Fixing the Eldritch Trickster Racket

[Teridax]

Springboarding off of the samurai and ninja thread, one subtopic of that discussion was around the Eldritch Trickster racket: in short, the racket has the potential to satisfy a bunch of character concepts, among them that of a magical ninja, but more broadly that of a sneaky magic-user who uses a combination of skills, spells, and sneak attack damage to fight and navigate different challenges. In practice, though, the subclass has a few pitfalls: for starters, it interacts quite poorly with archetyping and Free Archetype rules, as it locks you out of other dedications. Perhaps even more importantly, though, it doesn't make the Rogue that much better at using spells in a sneaky way, as Hiding by default doesn't benefit your spells and the racket offers no special benefits in that regard. For this reason, the Eldritch Trickster has largely been left behind in favor of simply picking up a spellcasting archetype at 2nd level, and the racket hasn't been updated for the remaster either.

In my opinion, there are different ways in which the racket could be updated, but here I'm opting for an economical approach which I've outlined in this brew. In summary, here are the changes I'm proposing:

1. The dedication feat you get from the racket doesn't restrict you from picking other archetype dedication feats.

2. The racket makes enemies off-guard to your spells in situations when they would be off-guard to your Strikes. Additionally, enemies off-guard to your spells take an equivalent circumstance penalty to their saves against your spell DC.

3. The racket special-cases the Hide action so that you only become observed to creatures after resolving a spell you cast, while causing creatures you were hidden from to be off-guard to your spells. I honestly feel this could be made baseline to the Hide action, but it bears mentioning here.

4. The Magical Trickster feat, which normally just lets you add your sneak attack damage to your spell attacks, is replaced with the Spell Sneak feat, which adds your sneak attack damage to the damage roll of any spell (including save spells!) against creatures off-guard to your spells. It also uses the wording from the Sorcerer's sorcerous potency feature to ensure this only applies once per spell and per creature.

In summary, the above version of the Eldritch Trickster is made to play much better with Free Archetype rules, the Hide action, and a wider variety of spells, allowing Eldritch Trickster Rogues to benefit much more from all the usual means of rendering creatures off-guard. There are in my opinion many more ways to update the Eldritch Trickster, and another potential direction could be to make the racket a bounded spellcasting class archetype, but this version is aimed to be simpler and a bit more general-purpose, accommodating players who want to use their Rogue's spells for general utility or crowd control as well as direct damage.

[Unicore]

I really enjoyed playing my eldritch trickster rogue once I could get magical trickster and the (then) scorching ray spell, which I had on scrolls by mid way through level 2. Surprise strike is a really powerful feature for going nova in round one.

Which makes me think throwing a fireball in round one against all off guard enemies to get both reduced saves and extra damage is just way too powerful.

The issue with the eldritch trickster is that by mid levels, any other racket can give you everything you are using from the ET racket and none of your follow up rogue feats are worth not just taking archetype feats or good universal rogue feats

[Teridax]

I think the main counterbalancing factor to sneak attack damage on spells is that you're only working with archetype-grade spellcasting, so your spells will be of a lower rank than the blast spells full casters can use, and you'd have far fewer of them to cast as well.

To stay with that fireball example, the earliest you'd access that spell from your spell slots is at 8th level, where full casters can cast 4th-rank spells with expert proficiency. If you were to catch enemies off-guard with surprise attack and then cast your one fireball for the day, the circumstance penalty to the enemies' Reflex saves would make your spell as accurate as that of a caster with expert proficiency, and your 2d6 sneak attack damage would give it the equivalent of 4th-rank damage. This is also pretty much as ideal a case as it gets, as afterwards your spells end up being 2 ranks behind those of full casters. All that effort to do that just once in a day, when a caster of that same level could cast four 4th-rank fireballs at the same level of effectiveness for far less effort, I think is okay for a subclass.

Besides that, however, I do agree that the Eldritch Trickster needs benefits that you can't get just by picking a multiclass dedication, which is why I think it's important that it gets to exploit the off-guard condition better for any kind of spell. I'm not super-worried about the damage a Rogue might deal with this, because the class can already deal a lot of damage with Strikes, so as long as the benefits match up to those you'd get from other rackets (which, considering the strong benefits you get from the Thief, is a decently high bar), then I'd be happy.

[Ascalaphus]

Well eldritch trickster is definitely an itch people want to scratch. There's a legitimate theme here to design with.

The problems I see with it are:
- Your proficiency will start to fall behind your proficiency with weapons. Why are you using spells when your weapons are more accurate?
- Applying sneak attack at range takes considerable work, especially if you want to be able to do it reliably.

