Talent or Skill Trees in Tabletops

Skills trees is just about one of the worst things in Pathfinder.

Legacy from 3.5 feat tree.

I'm more fond of sidegrades than upgrades myself.

What do you have in mind for sidegrades

...

>Are there any ideas to incorporate something like this into D&D 5e?

Have the sphere's be tied to an alter in an ancient temple, these alters become tied to the... essence or whatever, of whoever spills their blood into/onto them. They need to find a kind of precious mineral, something maybe spiritually charged? And that mineral has to be either found sphere shaped (exceedingly rare but provides extra bonus' because the artisans who made the alter made it and so they're master crafted) or have them shaped (varying degrees of quality) to fit the sockets on the alter. The sphere's could be the size of a walnut for simplicity sake.

An ancient tome tells them what each sphere socket does, or maybe they just know because spiritual connection or something, and each socket provides a small bonus to a feat or a skill or something, and every once and a while you'll find a very large slot (fist sized?) that can effect core attributes.

If you want to really go wild with the concept, have some sphere sockets come with penalties too, either in the form of story or gameplay.

I'm pretty drunk and this is pretty shit, but I hope you get something out of it.

Let's talk about skill trees in general. What do you guys like about them? What don't you like about them?

One issue I have is the completion paradox. Once you go down a given path, there's a limited number of reasons to stop going further. Once you start sinking points into a multi-point node, there's a limited number of reasons to not sink the max number. It's an unfortunate war between completion feeling satisfying and choices only mattering when meaningful.

I wonder if some kind of gem slot thing would be better for this kind of situation: Instead of picking a point in a tree and then going left or right, you acquire a skill from elsewhere and slot it into the tree, with the specific pattern having an impact.

Of course, this provides a number of weird issues. For one, video games can often get away with +3% armor or 1% chance to cast level 3 fireball because the computer doesn't give a shit. For an actual tabletop, you're usually pretty light on minor bullshit slots; each +1 is often too significant to be handed out casually, and nobody's going to remember to actually apply the super situational stuff.

For another, if you're getting the skills from somewhere else, what the fuck is the tree itself doing? Do you need to slot skills to use them? That could be neat, but now it's less a tree and more a, what, loadout graph? Do you play Cleave in attack position, hoping enemies will activate the Healing Word located in your Trap Slot?

That sounds kind of neat, actually.

Very JRPGish and fits with players' insatiable need to be magpies. I like it.

A more modest proposal might be some kind of class token bullshit, wherein leveling up (or whatever benchmark or condition we're using) grants you a number of tokens based (randomly if desired) on your character choices. Might be stats, might be class levels, depends on what system we're using for the rest of it.

Anyway, you get your tokens and they're basically your stat/skill points, but where you slot them matters not just for the actual skills taken, but for the type of token slotted.

If you wanted to be a lazy shit, you could make this a simple efficiency thing; maybe there's a Warrior/Thief/Mage tree, and they get bonuses for Strength/Skill/Intellect gems. So a thief CAN learn a bunch of spells, but he'll suck compared to a mage. This is precisely my beef with such a system; I'm not a fan of inefficiency as a motivator in builds. If you make retraining nice and easy that solves much of my trouble, but then you get characters shifting from gentle healer to critbuild reaper of souls over the weekend, which is a little odd.

A better solution would probably be to get all mystical and make the actual slots a little on the bland side, taking much of their impact from the type of gems you slot. So maybe there's an Attack Ring on your tree, but if you slot Red Gems it becomes Fireball and if you slot Blue Gems it becomes Frostbolt and if you slot Chartreuse Gems it becomes Witty Quip. This could potentially allow untold nuance, as you custom-build your moves to be weakass attacks that heal you, debuffing AoE strikes, really good against foes already on fire, etc. It could also potentially solve the completion issue I mentioned far above; you could get a cleave-style move up high enough to take down mooks and then decide you don't need it any higher even though you have slots to spare, for instance.

I think the biggest thing vanilla through Wrath talents had going for them was the feeling of slowly gaining power each level in a way you could see and interact with. When you'd ding, you had a choice in how you grew stronger and you could see and feel it in a more real way than the simple stats increase you'd normally get. The new talents are technically better and more easily balanced, but you just feel like you've got a bunch of levels between talents where you don't really grow. That, and they never truly solved the whole "false choice" problem old talents had of most of them just flat-out being inferior to other talents. Even now most specs have one good build and maybe one extra build for a different situation like needing more AOE/cleave, or a gimmick build that only works due to certain legendaries in Legion making them viable. It also doesn't help that right now, most of the talents that actually give you new spells are just worse than grabbing a passive +% damage passive talent in that same row.

In the end trees provide less depth than just straight system where you buy skills. The function of a tree is to make it to limit how many things you could pick from at any given time. You have to buy things in a certain order and they encourage specialization so you tunnel vision on a narrow aspect of the whole skill pool. It's designed to make it accessable.

The start of making a character might be as simple as choosing between 3 trees and than picking between 2 first tier talents. Much easier to grasp than having the whole pool of skill points open to you at once.

Edge of the Empire is the best implementation that I've seen in tabletop.

The system is essentially point-buy: you gain XP and spend it, no levelling up. The 3 things you spend XP on are abilities, skills, and talents. Abilities can only be bought at chargen, and you start with a certain number of free skills. Skills and Talents are the primary way that most characters advance during a campaign. So, when you spend XP, you're often getting some cool new power.

The nodes themselves are mostly good and meaningful.
Around half of the nodes on each talent tree are active abilities. They're fun and exciting to use. They make your character feel significantly more powerful and distinct from the other characters.
Most of the other nodes trigger in specific, thematic ways: gadgeteers never run out of ammo, survivalists can move freely on rough terrain.
There are a few filler "+1 to repairs" and "+2 HP" nodes on each tree, which are kinda lame.
The "+1 boost dice to X skill" are slightly more interesting, because they effectively give you more "advantage" for that roll while only slightly boosting your "success" rate. So, they're a nice complement to spending XP to upgrade skills: ability dice make you more powerful, boost dice give you more finesse.

Each talent tree is locked to a different class. Unlocking a class requires a starting investment of XP, but gives you a few free skills in return. Your character will generally have access to 1-3 talent trees, so you have a decent amount of options without being totally choice-overloaded like a FFX or Path of Exile sized skill tree.