Hi, player, GM, and arm-chair designer here. I really appreciate you looking back at the older design choices with hindsight and a critical eye.

As a long-time DM this series of retrospectives has been really validating. I've had to work around basically every issue you've talked about, and I've always wondered if I was just playing the game 'wrong' in some fundamental way. I'm glad it's not just me.

What you are talking about is niche protection, and 5e flushed that down the toilet.

I mean... fwiw 4e successfully did that: wizard spells were more AoE and damaging and had more movement/control debuffs while cleric prayers we focused on supporting allies (hit, damage, AC/defenses, saving throws, and on) and their few - weaker - AoEs wouldn't hurt allies. They felt super different (even the implement cleric compared with the implement wizard). Heck, Sorceror and Wizard felt totally different and the old Fighter vs Ranger issue played differently as well. Nothing's perfect, but those designs did a lot right.

5e warlock invocations were so great, and also drove me crazy. That design space could have applied to each class so well (having customization option features like rogue tricks, natures secrets for druids, oaths for Paladins). Warlocks have so much more customization

I loved the design of the 5e warlock, it was so different than any of the other casters I had played before. Other people pushed me to make a pact of the blade warlock for damage dealing, but I loved the pact of the tome and the fact that I could cast any ritual spell if I found one with the Book of Ancient Secrets invocation. I decided my character was chosen by the Great Old Ones and all his spells were lovecraftian flavored, and I even made my own grimoire book of shadows to write down all my character traits and spells in. But here’s the kicker—I kept the notes as if I were a doctor talking about a mental patient, because my character actually WAS originally taking care of a pact warlock and was driven mad by what he heard. So now he is constantly seeking out the truth about this abominable force that he has learned too much about, but he doesn’t realize that he’s the patient he’s trying to cure. Loved playing him, loved reaching into others’ minds to speak to them because it would creep others out to know that there were unseen tentacles writhing around inside their heads every time I used the ability.

Wizardry completely consumed my life for about 4 days. Skipped school (college). Only left the room to eat or use the bathroom. Barely slept. It was amazing.

Armor and shields not having downsides to Spellcasting was also a bad idea. Its too easy to get heavy armor on full spellcasters and it means that somehow, fullcaster with heavy armor and of course, shield spell(which wasn't balanced around having heavy armor) is everywhere. 18 AC that doesn't require Dexterity and +5 AC for the round on tap is ridiculous unless there is some effort/compromise to get the combo.

I can see how with the wrong players either the Wizard player or the non-Wizard players would find it unfun, but to me a Wizard starting to cast a spell and then combat centering around the enemies trying to interrupt it and the party trying to protect the Wizard sounds like a good way to make combat more dynamic.

Good to hear discussion of the game from an actual design perspective rather than a discussion of session/campaign content.

Great to hear one of the principal designers reflecting on their work. Very helpful for everyone designing in the the 5e space. I've always thought that "batching" spells would be a worthwhile design direction: A player selects (or is given) not one spell, but a thematic category of spells. So you pick the School of Fire rather than "Fireball". These batches (called schools or spellbooks or another fitting name) would be a collection of damage, buffs, debuffs, utility and summons. For new players it would be much easier to say "I'm THE Shadowmage" than looking through endless spell lists. The batches could have some power spikes, like fireball, but having them offset by weaker spells. A subclass all about one thing could give the appropriate spell batch, like a Shadow Monk might be able to cast Shadow Spells using ki points. Narrow spellcasters like bards and sorcerers would have only one pick from a class list of batches and a second one from their subclass. Wizards could have the most batches active at once.

brainstorming my own game my main obsession has been strong character identity so like you said with invocations if you have classes the more mechanically distinct the classes are the better

I find these fascinating, and hope dearly you make more. Bringing on others to discuss might also be fun? Food for thought