On the other hand, you have a lot of stuff casters would envy; the rogue class chassis with its saves and perception is really good. It's like on a line from cloistered cleric, war priest, harbinger - the eldritch trickster is another step further toward full martial. If you want to raise the quality of the caster, maybe you need to pay for it with a reduced chassis.

Maybe this is a good case for a class archetype, in the sense of actually trading out some class features? (I am in favor of making class archetypes non-blocking tho.)

[Teridax]

Proficiency for sure was a major reason why I wanted to have the above take on the racket apply a save penalty to off-guard creatures. Applying sneak attack more easily to spells is a key reason why I wanted to make Hiding better for the racket, but ultimately I don't actually believe that an Eldritch Trickster needs to be ranged: there are plenty of ranged casters already, and I think it'd be much more interesting for a magical Rogue to be able to exploit all the ways of rendering targets off-guard, such as flanking or Feinting, for their spells, even and perhaps especially if that leads to a character that prefers to casts spells at melee range.

I do agree however that there is a good case to be made for a class archetype here, and in general I think the different ways to implement a magical Rogue sit on a sliding scale: on one end, you have your "bag of tricks" character who still fights mostly with Strikes, and uses a bunch of small spells for utility, and on the other end you have your "magical assassin" character who uses spells as a primary means of dealing damage in a fight. I'd say the above implementation sits slightly away from the former end of the scale, as they do offer combat benefits but mainly offer a lot of smaller spell slots to play with, whereas there's room for a class archetype on the latter end of the scale that offers bounded spellcasting. The one caveat to all this though is that I think the end result ought to play meaningfully different from a Laughing Shadow Magus, which already combines roguish elements with bounded spellcasting and caters to a similar niche.

[Ascalaphus]

Yeah, the comparison with laughing shadow is pretty relevant. This should be different enough. Which does cause some challenge because if both of them deliver a spell into a flank, one of them adding weapon damage through spellstrike and the other sneak attack, how big of a difference is there really?

I really like the idea of making off-guard also help against saves, because that makes this feel like "still a rogue, still obsessed with off-guard". And that gives you a definite bias toward melee, without absolutely forcing it.

If we go the class archetype way, we could do something like:
- pick a tradition and a key stat (maybe with some restrictions, no arcane-wis?)
- there are basic spellcasting feats etc, using the standard caster multiclass archetype building blocks
- there can be some unique archetype feats as well
- there may be some things you lose from the rogue chassis (perhaps slowed weapon progression) while your casting progression pulls close to war priest/magus levels

So you're not taking a "bard" or "cleric" archetype, you're taking "occult" or "divine".

Compared to a magus, you're not getting the high-level spell slots, but you might get a lot more of the mid and low level ones, emphasizing the utility/trickery aspects. You're going a bit broader but not as tall.

Some ideas for unique feats/chassis things:
- once per day pulling a scroll out of nowhere that "you'd been carrying around all along" with just the right utility (arcane).
- getting sanctified and a top-rank Heal/Harm font (divine)
- a better version of the usual "breadth" feat that gives you a lot more slots of your lower ranks
- getting some kind of critical specialization effects on your spells
- easily adding subtle traits to spells (occult?)

[Teridax]

I'm on board with a lot of this. Some of the proposals look like they could be satisfied by existing options, e.g. Predictive Purchase to pull out a handy scroll, whereas others would justify the larger-scale changes you could only get from a class archetype. Giving the class archetype Subtle Spell and even some extra action compression on top would definitely be thematic appropriate, and I agree a class archetype would make it easier to make some tradeoffs from the base chassis to accommodate more power elsewhere, such as with bounded spellcasting.

[exequiel759]

I honestly think this brew is fine as is. A bit stronger and front-loaded than other rackets? Maybe, but it loses its power really fast and pretty much requires free archetype to function. I feel the comparissions to laughing shadow magus don't make much sense because the magus eats the rogue's lunch in regards to spells even in this brew in a way that isn't even funny. Laughing shadow has the highest damage out of all hybrid studies, and spellstrike is the strongest nova attack in the game. A rogue in this brew at best is dealing an extra 4d6 SA to their spell's damage, the magus can potentially add 20d6.

It also doesn't need to be an archetype to have bespoke feats. All rackets have bespoke feats and they aren't archetypes.

A slowed weapon progression would make this racket a trap option once again. The only real unique feature this brew has is to add sneak attack with their spells against off-guard targets, because the access to a multiclass spellcasting archetype is something everybody has access to. A different thing it would be if this racket granted you all the spellcasting feats for free, but it doesn't. It pretty much gives -2 to saves to off-guard foes and 2 cantrips. I'd say that's pretty fair.