So I've been playing 5th edition for a little while now, with a couple of different characters. I'm really enjoying it. It seems to make a very nice job of uniting things that were good about previous editions, improving game balance, and keeping everything flavoursome. Good job, WotC.
I just wanted to make a few observations based on my play so far. We've only hit levels 3 and 5 respectively in the campaigns, so it's early days yet. I don't claim particular expertise, and my notes will inevitably be coloured by my personal experience, as the stuff I've actually read in detail and thought hard about tends to be my own characters. I don't even own the DMG or Monster Manual.
For reference, those characters are:
- a 3rd-level human fey pact warlock ex-wheelwright who just got his sprite familiar (with cloth cap and tiny, tiny fey whippets), primarily distinguished by rolling really poorly on spell attack rolls and astonishingly well on fey charm rolls.
- a 5th-level elven ranger/draconic sorcerer/monk Gap Decade traveller who talks his way into bizarre situations and then is deeply bemused about why he suddenly has to fight his way out of them, using the motley collection of skills he's picked up between National Service, natural elven affiliation for magic, and other cultures' amazingly authentic and deeply spiritual practices that also involve flying kicks.
In general the experience has been extremely good; inevitably that means my comments here will tend towards the critical, because it's really hard to pin down why I enjoyed stuff, but easy to spot the things that jarred on me.
Spellcasters in 5e are great narratively, as well as their inevitable mechanical awesomeness. I think Wizards (drawing, admittedly, on other people's ideas, it seems) have done a great job of making the different caster classes feel seriously distinct from one another.
Wizards have a potentially immense array of spells available, and (due to a tweak about how spells are prepared, and ability to cast spells in a higher-level slot to greater effect) no longer have serious worries about preparing some of the niche but useful spells. You can memorise fireball once and cast it in all your mid-high level slots, giving you room for weirder stuff like protection from energy or major image.
Warlocks have a very limited focus, being primarily death-dealers, but the combination of pacts and invocations lets them take a range of interesting and flavourful powers. My warlock is a bit all-round weird, with at-will illusions and detect magic giving him an eldritch sort of touch. They have very few spells, but always cast them all at maximum level, and regain them every short rest. A warlock can use minor abilities all day, or have a few powerful abilities, or even enhance themselves with supernatural powers.
Sorcerers are seat-of-the-pants casters, with a small spell selection, but the ability to transform these in various ways on the fly. Their focus is on the sort of spells that you might manifest without training, essentially, although it's a bit hit-and-miss. Broadly speaking, they're more combaty than wizards, but less so than warlocks.
The sorcerer seems a bit cramped
Sorcerers are no longer specialists in a few spells that they can pump out all day. Warlocks are the untiring hurlers of magical death, with powerful cantrips and rapid spell recovery. Moreover, wizards now have the same number of spellcasting slots as a sorcerer gets, and the ability to regain spells once per day. Largely freed from the problems of spell preparation, this means a wizard can use more spells per day than a sorcerer, unless the sorcerer burns sorcery points to buy more spells.
However, that is a bit of a trap. Sorcery points are the main shtick of sorcerers, allowing them to partly compensate for a severely limited spell list through on-the-fly tweaks and enhancements to individual spells. Burning them to buy spell slots is expensive, and prevents the sorcerer from doing cool things with them. The wizard's ability is automatic, requiring no trade-off. The end result is that a sorcerer can sacrifice all of their other potential to cast a maximum of one more spell per day than a wizard can manage, while still having a far smaller range of spells to cast, and a vastly more limited spell selection in the first place.
Alternatively, the sorcerer can cast fewer spells than the wizard, but gain some mechanical advantages to each one. Don't get me wrong, some of these are nifty. Turning a touch spell into a ranged spell is nice, but there are very few touch spells for single-classed sorcerers - the unexplained lack of vampiric touch, for example, means you can't play a life-draining sorcerer. The main uses actually appear to be giving buffs to allies at range. Quickened spell allows you to cast two spells in a round, providing one is a cantrip, which means it's a useful slight boost to damage; it also allows you to escape combat and still cast a spell.
