How does Veeky Forums feel about skill trees?

I hate skill trees because
>Useless prerequisites for skills you do want
>Dead ends following up skills you do want
>The skills are not all necessarily related, but crammed together for symmetry with other branches of the skill tree

One way to look at it would be stepping stones of progression. There's a bit of a power gap between "punch so hard it stuns them" and "punch so hard it blows them up".

Do you think classes become redundant with the introduction of skill trees? Both are being used as vehicles for character identity, but would both of those elements together cause something of a conceptual rift?

The prereqs I'll give you, but the deadends represent the culmination of skill in that particular branch (ideally). That third point sounds more of a flaw in the design of the tree, not necessarily the tree itself.

>Do you think classes become redundant with the introduction of skill trees?
It depends. If designed with that specific goal in mind, a skill tree within a class can be used to define the class while also defining individuals within it. The fighter gets one set of options, the magic user gets another. But it's also true that if you have one, you don't exactly need the other.

I really wanna keep stressing how important it is that any skill tree you use be well-designed.

Almost all systems include some other form of progression along-side that of specific skill sets, so that as you advance 'punch stuff' you also advance' hit stuff in all ways'.

There's no reason to have a long skill-tree that only lets you punch stuff real hard.

As far as class redundancy goes, skill trees and classes pretty much fulfill the same goal; a set path of progression and way to make characters distinct. If I wanted those things, I'd go with a class system, not a skill tree system.

Let me tell you about skill trees.

>the deadends represent the culmination of skill in that particular branch (ideally)

But suppose the tree is Elemental magic is split into three, Fire, Water, Air, and you specifically want Air Magic because of the supportive enchantments.

If your skill tree alternates between attacking skills or the final couple skills are offensive magic, your line is dead. You want to utilize your elemental affinity as a utility role, but you have to spec stupid prerequisites and there aren't any later level, worthwhile spells. You either have to spec for Air offensive spells or sink points into the other branches of the tree for more support.

Just an example, but I typically only like a select few spells/features in any tree and none of the highest ranked shit is appealing to me.

Not to mention the sphere grid.

A friend and i are working on a sci fi setting with premade characters and skill trees, we're using gurps though. The skill tree's aren't too complex and over the course of the campaign they'll accumulate enough points to max out atleast 2 skill trees. The idea is to let the characters get obscenely powerful, and at the end of each skill tree they have to choose one ultra skill.

I pretty much jammed a skill-tree type system into my own game. Every character gets a magic doohicky that contains a number of abilities for them to use. They invest points into the magic doohicky and unlock new abilities, or spend points on a sphere-grid type tree unique to that magic doohicky to unlock passive bonuses or upgrade the existing abilities.

These magic doohickys are modular, the party finds new ones, and can sell old ones if they want, and they equip them like any other magic item, though the number they can equip is limited.

It's gone over middlingly. A couple players have essentially ignored it, another couple have gone all-in, and the others use it as they are prompted. Overall, I'd say it's a good prototype for a full homebrew I'm planning to make, but I'm gonna finish the campaign before I make any solid judgements.

This build is garbage, what are you even trying to do. Dual CotB spark?

Nigga get more defense, you gon' die. This isn't even gonna be Dried Lake viable.

I hear the stress and I'm on the same page as you. What do you consider a "well-designed" skill tree?

Aah, I see what you mean. The knee-jerk design reaction to this would be "just design the air tree in to separate utility, offense, and whatever specialized trees", but the inherent problem with that is just how much content that would be. Because then you'd have to apply similar designs across the board.
I don't think that's necessarily insurmountable, but I'm glad you brought that up.

If you're going to have a skill tree at all, I would prefer something more similar to, say, Witcher 3, where skills are organized into tiers and you can't pick skills from higher tiers unless you have a certain number of points in the tree. This removes a lot of the linearity from your tree, which is my biggest gripe with skill trees.

As far as a combination of skill trees and a class system are concerned, I think this is where 4e is a step in the right direction. Every class has its own tools, and the player's job is to figure out which tools he wants to emphasize.

Nothing says "I am playing a game" quite like a skill tree.

Much like levels in general, skill trees are a mechanic that exists to provide a gated progression that looks like progress.

I guess I don't have a problem with the idea of a skill tree, but the end products.

