[RPG] As a fighter human is it possible to have a pixie as a companion

'Balance' covers more than just 'balance', if you see what I mean

It's a good thing to ask whether your creation balanced, as in whether it is more or less powerful than existing options.

But there's also the question of balance as in 'is a good addition?', 'is it out of place?' 'is it useful?', 'could it even have been published in 5e in the first place?' and wider questions. I'll address the latter first.

Why are you adding it?

If the answer is as simple as "because it's fun to try and make stuff", fair enough, by all means homebrew it, and if you're with a bunch of mates rolling around casually, then often no-one's going to mind. Homebrewing new content can be enormous fun for the brewer. I make new classes all the time.

However, for a more serious campaign where balance is being considered carefully, as it sounds you are, then a lot of fellow players [or your own players, if you're DM], might want more justification than that. And to do that, you need to answer 'yes' to the following questions:

• Does it let my play a character (as in mechanically) I otherwise couldn't?
• Does it let me play a character (as in person) I otherwise couldn't?
• DM: Does it add to the story of, or the story happening in, the world?

I've found most players and DMs are as concerned with 'why' than the nuances of power balance. The more detailed and difficult questions you need to think about for the latter are as follows:

(Please note I discuss these aspects of balance checking generally, using hypothetical examples and those from my own homebrew, before discussing the Alchemist specifically each time.)

"Does my class step on another class's (story) toes?"

This depends on how specialist you want classes to be. Are you aware of the 3.5 'Basic' variant where there were 3 classes? (Warrior, Expert, and Spell caster?). The 5E basic rules divide spellcaster into Arcane and Divine. The 5E full rules split arcane into two, divine in two, and warrior into three (fighter, barbarian and monk) before adding magic.

In my high-magic, high-tech, gods-at-war setting, I have added two new divine classes, a Favoured (Soul) and an Invoker (or 'Siphon', TBC). These are just making new specialisms of 'Divine Caster'. The Favoured is a servant of a deity, just as a Cleric is. Oh! I'm stepping on story toes. Or am I?

In my setting, Clerics are trained, organised servants of the pantheon, whereas Favoured characters are untrained, devoted to one master above all. So am I interfering with the Cleric's role? Certainly some PCs and NPCs who would have been Clerics might now be Favoureds. I'm stealing some of the Cleric's player base, perhaps. Is that a problem?

If your class is stealing a player-base (even just for story purposes), it is something to be aware of and that many could be uncomfortable with. But then, introducing Bard, Sorceror and Warlock into the basic rules is stealing the Wizard's players. The Nature Cleric and Ranger steal Druid players. Where do you draw the line?

I would draw the line here:

• The class does not replace an existing one, even if it takes some players (unless, for example, gods and Clerics don't exist in your world, so you make a Runepriest as an arcane gish/buffer/healer)
• The class does add a mechanical way of playing the alternative character type (this could be a mechanical combination, rather than a unique feature. See the example of my Adept class below.)

Does your Alchemist step on another class's (story) toes?

Are you stepping on any story toes with an Alchemist? I would say no. Casting a spell (examples: fireball, teleport) is not the same as alchemy (which goes wider than just potions - we could be talking self-augmentations, poisons, skin grafts, who knows?).

When people say "Just play a Wizard and call it alchemy"...

Opening asker, if you think, like me, that a character should be able to say "I'm a Dwarf Alchemist, behold my concoction!" instead of "I'm a Dwarf Wizard and this Fireball is actually me throwing a bomb really far", then your Alchemist isn't stepping on story toes.

"Does my class step on another class's (mechanical) toes?"

Ah, now we have to express our class's unique place in our story with a unique place in the mechanics. Probably: but on the condition you're not performing another class's mechanical role better, you may not have to be that unique to still have an acceptable class.

Let's take a setting which is relatively peaceful... why the heck should all clergy automatically have armour and weapons? Let's make a 'Priest' or 'Adept' class. Take the Cleric, strip out all armour and weapon abilities (including in the domains, maybe adding Divine Strike to cantrips), and give them Divine Recovery.

Yes, a lot of the time we're playing one of a Cleric's mechanical roles (buffer/healer), but then so do the Bard and Druid. Do we have a mechanical difference which reflects their story role?

Giving them Divine Recovery means their key mechanical difference is ingrained in the story difference. A Cleric is great at being in battle. Our Priest/Adept needs rest - needs peace and quiet. As the whole point is that we're making a peace-time, no-trained-for-battle sort of divine agent, the mechanic fits the story perfectly, even if it's not that unique.

Even when stepping on toes a bit, if your mechanic is at least a bit different, and mechanically expresses a story difference, then it's probably alright.

"Does your Alchemist step on another class's (mechanical) toes?"

Hmm. Are we stepping on the Wizard's mechanical toes? It depends on what mechanic you use.