The other problems is, as with all pointy stuff, that these are limited. You can only pull metamagic tricks a handful of times per day; it's not as though you're transforming all of your spells as you go along, like the warlock does. A constant lament on forums is that the evoker is a better offensive mage than the sorcerer, because the Sculpt Spells ability allows them to hurl evocation spells into combat and always deal no damage to their allies, whereas a sorcerer can spend spell points a few times a day to deal only half damage to their allies - although this can be used for non-evocation spells.
Essentially, the problem here seems to be that sorcerers have been given one resource to compensate for at least two problems, and I'm not convinced the calculation took that into account. Sorcery points might compensate for fewer spell slots; used for metamagic, they might compensate somewhat for reduced flexibility; I don't think they compensate for both. This restriction is roughly in line with manoeuvre dice, inspiration dice and ki points, but it's worth noting that those seem designed to be relatively minor advantages.
Inspiration dice are the closest match, being a core feature of a spellcasting class. The bard has only a few inspiration dice, but they recharge on a short rest from 5th level. A bard also knows 50% more spells (though fewer cantrips), has a boost to most skill rolls, provides unlimited free bonus healing to the party, and can counter charm effects. A sorcerer has only spells and metamagic to rely on. If my shtick is warping spells, or doing combat stunts, I would really like to be doing that most of the time. A druid can be an animal most of the time if they want, a wizard or cleric will be casting an array of different spells, a rogue will be sneak attacking whenever they can. A sorcerer or battle master will get to use their flagship abilities 3-5 times a day for most of their lifespan. One of those is a core class ability, the other is a subclass ability. This seems odd.
I'm also somewhat underwhelmed by the way metamagic is learned. Since the primary shtick of sorcerers is now metamagic, it seems pretty tight-fisted that you only gain two metamagic abilities initially, working towards a maximum of four. Again, this is in line with manoeuvres. My general feeling is that a class focusing on these kinds of tricks should ideally get a fairly broad selection of them, with some level-gating that allows for more potent effects. This would make the class feel less limited, but also reduce repetitiveness. If metamagic is supposed to compensate for an incredibly restricted range of spells, an incredibly narrow set of metamagic capabilities seems like an odd response. "Sure, Albius Thunderwing may know a dozen 5th-level spells, but I can cast fireball, a really fast fireball, or a fireball that goes twice as far! How d'you like that, eh?"
Another concern with the sorcerer is the spell list. It's quite limited, essentially one-half of the wizard list with a couple of warlock spells thrown in. Any spell with a person's name in has been excluded, even though the only reason for the names is setting-specific background that originates (as far as I can tell) in the actual games of D&D employees, rather than any particular property of the spells themselves. Beyond this, there's no obvious reason why, for example, the earthy vampiric touch or flaming sphere are excluded, but crown of madness (which has very wizardy fluff) is available.
Beyond the actual number, I'm surprised by the number of niche spells listed. Metamagic doesn't fundamentally change anything about spells - you can't turn offense into utility, or even change elemental type. A sorcerer really needs to choose spells that are very broadly applicable, which typically means offensive and defensive spells. The limited spells known also places a high premium on scaling spells (giving more utility to the higher slots) and on those which can be usefully metamagicked. Bear in mind that each of a sorcerer's spells known is an investment of from 50% to (at higher levels) 10% of their knowledge. That is not a light decision. Let's look at 1st-level spells.
- Burning Hands - a decent emergency spell for a sorcerer who expects to end up near combat. It scales, although the small die coupled with a tiny range make it suboptimal.
- Charm Person - useful, though in combat it will tend not to work. It's problematic in social situations, though, because targets know they were charmed. It scales nicely by allowing more targets; however, higher-level targets have better saves, which makes advantage even worse. It's slightly better than friends, especially with scaling, but that's at-will. I'm undecided on this one.
