While I get the principle you're outlining, your specific example doesn't fly with me in that this player seems to want the benefits of both classes at once, and without the drawbacks of either. Sorry, bub, not gonna happen. Pick one.
Not at all. It starts and stops at "I want to write 'Priest' at the top of my character sheet and be a holy warrior that breaks face." These are, of course, not classes that exist in 5e or any WotC edition of D&D. ("Templar" was a subclass of Cleric in 4e, but that's not really relevant here.)
Then the player has a perhaps-unwelcome choice to make: play a Ranger and get the pet, or play a non-Ranger and eschew the pet. Or play a multiclass.
But that's exactly what I'm telling you that this player rejects.
This was something people LOUDLY lambasted 4e for making them choose.
This one's not an issue for me, in that the player here would be told: "Sure, go ahead and roll it up!". I don't (and never will) have the Drizz't-based Dex-first 2-weapon Ranger archetype in my game, and archery-spec'ed Dex-first Fighters have done very well in the past.
Okay. Can you ignore your personal table rules and consider it system-agnostically? Because that was kind of the point, and why I led with an example that never occurred in any actual D&D edition...
Here I'd say "You're flat out of luck. If you want to do any healing you're going to be a caster, end of story. A Fighter-Cleric or maybe Fighter-Druid is probably your best bet."
Again, I get that's
your way of how D&D would work. Please consider it from the context of there being people who want things out of D&D that aren't your preference.
Now that's a cool one.
I can't see why this couldn't be done. Sure it'd be an oddball Wizard who maybe had very few spells in its book, and it would be reliant on the DM/game to provide a steady stream of items for it to use (and then hope the rest of the party don't want those items too!); but I could certainly see a way to building something like this on the Wizard chassis.
No, you misunderstand. Not "I mostly use just regular magic items." They want to cast spells normally. They just want all of those spells to
be items, with mechanical support for them being actual items, and acting a little different compared to normal spells, but still also spells with all the rules of spells. And they want to be called a Wizard the whole time.
You saying, "Oh, okay, we'll give you magic items" would be just as much a problem, albeit not the cage of names. You would be taking away their ability to choose spells however they liked, and forcing them into dependence on whatever magic items the DM happened to give them. Whereas, if they played Artificer, they would have all this and more, even in 5e.
That's a houserule/kitbash circling around looking for a place to land, IMO. It wouldn't take much to strip casting away from the 5e Ranger and give it a few more combative abilities (e.g. heavy armour use, access to a few combat feats, etc.) in return.
Sure. But in official 5e, it doesn't exist.
Also rangers can wear all the heavy armor and take all the combat feats. And they get fighting styles. So...not really seeing how you're offering much that the Ranger doesn't already get.
Having the four basic classes as, well, basics is IMO a good thing both as an on-ramp for new players and as something for the casual player who's not interested in complexity. There's loads of sub-classes and alternatives if-when one wants to branch out.
Just because they're basics does not mean we must water them down until they are everything to everyone. They can still be clearly designed for purposes. You'd just want ones that are in high demand. And guess what! Fighter has always been cast primarily as a defensive wall between the dangerous world and their fragile teammates, going all the way back to OD&D. Other stuff they do has always been secondary--and in general is better served with actual mechanics that support doing those other things, rather than trying to scrape thin a single class that can do everything. The more archetypes you try to bundle into one single class, the thinner sbd more diffuse it becomes. That's why pure point-buy systems struggle with class fantasy; in a very real sense,
there isn't any. But cramming as many things as you can into a single class is very, very much pushing in that same direction.
And while I agree that subclasses are a very useful tool for flexing things in new directions, they're much more limited than you seem to think (which is, frex, why the "Warlord Fighter" has always been DOA, folks just keep burying their heads in the sand about it). You can sometimes make some pretty significant divergence (e.g. the 5e Warlock's Pacts), but the vast majority of cases are grace notes on the base structure, not meaningfully altering the class.
Believe me, I wish that weren't true. I have tried proposing such flexibility. For example, the Zealot Barbarian would be a shockingly close fit for an
actual 5e Avenger (the class from 4e), except that it would need to work with Dexterity rather than Strength for damage. Everyone I've ever proposed this to, even as a feat, has instantly and vehemently rejected it as totally unacceptable, broken garbage that only a powergamer could want. Just that tiny tweak of "let me use Dex for attacks and still get Rage stuff" is absolutely verboten.
This sounds like an argument for fewer classes and options, not more.
Not at all.
If we are talking about creating a new game, not the design of existing games, which I have not been doing this far. It is an argument for a Goldilocks zone. And four is
definitely too small.
It means that the things we design should actually be GOOD FOR the things we design them for. We don't need hypermegaultrageneric "Fighters" that have to be everything to everyone. We can have Fighters be some specific set of things, and actually be GOOD at them, and branch out
later if the player wants, using feats and (if I had my druthers) short, focused, well-made PrCs, something I know many are allergic to today because of how badly they were made in 3e.
The fact that
most people's preferences can be accommodated means we should make a sizable but moderate amount of classes. A tiny number will leave too many out in the cold. A massive number will be both "bloated" (a word I have come to despise because all it means is "I don't like this," it has no actual critical content) and, much more importantly, too
diffuse, making it harder to actually deliver on the design concepts of each class.
To give some examples, there's no way you could fit the diversity of Druid character options into merely being a subclass of Cleric, the Nature domain. You would axiomatically be eliminating relatively popular archetypes. Same goes with the ever so classic (and ever so frustrating) claim that Paladins are just Clerics with some extra stuff tacked on, or that Sorcerer could just be a subtype of Wizard. People enjoy the character fantasy of magic in the blood or the anointed knight (two archetypes which actually challenge "mundane but charismatic fighting-man" for the crown of "most common main character archetype"!), and the Wizard and Cleric would be garbage at delivering on that archetype. (Wiz especially, since it's already garbage at delivering on its
own archetype of the Hermetic researcher.)
Ultra-genericized classes weaken both the thematic expressiveness of those classes and their mechanical ability to actually
deliver those themes in enjoyable acts of gameplay. But having
too many classes makes everything a blur, for both players and designers. There's a happy medium somewhere in the middle, where there are enough classes that they cover
pretty much all the archetypes and flexing can get you essentially all the remainder, but few enough classes that each can be clean and distinct and pursue it's thematic identity with distinct and effective mechanics.
I think 4e had perhaps a small amount too many, and 5e has a noticeable amount too few. IIRC, 4e has something like 25 classes, and some really didn't need to exist (e.g. Seeker, Runepriest, and Assassin could—should—have just been variations of Ranger, Cleric, and Rogue respectively.) Meanwhile, 5e is missing at least the Warlord (as much as I know you hate it, it is one of the most commonly requested additions), Swordmage, Psion, and (debatably) Priest aka what 4e called "Invoker," the "clothie" miracle-worker type rather than the "armored cudgel-swinger" that every Cleric necessarily is. Since 5e offers no (official) support for class feature swaps, IMO a dire fault that hobbles the subclass system immensely, it's not really possible to do that. Best you can do is a Light Cleric who should still always be wearing medium armor.
If they actually supported the option of
replacing basic class features (not merely adding new options, actually
losing a feature you have and getting something else in its place), subclasses would be far more effective as tools for representing archetypes. But because WotC is
at least as allergic to the concept of ACFs as they are to the concept of PrCs, I doubt we'll ever see it, and 5e subclasses will remain almost exclusively grace notes.