As you're using spell slots, you need to ask yourself how that mechanic is actually reflecting a story difference between alchemy and wizardry. Personally my thought would be to only duplicate low level spells (2nd attacks, 3rd buffs, or something), but have it almost at will. Why, if I can make and carry one fireball-potion today, can't I make one a day for a week and then go nova with 7 when the fight comes? If I'm going adventuring, why can't I have a backpack full or kit?

Perhaps the alchemical components spoil, AND require the brewer's own blood, or something along those lines. I'm toying with an alchemist myself, and one possible idea is to make the proficiency or constitution bonus worth of potions etc. on a short rest, which last 8 hours before spoiling. (This is in addition to duplicating cantrips as long as they have a standard toolkit on them).

Is it too powerful? Is it too weak, or partially too weak?

Honestly that all depends on what you do with duplicating spells. If you're duplicating up to Fireball (which would be really nice), maybe you should only be able to get that at-will at level 18-20, because that's when a Wizard gets level 3 spells at-will. However, level 1 and 2 spells, the Warlock gets at earlier levels.

Your cantrip system up to 4d6 is less damage than other classes. Even if he can whip out literally any elemental type, bear in mind other get 4-6 cantrips so could have most elemental types available whenever they want. Just make it elegant. (going from 1d4 to multiple d6s is a bit odd: keep the 4 or 6 consistent)

It is hard to give you feedback without seeing more - but then it's probably hard for you to make more without getting feedback, I know.

Note, I've compared mostly to the Wizard, but if you brew stuff on a short rest or something, you will need to compare to the Warlock or Cleric Domains too.

As an exercise to try out balance, put your class in a hypothetical story situation and compare her to how a Wizard/Warlock/Trickery Cleric or whoever would do.

• Level 5: A mob of bandits descend on your caravan. Defend yourself! (Compare: Wizard's Fireball, Cleric's Haste on the caravan guards). How do your Alchemist's class features compare in a straight up mob fight?
• Level 10: The enemy are executing a flank attack. A company of hobgoblins will be crossing the long, fragile stone bridge leading to the village in one hour's time. (Compare: Illusionist making the bridge look out, Druid summomning creatures to re-flank their column) Maybe your Alchemist can cast Fireball but as a ritual, and take out the bridge or something. How does your class handle a large choke-point defence with some time to prepare?
• Level 15: The cult of the stonelord have taken up position in an old dwarf fortress. We can either assault the battlements or try and infiltrate an abandoned mineshaft under the water supply. (Compare: Wizard puts greater invisibility on a strike team, Bard makes a mass suggestion for one sub-faction to fight another, Sorceror sends Careful Spelled cloudkill, incendiary cloud, chain lightning, etc., as mob clearance) Can your class take on one of these roles and do it as well? Could they do several of these but not as efficiently each? Can they add something new? (Poison the water supply? Transform the overhanging cliff into a tsunami of acid?)

Maybe you can't come up with a specific plan, but thinking over those scenarios should help you decide what your class should and shouldn't be able to do, so you can think "Actually, if I give her delayed blast fireball, that just makes me a Wizard with Con instead of Wis, so maybe I'll make all her spectacular effects require preparation and slow application"

I hope that gives you some ideas to think about for balancing homebrew classes generally and some specific feedback on your Alchemist ideas.

Prioritize number of attacks and Feat combinations

Ok, a quick addendum before we go any further. There is no such thing as one perfect build for every situation. The environment you'll be fighting in, the type and number of opponents, and even the nature of the story you're engaged in can all change which characters are "best" suited. For example, a character who hits once but hits extremely hard (like a rogue) may have higher DPR (damage per round) than another, but against a horde of weak enemies (e.g. Goblins) they may find themselves less useful than their DPR would suggest they would be, since a large amount of their damage each round is wasted hurting an already dead enemy.

All that being said, I've done a bit of research and found that at this particular level, there is a build that will work very well in a lot of scenarios, and meets the criteria you've set (survivability and DPR without any other players' help) in a measurable way. So enough hedging: let's look at the details.

The Build

Stout Halfling (counterintuitive, but hear me out), Eldritch Knight Fighter 13, Rogue (or Wizard) 2

Str: 8 Dex: 20 Con: 16 Int: 14 Wis: 12 Cha: 8 AC: 18, 20 with Haste, up to 25 with cover

Feats: Sharpshooter, Crossbow Expert

Favored weapon: Hand crossbow

Favored Spells (in order of importance): Haste, Mage Armor, Shield, Find Familiar (or another 1st level Wizard spell), 7 more evocation or abjuration wizard spells of your choice (you'll rarely be casting them).