- Chromatic Orb - exactly the sort of spell a sorcerer needs. It's very flexible, has good range and scales reasonably.
- Color Spray - a pretty poor spell in the first place, which can blind a creature, or a few very weak ones, for a single round. It scales reasonably well, but is seriously hampered by short range and short duration.
- Comprehend Languages - almost no campaigns emphasise language enough to make this worth a precious spell slot.
- Detect Magic - situationally useful, but is it really ever worth giving up a very large proportion of your spells (1/2 or 1/3 at the levels you're likely to be picking level 1 spells) to know if there's magic around? No scaling.
- Disguise Self - maybe useful in a very intrigue-heavy campaign, but it's pretty weak even then and doesn't scale. If you actually want this kind of capability, there's basically no reason not to skip this and take alter self next level.
- Expeditious Retreat - quite useful if you need to run a long way suddenly. Handy for escaping melée as it's a bonus cast; more useful if the rest of your party have similar abilities (e.g. monk or rogue).
- False Life - quite useful at lower levels. It does scale, though the ability to soak an extra 5hp per spell level is a relatively poor trade most of the time - casting a nastier attack spell is often a better way to survive. It's okay, and quite thematic.
- Feather Fall - amazing when you need it, but falling is rare enough that it's a poor investment. If you genuinely need this spell in your campaign, play an aarakocra.
- Fog Cloud - it's okay. It can seriously hamper archers and other ranged enemies, or create confusion while your party slips through. Scaling makes this potentially able to fog up an entire enemy force.
- Jump - why in the name of all that's holy is this dross competing with chromatic orb for space?
- Mage Armour - a solid choice for most spellcasters, although the draconic sorcerer can skip this one.
- Magic Missile - automatic damage is nice, and force is rarely resisted. It scales as well as any other low-level damage spell. Cantrips are nearly as deadly, but when you 100% need to hurt something this is invaluable.
- Ray of Sickness - reasonable, scalable and with an okay rider.
- Silent Image - a fun spell, when it's a trivial part of your abilities. My feypact warlock gets a lot of flavour use out of this as an at-will invocation. I can't see it being worth a major fraction of your casting.
- Sleep - a classic AoE spell, which also scales. Because 5e encourages using low-level enemies, this should remain useful.
- Thunderwave - I'm torn on this one. It's a crowd-clearer similar to burning hands, and its pushback can get a mage out of trouble or cram enemies into a hole. It also scales and deals an unusual damage type. On the other hand, most sorcerers should avoid being in melee range. It also makes one hell of a noise, which if you're playing "realistically" should cause all kinds of problems. The pushback is limited and doesn't scale, like most riders. Still, it's an okay choice.
- Witchbolt - this is a classic Sith lightning deal, which I really wanted to like, but can't. You can twin it and everything! But... the damage is barely better than your cantrips (especially as you level up), and you need to stay within 30' to use it, and it takes concentration. It leaves you within attack range, and enemies can easily escape it. There are just very few situations where this is your best option; lighting-vulnerable enemies, or paired with true strike against a really tough target that can't move fast.
That's 9 of 20 1st-level spells that are, at best, very questionable investments, and several more that are borderline. At higher levels, things improve a lot, though there are many odd omissions - why can a sorcerer animate objects, which feels quite wizardly, but not turn people to stone, which feels like primal magic? Why polymorph, but not true polymorph? The power word spells seem more like word-based wizard magic than innate sorcery. Delayed blast fireball seems like a wizard trick. What's wrong with wall of ice when all the other wall spells are available - other than force, which is (not unreasonably, I think) treated as "pure magic" and left to wizards?
I appreciate there's no particular reason why sorcerers should have spells unique to them, but most other classes do. It would be nice, but it's not essential.
Perhaps a bigger problem is elemental bias. There are two sorcerer foci, Draconic Bloodline and Chaos Magic. Draconic sorcerers are, I think, going to be the majority.
