D&D 5E Sorcerer experiment - your thoughts?

[Quickleaf]

My thoughts:

Mishaps: So you need to be careful with these, the second you create a % chance that the sorc can no longer "do their job" in combat, you seriously cut in to the effectiveness of the class. I would recommend two key guidelines here:

• The spell always does what it is supposed to do. Now that doesn't mean it can have some additional negatives, or some lashback, or things like that. But....never actually make the sorcerer spend an action and do nothing (or even worse, do nothing and hurt themselves).
• Mishaps should not hurt your party members: Who the heck would ever want to adventure with someone who might blow up at any moment and kill them? (this is the number 1 issue with the wild sorc, they can drop a fireball on their party at 1st level and TPK the whole group....not ok).

Overchannel: The fact that they can cast any sorc spell higher than them is actually seriously cool, not just for the power, but for the versatility. You might consider adding the ability to use sorc spells of a level you can cast, but not ones you know. One of the few things that triggers an exhaustion that is honestly worth the effort. That said, I don't see any reason to force a player to put this on their list, that creates needless tracking and pigeonholing. This sorc is chaotic, sure they cast that reverse gravity that one time but when their powers matured they actually got crown of stars or whatever. Nothing wrong with that as far as I'm concerned. Why force the player to play a certain way, let them enjoy the flexibility that their very expensive exhaustion cost grants them.

Spell Emulation: Does this work with higher level spells as well? (in which case its need to be noted about the overchannel penalties), or can they literally just cast any spell they have seen?

Great feedback! I appreciate it, Stalker.

1) Mishaps

While I deliberated avoided any overarching rule that might say "if you fail your cast roll, your spell doesn't cast", you're right that several of the miscasts really negate the sorcerer's action. Definitely miscasts #4 (random target), #6 (reversal), and #10 (delayed) I need to add a proviso at the end: However, you regain the action you used to cast the spell, and can immediately cast a spell without needing to roll to cast it. I think that might meet my intent (that sorcerer's spells can go very wrong), while also meeting the game need (not to negate the sorcerer player's turn).

2) Overchannel

Exactly, it's probably more for the versatility than strictly raw power...yeah, your suggestion about casting sorcerer spells of a level you can cast but that you don't know synchs up with what I'm aiming for. While I really like @TwoSix 's idea about "learning by doing / developing spell known through play", I agree with you that's probably too much needless burden of tracking on the player.

3) Spell Emulation

Yeah, the sorcerer could use Spell Emulation concurrently with Overchannel to (a) emulate a spell they see, that is (b) of higher level than they could normally cast. Maybe to clarify this synchronicity I should include language under Spell Emulation that "for 10 minutes that spell is added to your sorcerer list until you cast it"?

EDIT: Oh, I can just remove the "sorcerer" bit from "You may attempt to cast sorcerer spells beyond your normal ability" and it all works fine.

The idea is that all of these features start to play together the more an encounter goes on and the more the player plays their sorcerer, so it starts simple enough, but then the mishaps start to pile up, they start taking bigger risks, start burning through spell slots due to mishaps, burning through Sorcery Points to try to avoid mishaps / mitigate Overchannel exhaustion, etc. Like a flaming ball rolling down the hill gaining momentum, I hope the play style feels more frenetic than other spellcasters.

 

[Quickleaf]

agreed, I would definitely have negative, neutral (something different happens) and positive effects based on the result.

Personally I’d probably roll in caster level and spell level as well, so low level spells tend to get stronger when cast by experienced magic users. Never formalized this though

Yeah, that makes sense. I'm trying to avoid straightforward "you upcast the spell 2 levels" type of benefits, and make them more double-edged. How do these two "mixed benefit/drawback" miscasts look?

11= Metamagic backlash! You may apply one form of Metamagic (even one you don’t know) to the spell for free. However, you suffer a major effect related to the spell, lasting for 1 hour. For example, comprehend languages might cause you to speak backwards or modify memory might inflict amnesia on you.

12= Unstable conduit! You radiate magic for 1 minute, and any spell you cast in that time is randomly upcast 1d4 levels (without increasing the DC to cast). However, you have disadvantage on your rolls to cast in that time.

 

[Stalker0]

Yeah, the sorcerer could use Spell Emulation concurrently with Overchannel to (a) emulate a spell they see, that is (b) of higher level than they could normally cast. Maybe to clarify this synchronicity I should include language under Spell Emulation that "for 10 minutes that spell is added to your sorcerer list until you cast it"?

I think that language works nicely. Its clean and then works well with your other mechanics.