3.X addressed the idea of spellcasting classes having different power levels of spells by differentiating a shared spell as being Wizard X / Cleric Y.   So while multiple classes may get access to a spell, it would come at a different character levels and required different resources.  That would also then favor certain builds.   For instance, clerics  made better horde necromancers than wizards because they got Animate Dead as a 3rd level spell while Wiz/Sorc had it as a 4th level spell. That means clerics could make skeletons and zombies at character level 5 and wizards had to wait until level 7 AND use a bigger resource for the same skeleton. (Sorcerers had to wait to 8, but thats because 3.x had sorcerers get more spell slots but gain higher tiers slower)

It was several years before I realized 5e doesn't have a magic system at all. There was some discussion online about bards and I wanted to double-check how 5e defines arcane and divine magic before commenting and realized...it doesn't.  (I was basically carrying my assumptions from 2e with me).  It drives me crazy now, how wishy-washy spellcasting is...the way each class has it's own bespoke spellcasting ability even though they are basically doing the same thing mechanically. I'm guessing the idea is that the setting would say how magic worked...but I've found players to be resistant to settings putting any kind of restrictions on their characters.

I'm not a fan of spell slots being used as a resource for limiting abilities either. From a game design perspective it makes sense not to have multiple resource counters...but from a narrative perspective it is confusing. In fact, much of the (lack of) magic system is a confusing mess, especially to new players....What is a spell slot? What does it mean to "prepare a spell"? Why can't I just cast any spell I know? What is the difference between a Cleric and a Warlock? Why do Sorcerers get fewer spells than Wizards? etc, etc. Questions I could have answered in pre-4th editions, but not now.

Been gaming since the 90s, and been in the design space since 2014. Listening to your videos and podcasts have been a helpful introspective of the game space from a designer's point of view. I too have those hindsight moments where you can't believe what you put out, and how you'd want to do things differently. I've only managed to split the atom once, and that one project I know I could never improve upon it, but I haven't been able to replicate it either. So hearing your perspective has been insightful.

There's a LOT to unpack regarding 5e design compared to older editions and mistakes that were made, especially with regards to spells. Fireball and lightning bolt shouldn't have leaned into the meme, and hypnotic pattern is just unnecessarily overpowered. I know you dislike forcecage, but that's a higher level spell. I can make just a slight exception for that spell only because of that. Hypnotic pattern has no excuse.
As someone who prefers 3.5, I liked the distinct differences in levels between class spells like how finger of death was Druid 7, Sor/Wiz 6; delay poison was bard/cleric/druid/paladin/sorcerer/wizard 2, ranger 1; and lesser restoration was cleric/druid/ranger 2, paladin 1. However, where making all spells the same level came in handy was cure light wounds used to be ranger 2, but was only 1st level for all the other classes. I always hated that. It finally being ranger 1 felt vindicating as a ranger fan.

I definitely understand where you're coming from regarding limiting spells to specific classes. That was one thing a player had to worry about in 3.X when playing a cleric as if you didn't take the right domain, you were either going to just a wallflower, a blaster, or combat fanatic. 5e isn't much different as one of the more popular domains for cleric is the Light Domain so you can access fireball, but Life Domain was very much the de facto healer and not much else. Now, with the Arcana Unleashed feats, you don't need to pigeonhole yourself to the Light Domain, even the Life Domain cleric can blast everyone to smithereens.
Also, I do like your idea of sorcerers attacking with magic, wizards being more formulaic, and warlocks having cantrips and invocations be the spells. In that regard, warlocks should have been the only class that saw cantrip increase in damage, not what we have now which is a fighter 19/wizard 1 can deal 4d10 damage with a firebolt cantrip. Again, that concept feels closer to 3.5, which did kind of negated the need of warlocks with their at-will (scaling) eldritch blast thanks to the Complete Mage's Reserve Feats which gave you the equivalent of 5e cantrips so long as you had certain schools of magic prepared.