Rationale

In descending order of importance, the three factors you need for maximum ranged weapon damage dealing are:

• Multiple attacks
• Accuracy & Damage Bonus (about equal: both come from Dex)
• Base Damage (weapon's dice)

Multiple attacks are almost always the most important feature. A typical enemy in DnD will have an AC which can be hit on a natural 8 or better. So usually, an increase an increase in accuracy (say, the 2 point increase from the fighter's "archery" style) will grant you about a 16.666% overall boost in expected damage (hitting on 14/20 attacks rather than 12/20), while a boost in number of attacks will grant you at least a 20% boost in damage (going from 5 attacks to 6, which usually can only happen at 20th level anyway). Although enemies with particularly high AC may alter this math, on average multiple attacks will be of more benefit. That being said, we're going to get both as high as possible.

With all that in mind, you're going to want the following build:

• 13 levels in Eldritch Knight: 11 for 3 standard attacks. At level 13, trade in the spell you gained at 8th level for "Haste", and get another single weapon attack (4 per round total so far).
• At level 4, take crossbow expert feat. Now, according to sage advice (http://dnd.wizards.com/articles/features/sageadvice_feats), you will be able to use your bonus action for another hand crossbow attack. Although the hand crossbow has the lowest damage dice of the martial ranged weapons, this feat makes it the most damaging in most circumstances. (5 attacks per round by level 15).
• At level 6, take sharpshooter. Now, you can shoot your hand crossbow just as well up to 120 feet (and since most combat in a dungeon crawl happens within 100 feet, you're golden). Also, as long as your target's AC is less than 22, you always want to take the -5 to roll for +10 to damage (the average damage per round will be higher).

Your remaining two attribute increases should got to getting a 20 dexterity. So far we're up to: 5 attacks per round, expending nothing other than a 3rd level spell slot (2 per day available) and ammunition. Each attack does 1d6+15 damage, or 18.5 average damage per successful attack. That's (averaging out the rolls) a whopping 92.5 damage per round, without even using action surge! Even more amazingly, you could still be dealing an average 37 damage on a turn in which you are DODGING (haste action + bonus action attacks)! Even without haste, we still could be doing an average of 74 damage per round (assuming all attacks hit) with the hand crossbow, expending no resources other than bolts.

For the final two levels, there are a couple of options; but probably the best one is to take two levels of Rogue. You'll get an extra 1d6 per round from sneak attack, and cunning action (which, although it would lose one weapon attack, would allow you to disengage, dash, or hide as a bonus action: it'll give you considerable versatility on the battle field, and allow you to easily keep your distance). Also, you could get expertise in stealth and acrobatics, which will be a huge boon if you're hoping to scout or hide in combat, or avoid being grappled. Our damage total (without action surge) is now up to 96 damage per round. If you wanted to Haste more often rather than do a bit more damage (and losing some battlefield mobility), you could take 2 levels in Wizard instead, allowing you 3 castings of haste automatically, and an Arcane Tradition (War Magic could fit nicely here, or Bladesinger if your DM allows it for a non-elven character. Warning: a DM may consider an attack with a hand crossbow to be using "two hands to attack with a weapon", which ends the bladesong. Check with them before choosing a subclass).

The Halfling's bonus to dexterity is essential, but you could get this from other races. A real benefit is the +1 bonus to constitution. It may not look like much, but this build is very much about maintaining the Haste, and has no attribute boosts to spare. If you take this build, DC 10 concentration saves (which is what most attacks will incur if they hit you; only attacks dealing 22 or more damage will up this difficulty) will only fail on a natural 1: and because of Halfling's "Lucky" trait, this means the standard concentration save will only fail 1 in 400 times. Also, the Halfling ability to move through the space of medium hostile creatures makes this build insanely mobile. If you went with 2 levels of Rogue, even if your character was completely surrounded by (medium sized) guards, you could Disengage (cunning action), Dash 45 feet (Haste dash action, speed doubled, five feet of difficult terrain for moving through an enemy's space), make three attacks (action), then move up to 50 feet more, all without incurring any opportunity attacks.

For your non-abjuration/evocation first level spell gained at level 3, I'd suggest Find Familiar (an Owl using the Help action in combat can be very useful). You should pick up Mage Armor right away at level 3. With mage armor and Haste activated, you'll have an AC of 20 (13+5 from dex+2 from Haste), before you even take cover (which you should be doing as much as possible as a ranged character). This will make your standard AC between 22 and 25. If you pick up Shield at level 5, you can even bring this up to 30 for a round, in REACTION to being hit (which could cancel the hit in question) at the expense of a first level spell slot.

There are other options that make for powerful ranged damage dealers at this level (Level 10 bard with swift quiver and heavy crossbows will fire 4 times per round, and ranger5/rogue* beats out fighter at most levels). But this build has a slight (or major) numerical advantage over all of them I've been able to calculate at precisely level 15. One last consideration: if you went Fighter 14, Rogue 1, you'd get the same max damage, lose the cunning action, and get access to one 3rd, 2nd, or 1st level Wizard spell of your choice. Mirror image alone might be worth this: but it all depends on how much flexibility you want on your turn. Personally, I'd go Rogue 2 (cunning action is fun and useful).