True, but the gaming component can pretty generally be segregated from the roleplaying experience where applicable. A compelling gaming component (not necessarily a good one, see: 4E DnD) that involves mechanics that are very "gamey" isn't necessarily indicative of a poor quality RPG. It's just a vehicle for narrative.

And that's totally fair, and really I encourage you to talk about the parts you don't like- all the better to subvert them.

Trying to rack up crits with arc.

Yeah, I'm a fuckin scrub. I'm thinking to ditch the elemental shit at the top and invest in in something else. Like energy shield and life. Because I'm at Kaum and the Grand Arena and I just can't stop dying.

Forgot to reply to this post:

Haven't played W3, been told not too because it's glaringly repetitive, but I might give it a shot if it's got a good skill tree.

As far as skill systems go, I really only have video games to compare to. None of my experiences with skill trees left a good taste in my mouth.

I'd much rather play a system like Guild Wars where you could select whatever skills you wanted and then adjust their relative power by dumping points into the relevant attributes.

If not that, then individual skill buying based on points is my next favored. Stronger skills cost more, but at any time you have enough points to buy said skill you can.

I like a system where the reason I picked one skill and then another and then another is because I decided they synergize, not the game. I at least have to FEEL like I decided they synergize. It's not clever if all the necromancer's talents are easily categorizable as "ghost fucker" or "zombie fucker".

Trees are a way to gate options that is often not compatible with other game design choices.

Ultimately any skill tree or level/class system is dependent on the designer to do it right. At worst they can kill imitation and balance, at best they prevent cherry picking.
What are your thoughts on the FFG Star Wars RPGs?

That's the consensus I'm getting and I see the logic behind it.
The side effect of this is that I'm looking at 4e in a more favorable light: The abilities one can choose aren't predicated on the abilities they had previously chosen and are instead merely level dependent.

On my to-read list! Mind giving me a quick rundown of what those look like?

You can't just go all-in on damage, the game slaps your shit for trying to do that.

You need to drop like 60-70% of those damage nodes and invest in either tons of ES or tons of health, along with good ES or life mods on your gear. Additional mitigation such as an enfeeble blasphemy or block is good.

Kaom and Daresso are basically where you find out if your build isn't shit. Daresso in particular acts as a brick wall with his bullshit nigh-unavoidable anime swordwave attack that deals gorillions of damage. You NEED survivability to beat him because he one-shots you if you focus damage.

>What do you consider a "well-designed" skill tree?

Not the guy you were talking to, but here's my two cents:
The biggest pro of a skill tree (or any linear progression system) is that, by the time you invest in it, you end up with a well-defined, cohesive character. The problem is, these characters can feel prepackaged when compared to another character of a similar build. For example, there might be three separate paths in the Swordsman tree, but you either end up as the Big Sword Guy, the Sword and Shield Guy, or the Two Sword Guy.

A freer progression system affords the luxury of allowing the player to, ideally, create any type of character they want. The problem with this is that you lose the sense of focus; what may have been a cool concept in theory may in practice play like a collection of pieces from different puzzles that have been stapled together.

The ideal progression system, then, would be something in the middle, where it is focused enough to produce unique characters, but free enough that the player gets to define their unique character rather than pick one from a premade list. Ideally, under a system such as this, you could build your sword and board fighter such that he will feel different from someone else's sword and board fighter in more than a token way. (So not "his fighter does slightly more damage and mine has slightly more defense.) Of course, such a system would have to be quite complex, which can be a huge detriment, especially in pen and paper games.

TL;DR: Complex, but not confusing; deep, but not impenetrable.

It's by no means amazing, it was just the first comparison I could think of.

>There's a bit of a power gap between "punch so hard it stuns them" and "punch so hard it blows them up".
Yeah, but power gaps are just a matter of keeping pace with the system's power-scaling math, whereas choosing a variety of skill is a matter of character versatility. That means you're spending the same resource to participate in the game as you would pay to make your character more interesting.

Seems like most anons in this thread are talking fantasy games, but I think the Star Wars FFG games handle trees pretty well. There are multiple branches in each specialization that have lots of crossover so you can definitely focus on abilities you want. Also for talents that don't stack on themselves you are considered to already have them if you buy into a different tree

You bring up a super interesting point, because you're right: with two players that go through similar progressions end up being identical. Subverting that is important if the "identity" is to actually be meaningful.

>That means you're spending the same resource to participate in the game as you would pay to make your character more interesting
Could you elaborate? That didn't sink in like it should have cause I'm dumb as hell.