This is because, essentially, I think the chaos sorcerer is problematic. Its main shtick is rolling on the Wild Surge table for random side-effects, but the DM is left to choose when they can do this, with no guidance whatsoever from the authors. A kindly DM might allow you to constantly use Tides of Chaos and then roll on the Surge table, letting you have advantage on a large proportion of your rolls while also gaining (hopefully) benefits from chaos magic. Another might only occasionally let you roll a d20 for a 5% chance of getting a surge, and never allow you to regain Tides of Chaos between rests. There's no discussion of what effect this will have on play and balance. The unpredictability of the table also adds some problems, as a chaos sorcerer can wipe the party with an unlucky Surge roll. The 4e chaos sorcerer seemed like great fun, unpredictable yet largely under the player's control; I'm not impressed with this version.
So, dragon sorcerers. You gain a small roleplaying bonus when interacting with dragons, and ability to use Draconic. You also get extra HP and, most importantly, a bonus to AC for being scaly. Awesome. At 6th level comes Elemental Affinity. This allows you to add your Charisma modifier to spells dealing elemental damage of the type aligned with your ancestry, and also spend 1 sorcery point to gain temporary resistance to that damage type. This is particularly useful for powering up your cantrips, which are normally dice only.
Unfortunately, whoever was writing the rules forgot to consider how this would interact with the spell list. The sorcerer list is crammed full of fire spells, and at a loss for almost everything else. Poison and acid come off particularly badly - there's exactly one poison spell (cloudkill), no acid spells that I can find, and chromatic orb. This means you'll see far less benefit from this ability compared to a fire sorcerer, who has a fire spell available at almost every level, most of which can also be scaled to fill gaps, many of them powerful AoEs, and one of which (scorching ray) allows you to hammer a single target with this benefit multiple times. It's just theoretically possible that this was judged on the basis of all those monsters with fire resistance? Hmm.
The Elemental Evil Player's Handbook offers up a stack of elemental spells, many of them very nice. This could have been an excellent opportunity to address this. This provides, in total: one acid spell, one lightning spell, three cold spells, and another six fire spells. Poison, of course, not being elemental, gets nothing at all. There are several nice air, water and earth spells, mind.
There used to be, somewhere, an ability that let you change the elemental type of a spell. An obvious perk for a draconic sorcerer would have been an ability (perhaps limited) to transform spells to their own elemental type to address this imbalance. As it is, choosing to be an acid or poison bloodline is putting yourself at a massive disadvantage, while choosing fire is a significant advantage. It doesn't help that fire also has most of the multi-target spells, making extra damage per hit particularly valuable. Scorching ray exaggerates this even further by allowing multiple hits on a single target, potentially dealing massive damage.
Because of its limited selection of spells known, the sorcerer is unusually dependent on scaling spells. The warlock has similar issues; although all warlock spells are cast at the highest possible level, taking a non-scaling spell means you're essentially cheating yourself out of additional power. You can swap it out, of course. In contrast, the sorcerer aims to scale up certain spells to higher-level slots to fill gaps in its selection. No 5th-level AoE? No problem, just scale up that fireball.
One thing I noticed as I was looking through the lists is that this... doesn't really work.
The thing is, higher-level spells are just flat-out better. It's not simply a case of damage - damage tends to scale reasonably if you cast spells in a higher-level slot. Higher-level spells affect more targets, last longer, and have more potent riders. Let's compare a bit.
Burning hands is a 1st-level evocation spell that deals 3d6 fire damage in a 15-foot cone (Dex for half). Cast in a higher-level slot, it deals 1d6 damage per extra level.
Cone of cold is a 5th-level evocation spell that also deals damage in a cone. This deals 8d8 damage in a 60-foot cone (Con for half), and scales 1d8 per level.
