a big power swing, though even a -1 or -2 on the next attack applied over a bunch of attacks in a day has a big average impact.  Sadly, this approach seems awfully boring.  It's not likely to result in any hero moments, just disappearing into the game mechanics.  

I'm tempted to just replace the whole occultist calculation with the bard's inspiration die progression.  The variability of that approach makes allowing it after the attack is fully evaluated a real choice for the player, with the exception of an attack that just barely hit. While the much lower average impact of the bard's ability is in keeping with bounded accuracy.

A Step Further

The essential effect of having the player choose to use this ability after the roll and before the results are announced is the introduction of uncertainty.  The player knows if the d20 roll was good enough to maybe or perhaps certainly to hit the target before they commit to using the ability but they don't know for sure and they don't know the stakes (was it a high or low damage roll). 

Allowing the ability to be invoked after the attack eliminates that uncertainty and with it one of the tasty gameplay elements.  But what if we could retain some interesting uncertainty on that choice.  Perhaps the ability just fails 15-25% of the time? That might make the choice more interesting.  

Consider a boost to the size of the die rolled for the effect, a d4 becomes a d6, a d6 becomes a d8, a d8 becomes a d14 (a VTT isn't restricted to a physical die's reality concerns), and so on.  A larger die increases the variability.  

But the average result is significantly bumped up by this.  That d8 averaged 4.5 but the mythical d14 pumps out an average of 7.5, or 3 points higher.  Well, to keep the average in line, apply a penalty of that average shift.  A d8 roll becomes a d14 - 3 which still averages 4.5 but has a 21% chance of yielding zero or less (which is equivalent to zero since it would never be used on a missed attack).

This chance of zero or less is a rough approximation of the uncertainty eliminated by allowing the choice to be made after the attack is completed. 

The larger die roll does increase the maximum possible result, which might be interesting, but is likely irrelevant as players will most often use this when an attack just barely hit.  But still, more variability isn't bad and this is a bit of an upside.

I think this best be thought about a bit longer.