 

[Stalker0]

The biggest thing as a DM you need to worry about with overchannel is downtime. In an adventuring day scenario, the exhaustion is a hell of a penalty, it balances itself.

But lets say the sorc has a weak of downtime. A 1st level sorc could cast Wall of Stone (which is now permanent), and then just sleep off the exhaustion during the week with no downside. Suddenly your low level characters are hitting way above their paygrade, suddenly high level divinations like commune become an option, etc.

 

[MoonSong]

I don't love it. I'd prefer an evolutive approach towards the sorcerer. So far every new iteration has just remade the wheel instead of iterating on it. I'm not against experimenting, but if we only experiment, we'll never have a sorcerer that is just allowed to be and we'll never have a sorcerer that isn't hampered by experimental new mechanics. Particularly, I'm not a fan of picturing all sorcerers as chaotic, unstable timebombs. Might be fun for a once on a lifetime character, but it makes a class that is already hindered very much unplayable. Just by having so few spells known right now, the sorcerer is already a liability to the party if not optimized correctly.

I'd rather have people experiment with other classes instead.

 

[Quickleaf]

The biggest thing as a DM you need to worry about with overchannel is downtime. In an adventuring day scenario, the exhaustion is a hell of a penalty, it balances itself.

But lets say the sorc has a weak of downtime. A 1st level sorc could cast Wall of Stone (which is now permanent), and then just sleep off the exhaustion during the week with no downside. Suddenly your low level characters are hitting way above their paygrade, suddenly high level divinations like commune become an option, etc.

Brilliant analysis. Yeah, the intent is that none of the Overchanneled or Emulated spells may be permanent in effect. I'll have to figure out the wording on that.

 

[Quickleaf]

I don't love it. I'd prefer an evolutive approach towards the sorcerer. So far every new iteration has just remade the wheel instead of iterating on it. I'm not against experimenting, but if we only experiment, we'll never have a sorcerer that is just allowed to be and we'll never have a sorcerer that isn't hampered by experimental new mechanics. Particularly, I'm not a fan of picturing all sorcerers as chaotic, unstable timebombs. Might be fun for a once on a lifetime character, but it makes a class that is already hindered very much unplayable. Just by having so few spells known right now, the sorcerer is already a liability to the party if not optimized correctly.

I'd rather have people experiment with other classes instead.

Totally get your sentiment. If I had time and playtesters to design it from ground up I would probably lean toward some kinda Word+Noun / Improv Magic system, but time alas is not on my side.

 

[Steampunkette]

So... the two big pitfalls I can see are:

1) Too many rolls to resolve for speed of combat.
2) Too much Ad Hoc

For the first one, let's say your Sorcerer casts, or tries to cast, Melf's Acid Arrow. So you have to roll 1d20+Mods against a dc 12 to cast the spell and roll a 1. You still cast the spell, but it may do something silly. So you roll a d12 on the table for mishaps. THEN you roll your attack roll, THEN you roll you damage roll. THEN you figure out how it all went down and whether you also hit yourself.

That is a lot of slowdown at the table.

For the second one: Because some of the effects are "Reverse" or "Empower" or other "Mishap" that requires a per-spell decision point on how it goes silly, with at least 3 variations per spell on your list, you're gonna wind up with a lot of "Um... uh..." to extend the sorcerer's turn.

Setting up 12 "Set in Stone" results (or 20, or whatever) would definitely help with the back half of the slowdown. But you're still gonna wind up throwing at least an extra d20 roll every time a spell is cast, and a d12 roll every time a bad roll comes up.

Ways to Streamline:

1) Bake the "Mishap" roll into the Attack Roll or Saving Throw.

If you fire off the Melf's you get the mishap if your attack roll is below 8 minus your Charisma modifier. So you start out with a 40% chance of Mishap but with maxed Charisma (+5 bonus) you're down to a 15% chance of mishapping on a 3 or less. For multi-targeting effects, pick an "Initial Target" for your attack roll to trigger Mishap on.

For a Saving Throw put it on the other end. If your enemy rolls a 12 or better (Modified upward by Charisma modifier) you get a Mishap, with minimum 15% chance to mishap. For AoE spells pick a single "Primary Target" in the area to check for Mishap.

Downside: Some spells cannot mishap. Magic Missile, Silent Image, Prestidigitation, Power Word: Kill. Some spells don't have attacks or saves.

2) Reduce chances for mishaps to "a Crit" and apply it to the attack roll or saving throw.