So in a 5th level slot, we can cast a 15-foot cone dealing 7d6 damage, or a 60-foot cone dealing 8d8 damage. With a 7th level slot, we could cast a 15-foot cone dealing 9d6 damage, or a 60-foot cone dealing 10d8 damage. This is a very substantial difference, and much of that difference doesn't come from the damage. For a squishy mage, 60' range is a massive advantage over 15' because it keeps you out of combat, while the cone of cold can also attack far more creatures, or targets that are more spread out. You can blast creatures out of the sky, or cast from a rooftop into the street, or aim through a window and take out a roomful of enemies.
Similarly, charm person is a 1st-level spell that makes a creature consider you a friend for 1 hour, gaining additional targets if cast in a higher-level slot. Dominate person is a 5th-level spell that lets you telepathically order another creature around, or even take mental control of its actions, for a short time, gaining duration if cast in a higher-level slot. Unlike charm, the creature doesn't (by RAW) remember being dominated, and you can force it to act entirely against its instincts and wishes - commit murder or even throw itself into a volcano, if you choose. Charm has its advantages, both in duration and target number at 5th-level; the point is that you can't simply scale up one and compensate for lack of the other, even with two such similar spells.
Unlike previous editions, 5e doesn't have spell slot affect the DC of spell saves. This is a sensible move to keep bounded accuracy, as it means high-level wizards can't produce nigh-impossible DCs (although presumably you could have adjusted the way it works to avoid this). One perhaps unexpected result is that it makes non-scaling spells even more problematic for a sorcerer. If a spell has a fixed effect, such as disguise self, then there's no point casting it in a higher slot, which makes these spells a worse investment for casters with limited selection. Under the older systems, you could at least use higher-level slots to bump up DCs, making your spells more likely to work, and clawing back some of that utility.
Both of these issues will tend to have significantly more impact on casters with limited spell selection, which primarily means the warlock and the sorcerer. Because the warlock auto-scales spells, they really care about this stuff. Half-casters have a similarly limited selection, but rely proportionally less on their spells. A sorcerer can slightly improve matters with one of the two spell-selection feats, Ritual Caster or Magic Initiate, although unless they're the only caster in a party it's probably not worth the loss of stat advancement - unless they really want that familiar, of course.
So overall, the sorcerer has:
- A limited selection of spells, which seems erratic and has many poor choices of very niche application. The only thing I can think is that perhaps these are intended to ensure sorcerers can use the relevant scrolls?
- Fewer daily spells than a wizard, and potentially fewer than a warlock, depending on short rests.
- A resource pool that seems designed to compensate for both of these, but mechanically only seems hefty enough to compensate for one.
- A heavy reliance on scaling up to compensate for gaps in its known spells, which doesn't actually work very well.
- A unique shtick, metamagic, which should make sorcerers stand out and compensate for limited spell knowledge, but which is so restricted in both range and usage that it doesn't appear to do so.
- A subclass with an elemental focus that interacts extremely poorly with the spell list, apparently just due to poor planning.
- Another subclass whose effectiveness depends entirely on GM whim, and which may be superlative or useless accordingly.
I have to say that so far I'm not particularly impressed. The sorcerer is a very cool concept, and it may be that in actual play a full-classed sorcerer somehow works out on equal footing with other casters. I don't really see how, though, and this wouldn't do anything for the problems with its subclasses. Don't get me wrong, I think a sorcerer will do okay, but I suspect it will begin to feel lacklustre because of its limitations.
Armour Class synergies
I'd been playing Charlie for four levels before we worked out that, due to legacy thinking, I've been miscalculating his Armour Class this whole time.
See, Charlie is a draconic sorcerer with monk levels. Both of these improve your armour class. Monks get to add their Wisdom to AC, while draconic scaliness improves your basic AC.
Unfortunately, in 5e, for very sensible balance reasons, they've really clamped down on AC-boosting. For example, most armour now sets your AC rather than adding to it. The end result is that wherever you have multiple ways to calculate AC, you have to pick one and stick with it. Draconic sorcerers get to use 13 + Dex (as though they had permanent Mage Armour, or lightish armour), while monks get 10 + Dex + Wis. I'd been thinking with previous editions, so my quick glance over the monk rules didn't pick up that the monk Unarmoured Defence ability now resets your AC and competes with the sorcerer ability. You don't get 13 + Dex + Wis, as I'd been assuming.