Another "mistake" I would want to say 5e did was making everyone spontaneous casters. That was what made every class unique. Unless you had known spells, you had to prepare each spell a specific number of times. Only memorized fireball once? It's gone for the day. Clerics didn't have to prep cure wounds unless they were evil because they could turn any spell in to a cure wounds of that level. There's no limitations anymore. Also, armored casting should have cost a feat, minimum, not just something that a simple 1-level multiclass could fix instantly. Arcane spell failure was a great limitation, as was provoking attacks of opportunity is you were casting in melee range. Combat Casting was practically mandatory for most casters. Finally, caster levels shouldn't stack when multiclassing. In 3rd, if you wanted to do that, you had to take feats like Practiced Spellcaster to negate the caster levels you lost. Not only that, but a wizard 3/sorcerer 4/cleric 3/bard 4/druid 4 can have 9th level spells slots, but fighter 4/paladin 4/ranger 4/barbarian 4 can't get Extra Attack?
Stepping back from spells for a moment, melee Dex based finesse builds should have required a feat in order to get all the benefits it has nowadays. Before, you needed Weapon Finesse. In 5.5, that could have been an Origin Feat.

However, you're right. Time is a limiting factor and 5e had to be made. I've gone through all the old One D&D and D&D Next playtests, and there was a lot of good things that could have brought the game back to some kind of balanced normalcy, that was scrapped, which is unfortunate. The ranger and monk especially got done dirty, and arcane casters would have been just that little bit more balanced.

Also, this isn't a knock on you, but WotC as a whole. I really dislike that their way of "fixing" martials is to just have you cast spells. If the only way to fix the game is to make absolutely everyone a wizard, that's bad game design. Even worse now that counter spell and dispel magic is basically useless in 5.5 because not only were they nerfed, but caster NPCs have supernatural abilities that don't involve them casting spells.

You brought up the comparison to older editions, and something I was surprised to see omitted was the class specific experience progression of 2nd edition, and how that changed the way priest and wizard spells were balanced, side by side.

Characters in ad&d 2nd edition weren’t weighed against one another using level, they were balanced instead by total xp.

When you couple that idea with the lack of spell-slot fungibility while multi (or dual) classing, you hit a sweet spot of class parity that I personally feel was a peak for dungeons and dragons.

In short, the wizard just leveled up slower at early and later levels, and that makes as much of a difference as any other factor.

There’s an xp-slinky effect in the mid levels between 8 and 13, where wizards catch up past clerics, but priest spells didn’t go past 7th level.
And by the time you reach the top, wizards were accessing 9th level spells at the cost of 1,050,000 more xp than the cleric.

Meanwhile, the thief and bard are keeping a comfy 3-level lead on all of them, but that’s a different story.

Now that I write this all out, I’m less surprised that it was left out of the discussion, but I think it’s an interesting tool we’ve jettisoned, over the editions, that might still have a use case.

Thanks for the vid :)

11:06 and we're back to the the 3.5e Warlock huh?

Shout out to Richard Baker for nailing it on the first try.

Damn what a great channel I’ve stumbled on!

I just entered the world of TTRPGS & CRPGS two years ago. Game systems fascinate me. I immediately started analyzing how different games approach things and theorycrafting my own.

As a newcomer, here’s my take. RPG characters seem to come in two main design approaches:

Build-based, where you mix and match abilities. It’s great for player freedom, but it leads to minmaxxing and exploits - which hey, is also fun in its own way.

Story-based, where the “flavor” leads the design and classes are explicitly distinct. This is how you described ADnD and Overwatch. More limited, but limitation breeds creativity, plus fewer choices is way more approachable.

IMHO the latter is better for tabletop.

It seems to me that 5e is an awkward mix of both and it’s fighting with itself.

Perhaps if there’s ever a 6e, the approach could be a much wider array of classes, but with fewer options to overlap with other classes. Eg a Witch and Druid would both share nature-flavored spells, but otherwise have completely different mechanics.

Thank you for sharing your thoughts and experiences. And thank you for all the cool stuff in 5e that are not mistakes, you did an overall great job :)