Nat 20 save or Nat 1 attack triggers mishaps. Nothing else does. But instead of a primary target, AoEs all trigger off any Nat 20 save, and multi-attack spells trigger off any Nat 1. Multiple Mishaps in a single casting become a wild possibility.

Downside: Some spells still cannot mishap.

3) Player-Controlled Mishaps.

Player declares a spell they cast will Mishap. Player rolls the d12 and gains Inspiration or a number of spell points back or some other benefit in exchange for having a mishap. Player cannot mishap more than once per round, and can only mishap a number of times equal to their proficiency bonus per long rest.

Downside: Giving player control removes some of the 'Wildness' of the mishaps. Though they are still perfectly 'wild' within the narrative of the setting/story/game, happening inexplicably as far as the sorcerer is concerned.

 

[Quickleaf]

So... the two big pitfalls I can see are:

1) Too many rolls to resolve for speed of combat.
2) Too much Ad Hoc

For the first one, let's say your Sorcerer casts, or tries to cast, Melf's Acid Arrow. So you have to roll 1d20+Mods against a dc 12 to cast the spell and roll a 1. You still cast the spell, but it may do something silly. So you roll a d12 on the table for mishaps. THEN you roll your attack roll, THEN you roll you damage roll. THEN you figure out how it all went down and whether you also hit yourself.

That is a lot of slowdown at the table.

For the second one: Because some of the effects are "Reverse" or "Empower" or other "Mishap" that requires a per-spell decision point on how it goes silly, with at least 3 variations per spell on your list, you're gonna wind up with a lot of "Um... uh..." to extend the sorcerer's turn.

Setting up 12 "Set in Stone" results (or 20, or whatever) would definitely help with the back half of the slowdown. But you're still gonna wind up throwing at least an extra d20 roll every time a spell is cast, and a d12 roll every time a bad roll comes up.

Absolutely. I appreciate you walking through some solutions! I'm ok with some ad hoc, but I'm with you on this slowing things down too much.

Initially, I thought I would have the miscast be incorporated right into the Roll-to-Cast so that this would only add one extra d20 roll to the process. That got tricky to design...

Roll a d20, and if the sorcerer fails the check, then the face value of the die is cross-referenced to the miscast table. No rolling a d12 or anything. The idea was, to use your Melf's Acid Arrow example, the sorcerer would roll d20 + Cha (let's say +3) + proficiency (+2) against DC 12. In this case, they fail on a 1 thru 6, and succeed on a 7 or higher. Therefor, consulting against the twelve-row miscast table, only entries 1 thru 6 might come into play for this spell. BUT, if the same sorcerer attempted Reverse Gravity, they would roll against DC 17. In this case, they fail on a 1 thru 11, bringing more of the table into play. So the higher numbered rows would indicate more risk, so probably would be worse effects, while the lower numbered rows would indicate less risk.

Hmm, actually... writing that out makes me wonder if maybe I might have a grasp on how to do that now that I've messed around with the design a bit.

 

[Andvari]

I was thinking of adding miscasts to get rid of tracking spell slots (or rather spell points, which is my current method), but with nasty stuff only happening on a natural 1 and perhaps just minor annoyances on a regular failure.

Since I also use both a spell casting check for the spell to even work, and possibly a save for the target, I made spells have a reduced effect even if the target makes its save.

Example:

Sleep*​

Range: 60 feet, Target: 20-foot burst

Creatures within the target area become drowsy and sluggish, or fall into a deep slumber.

Each affected creature must save vs. spells. On a success, it becomes drowsy, suffering a penalty die on all ability rolls for 1 turn. On a failure, it falls asleep for 10 turns, becoming prone and dropping any held items.

Sleeping creatures are helpless, but awaken if hit, slapped, or shaken. Loud noises nearby, such as from a battle, allow sleeping creatures another save attempt at the end of the next turn. Awakened creatures are drowsy for 1 turn.

I use a combat system similar to BECMI, where you can get hit before your spell goes off. Though this just makes it harder to succeed the spell casting check instead of making you automatically fail as per BECMI.

One issue with potentially having no price to casting except on a 1 can result in the player just repeatedly rolling until success outside of combat, which is rather boring. In my mind, if there is unlimited spell casting, there should be a reasonable sense of drama to each roll whether in or out of combat. I'm also not a fan of permanent corruptions a la Dungeon Crawl Classics.

Losing the ability to cast that particular spell for the day solves it, but then you are back to tracking, which the goal was to avoid. Though only in the case of a failure - so perhaps still better than spell slots / spell points?