The result was that rather than a hefty AC 17 (which makes for a decent frontline skirmisher despite suboptimal HP) Charlie actually had AC 15, which is much worse, considering that AC caps out around 20. Having spotted this, we agreed on some hasty rebuilding and got him up to a 16. It's not a major loss, but somewhat irritating.
I absolutely understand the balance reasons for this. A draconic sorcerer/monk or sorcerer/barbarian would otherwise be able to gain enormous AC through stat-boosting, since you could easily start with an 18 and reach ridiculous levels quickly. The disadvantage, though, is that it makes no sense whatsoever in the narrative. A monk can avoid danger by sensing and avoiding them. A draconic sorcerer is hard to hurt because of being scaly. These two things do not interfere with each other in any logical way, and so a monk who is also scaly should absolutely, as far as narrative is concerned, be even harder to hurt. This sort of thing upsets me.
This is the first edition where I genuinely like the paladin class.
The 5e scheme explicitly establishes the paladin as a courtly Arthurian archetype, with a lesser emphasis on their divine powers and a much reduced commitment to legalistic goodness. Although they're all supposed to quest against evil, the book presents a range of suggestions from shining cosmic champion, through heir to a questing tradition, to essentially Batman. This is all good stuff. Paladin powers now come as much from determination and willpower as from any god, and despite the fluff, there's no particular reason a paladin's god must be a relentlessly good one. Most gods consider their own ends good, after all. You can, if you choose, play a paladin who cares nothing for gods. They've also hacked back the paladin obligations from mechanical burdens to roleplaying principles, with different principles for each subclass.
Another nice change is to the evil detection. It was always frankly weird that paladins could innately sense evil; this put paid to a lot of scenario options unless you deliberately nerfed the paladin's abilities by taking specific magic items or spells that negated them, which makes the whole exercise pointless. It also caused weirdness because of the attitude to evil - a miserly shopkeeper or selfish wizard might read as Neutral Evil, while a deeply misguided king might read as Good despite an array of horrific actions. After all, torturing all those traitors to death was the right thing to do. The improved version simply reveals the location of nearby angels, fiends or undead, and of holy or unholy places. It has a strictly limited use, which is also much better. For my part, a paladin wrinkling their nose at the palpable stench of evil incarnate, or sensing great goodness and bowing the knee, is very much appropriate to the genre - mentally judging every person they set eyes on was not.
They've also shaken things up with the three Oath subclasses, allowing for very different foci. The Devotion paladin is the classic do-gooding monster hunter, with lots of protective abilities and light-based powers. Its Channel Divinity provides a potent magical weapon, or turns fiends and undead. It's also the simple option - all classes now include one quite simple option, with minimal resource management, which I do approve of. You can probably use this in most campaigns, although an all-evil campaign would probably struggle.
The Ancient paladin is essentially a fey knight, a force of nature that's hard to kill has lots of druidic abilities. It's commited to life and beauty, which offers the option of a kind of "drink and be merry, for tomorrow we shall die" paladin who revels in all the good things of life as much as in battle against evil. I really like this one, although the Channel Divinity options seem the most lacklustre of the three. Nature's Wrath seems very weak: restrain one target within very short range, which gets to choose either Strength or Dexterity to save on each round, one of which is likely to be good. Turning fey and fiends is nice, but honestly fey tend to be pretty rare as enemies, while this kind of paladin may well have fey allies. I confess, however, I'm not entirely sure how it's supposed to fit into a campaign. I feel it would work well in games focused around hunting down incarnate evils in a mostly wilderness setting, protecting elven forests from the incursion of nature-scorning settlers, and so on. It would make little sense in games that are primarily urban, or those whose sides aren't so clearly demarcated as the Good and Natural vs. the Evil and Unnatural.