This has been suggested and I'm quickly finding myself needing to read those systems.
>Also for talents tha tdon't stack on themselves you are considered to already have them if you buy into a different tree
Could you explain that in a different way? Not too familiar with the system (though I will be soon, it seems)

>Could you elaborate?
If you're upgrading one ability, you have the same number of abilities. You're essentially getting a mechanical bonus, going from +1 punching to +2 punching to +3 punching, if we reduce it to its simplest form. If you get another skill (not part of the tree) you go from being able to punch to being able to punch and kick. Again, an example over-simplified to the point of banality, but the point is that you have more options in the latter case. The former case isn't necessarily inferior if that's the only way skills work, since it's just a way to balance the numbers, and a system can allow you to get new skills a different way. But if you mix the two, as skill trees usually do, that means players can choose to either increase their one ability in effectiveness, or increase their variety of abilities. That means that from a game designer's perspective, you have to somehow balance for a character who takes each of those possibilities to their extremes. So you have the guy who can punch at +3 power, and you have the guy who can punch, kick, or headbutt all at +1. So... What level of numbers do you balance for? What level of versatility? No matter what you do, either versatility or specialization is going to be underpowered compared to the other option. If versatility is weaker (as it is in many games) you're encouraging players to make their characters as simple as possible. If specialization is weaker, you're encouraging the "batman wizard" paradigm where characters have solutions at the ready for everything and DMing anything other than hack and slash becomes a nightmare. Of course, either of those problems can be partially ameliorated by other areas of design, but it's not necessary if you just don't have skill trees. The gradient in power can be covered by passive effects of leveling up instead, or some similar mechanic.

Excellent, thank you!

This is essentially my opinion. I don't mind pre-requisites in the sense of having certain skills or ability scores first (can't be a rip and tear machine if you are weaker than a teenager girl) but the actual benefits should be one-off and stand alone. eg.
Okay
Feat A
>Must have 12 strength
Feat B
>Must have proficiency in heavy armour

Annoying
Feat A
Feat B
>Must have Feat A
Feat C
>Must have Feat B

I'll point out again that, although "punch so hard it stuns" and "punch so hard they blow up" are related to a similar idea (punch super hard), there's a powergap that has to be addressed. I'm not sure what the problem is if one skill requires another, less powerful and related skill before being taken.

Devil is in the details; a skill tree where you're not left feeling that you must take stuff that's of no use in order to get stuff that you will use won't go down well.

As I have something of a hate boner for planning out builds, I don't like skill trees very much in general. I like my mechanical character progression to stem from what the characters are doing in the story and how they're interacting with the world. Growth that makes sense in terms of the events unfolding around them, rather than growth that is based solely on what you wanted to get at level twenty or whatever.

People don't like being forced to take something they don't want in order to be able to get something they do want. It would be like if you had to buy frozen yogurt before you were allowed to buy ice cream. This is a problem unique to skill trees and it's all to do with the presentation of it. If the skills in question were merely level-gated, for example, if Weak Punch was a level 1 skill and Strong Punch was a level 2 skill, then anyone who wanted to make a punching character would probably take Weak Punch at level 1 and Strong Punch at level 2, no problem. The trick is, they didn't have to. They could have taken Stunning Punch at level 1 if they wanted, or even Dagger Slice if they were so inclined. Once you make Weak Punch a requirement for Strong Punch, however, you have removed that freedom to choose, and people hate when you take away their freedom to choose.

Obviously, you don't want your amateur fistfighter to have Nuclear Punch at level 1, nor do you want the dedicated fighter to take a high level wizard spell on a whim, but this is why you put different kinds of prerequisites on these things, such as ability scores, class, level, etc.

Like I said, this is all about presentation. Even something so simple as making Punch a skill that can be upgraded as you level, which serves the exact same function as the skill tree, comes across as a much more palatable option than the linear skill tree.

Each class has 3 (in the core book, there are 2 more in their supplement) in class specialtys that give extra in class skills and a skill tree of abilities, you start with 1 and can buy others but it costs more to buy an other classes specially than your own.

Thought I recognized RO from the thumbnail. Oh nostalgia.

I love skill trees and planning our characters mechanically. The extent to which I enjoy it gives me suspicions that I might be mildly autistic.