The Vengeance paladin is essentially the Avenger from 4e, or Batman. Conan could actually be skinned as a Vengeance Paladin. It has powers to help it hamper and damage its prey, track them down and chase after them effectively. It's a nice relentless bringer of justice. You can fit this into most kinds of game, even those where most characters have grey morality. You might even manage it in a game of villains, playing a paladin who's either willing to deal with the devil to achieve their quest, or who's focused on their own brand of "justice" but has no overall moral high ground.
I think the only change I'd have wanted for paladins is for skill lists to better reflect the subclasses. As it is, despite the fluff, you can't play a hunter-paladin trained in Stealth, a fey knight with any Nature at all, or a courtly knight with any knowledge of History.
Bounded accuracy is awesome
I'm loving the way the bounded accuracy, and other system tweaks, are making things play out. It really does feel as though the curve has been flattened, to the point where a lucky adventurer can (as in all the stories) actually defeat a very powerful enemy with a bit of cunning and skill.
At one point we ended up fighting a glabrezu and several de-levelled cambions, and despite being only 4th level and in no sense optimised, we bravely fought them off for several rounds, killing two cambions and badly wounding the glabrezu, before reinforcements arrived. Of course, this was partly due to the DM hopping the glabrezu from target to target to spread the damage, rather than going in as hard as possible, but that's the appropriate way to run an overpowered battle in my view. Let me point out in his defence that we were never actually intended to do much of the fighting ourselves, but our angelic ally rolled a string of really terrible saving throws and got himself stunlocked.
What I'm saying is, as a 4th level adventurer I was skulking through the planes of Hell and fighting off powerful demons in an effort to recover a war machine as part of a reluctant deal with a pit fiend, and the mechanics made it feel not completely inappropriate. It doesn't feel as much as though level 1 adventurers are lowly worms and level 10 adventurers burn cities in their wake.
This also addresses one of the issues I always had with the 3rd-4th edition settings: the pyramid. Given how hard it is to reach high levels, there ought to be a huge number of mid-tier adventurers around, each of which is capable of taking out a substantial town single-handed. Why do you never meet them? Why don't you regularly run into other adventuring bands? If they're the town mayors and so on, which is one solution you sometimes see (primarily to stop PCs overrunning towns, one suspects) why don't the level 10 fighter-mayors go and clear out that band of lowly goblins and stop pestering you about it? The world should have a hefty population of other heroes at all levels, and having similar interests and goals, you should interact with them sometimes, and at low levels you should frequently be overshadowed by them. With bounded accuracy, it feels less of an issue because the distinction between you all is much less, so you're not constantly in the shadow of superheroes. The smaller mechanical discrepancy means narratively you don't necessarily notice a huge difference in capability between a low-level warrior and a higher-level one - although it's always much more noticable with casters, because they gain specific identifiable abilities in the form of high-level spells.
Similarly, town guards have tended to be caught between a rock and a hard place - if they're low-level then they present no obstacle either to mid-level adventurers nor to the majority of threats worth employing guards against, but if they're higher level then why are they scraping a living as a town guard instead of running a town? Again, bounded accuracy means there's much less distinction in capability here, so one low-level guard with suitable equipment is probably capable of dealing with a minor crime, and a patrol of guards are a genuine threat to rowdy adventurers, mad wizards or wandering monsters. They aren't going to take out your average dragon, but then literature doesn't expect them to - that's where heroes come in. If it's just an ogre or a griffin, though, it's perfectly possible.
Multiclassing has been a thing since at least 2nd Edition, and 3rd edition turned it into a core feature with matching XP tracks. I've always liked the flexibility that it offers, as well as the fact that it undermines a tendency to consider classes (a mechanical abstraction) as some kind of universal truth. Nobody is a Fighter. 4th edition tried to cut back on the splat-inspired carnage by turning multiclassing into a minor affair handled by a couple of feats. That seemed frankly terrible, because many of the feats didn't remotely emulate the class they were supposed to represent - adding +2 to your damage once per encounter is in no way similar to being a freakin' sorcerer.
Fifth edition feels like multiclassing done right, frankly - it retains the flexibility of 3rd edition, but implements some canny rules on multiclass benefits that shut down most of the excesses.
I remain a fan of multiclassing. It's a great way to produce some specific character concepts, or simply add flavour to a class. The warrior who knows just a couple of spells, or the wizard who can handle a sword, are both classic tropes. A rogue might take fighter levels to learn some combat manoeuvres for that tricksy slippery warrior look. A barbarian can take rogue levels to be a raging, dirty fighter who hits hard and low, or ranger levels to be a hard-as-nails hunter, perhaps defending their tribe from ancestral enemies. A wizard or sorcerer might become a servant of one deity or other through their feats of magic and gain cleric levels. And so on and so forth. Another great use is to access particular spells that fit a character concept.
Multiclassing for spell range
In a few cases, multiclassing seems almost mandatory. This is primarily the case for the two spell-hampered classes, the warlock and sorcerer.
A warlock can expend one of a precious few invocations to gain the ability to cast bane or bestow curse once per day while also burning a spell slot. These curse spells are extremely fitting for the class flavour (which is, presumably, why these invocations exist); they're also highly useful, and apparently considered so powerful with auto-scaling spells that they need to be severely limited. So they don't appear on the spell list at all; they use a separate special mechanic, which limits their use to once per day, and also uses up one of your very few spell slots. Or they can take the Magic Initiate feat and gain the ability to cast bane only once per day, at 1st level (so no scaling for additional targets) plus two free cantrips.
Alternatively, they can multiclass into bard or cleric and gain several spells from the chosen list, including bane, plus some very nice cantrips that greatly expand the type of things the caster is capable of. If they're committed to multiclassing, which is my preferred approach, they can soon get hold of bestow curse as well. These spells can be cast multiple times per day, using the scaling warlock spell slots if desired.
This isn't about powergaming, mind. What these spells really offer is what you might call a halfway-viable alternative to hex, which is otherwise both a must-have for most warlocks and a block that essentially makes it pointless to take any of the many other Concentration spells on the warlock spell list. They don't boost damage in the same way hex does, but they do provide a reasonable alternative way to hamper enemies while laying down damage with your cantrips.
Multiclassing and Ability Score bumps
I'm significantly less a fan of the new rules for Ability Score Improvements (or feats), which now come off class level rather than character level. I assume this is aimed at shutting down dipping, so that taking one or two levels in a couple of classes seriously hampers your advancement, and it seems like a fairly simple and effective solution. As is traditional, the fighter receives bonus ASIs at various levels, allowing them to both boost their stats and pick up a range of other useful capabilities.
However, it does mean if you're multiclassing in earnest, you run into problems. Charlie being a triple-classer, he doesn't actually qualify for his first ASI until 10th level, at which point he gets three in succession, leading to some wild swinginess in relative effectiveness to the rest of the party. As it happens, my DM ruled that in our Planescape campaign, as a member of the Sensates (whose whole shtick is being open to new experiences and dabbling around) his Faction benefit would be basing ASIs on character level. Phew.
Since I used that 4th level to take the Actor feat, he may now be regretting that decision. Actor is possibly the most entertaining feat ever.
Things whose voice Charlie can currently imitate almost perfectly:
- A beholder (the barkeeper of an inn we visited)
- A major politician of Sigil
- The factol of the Sensates
- A pit fiend
- A planetar
- A giant adamantine wolf golem
- The leader of a huge brainwashed cult living inside an illithid doomsday weapon and worshipping the Mighty Beam.
- Any common animal
- Every other party member
For even more hijinks, it turns out our wizard took the opportunity to take the 'perfect memory' feat at the same time. He can describe exactly what happened and Charlie can act it out impeccably.
Luckily, Charlie is so bad at Deception that he hasn't yet ruined the game.
So, some qualms about specific aspects, but on the whole I remain remarkably positive about 5e.