Spitballing a branch system

Learning a new skill:
>unlocks 2 skills, which are either variations or combos that inherently include the use of the original skill to execute
>>Fireball->Will o Wisp(homing fireball) or Chained Blaze(multiple fireballs)
>unlocks 2 skills within the same field, often basic skills scaled up with brute force
>>Throw Rock -> Throw Boulder(bigger Throw Rock) or Rocket Punch(bigger Knockout)
>Unlocks 1 skill, something of practical value to [class] regardless of what else they've picked
>>Shadowstep->Reckless Ambush(when moving out of Stealth, lose your next turn to act again now retaining the Sneak Attack benefits)

If it's like that, I don't think it's necessary to ever put more points into a known skill.

Ok, so something like what 4e did would be a bit more forgiving: you can take the skills you want with the only prereqs being that you're the appropriate level to get them. Would that be a fair assessment?

Bad servers, terrible bot problems, power creep like crazy, pay to win, but y'know what? Fun game and great community.

In RO at least it prevented some classes from getting all the best skills in same build by making prerequisid abilities eat up skill points.
One of the private servers I played on increased the max skill point level a bit and it made certain classes extremely imbalanced.

Are there any PnPs that use a sphere grid for character advancement?

A Path of Exile styled leveling system is my favorite permutation of the talent tree concept.

Yeah but let's be real: every class only had about one or two viable builds. The rest were trap options that you had to research to avoid.
Not even to mention that the only discernible difference between one Spear Knight and the other was the quality of their gear.

Believe me when I say that trying to do a 1:1 comparison between RO and tabletop gaming is a fool's errand.

Skill trees, like all things, need to be executed correctly.
I played an Innocent in Hunter: The Reckoning. The skill tree was so gimped that it was almost physically impossible to progress, as gaining new skills required you to use your abilities in useful ways. And none of the basic abilities were useful.

Anyway, skill trees have to keep a balance. They shouldn't be fragmented shits where you question why it's even a tree, but they also shouldn't be constrained bullshit where you feel like you want a higher skill but NEED to get one of those shitty skills beforehand, even though you would theoretically have access to actually interesting skills somewhere else.
One of the worst offenders I've seen here is from vidja, Borderlands 2 specifically.
No, I don't want to invest five of my precious points into a gun type that I'm not even using, just to get that one upgrade that actually suits my style.
And that's in a game that allowed *any* of the earlier skills in the chain to add to the requirements of later skills.
It would have easily been fixed had you been able to invest more than five points per skill. So if layer 2 of the skill tree is shit, but 1 is awesome and I want 3, I'd skip 2 by investing additional points in 1, maybe with some kind of penalty like needing two points (which still count as two pre-requisite points) per skill increase.

If done well, skill trees make for good specializations, but the occasional off-tree skill allows you to add a synergy of your liking.
Specialized builds would have more power in one area, while fragmented builds would make good use of synergies, which would increase their base line while their top line stays at a lower level.

>It would have easily been fixed had you been able to invest more than five points per skill. So if layer 2 of the skill tree is shit, but 1 is awesome and I want 3, I'd skip 2 by investing additional points in 1
Interesting suggestion- it would allow for super specializations without gimping players that wanted more of a variety of abilities. Good shit.

If Shadow Quest and Homeless Mutant Quest taught me anything, it's that Veeky Forums fucking loves skill trees.

If I'm reading the vibe right here its almost like you want your skill tree to develop opposite of how they are usually implemented. Instead of having a huge mound of garbage skills at the bottom that sharpen to specific capstone you should aim for a very limited set of starting skills (like say, use martial weapons, use magic, and use stealth) that continues branching and expanding the further out you go. This way instead of feeling lime you are working through a bunch of prerequisite you don't care about to get the skill you actually want you are continuously expanding your abilities, hopefully in such a way that going tall or wide allows you to access the same results and with multiple branches so that you have many ways to access the same high tier skill

I'm not sure if Borderlands is a good example here. It's a system that allows more modular advancement, gives you enough skill points to eventually max out 2 of your 3 skill trees, and let's you respec for a negligible cost. Also, most skills are just broad enough to suit several playstyles.

It's okay in Borderlands 1, but Borderlands 2 was very borked due to a shitload of balance issues.
The problem with the skill tree was a symptom of other facts, like the fact that pistols were objectively superior and elements were useless. Especially the fucking gyrojets, which are the only elemental guns that gain bonuses from Axton's skill tree.
His skill tree becomes a mine field because half the choices are useless because the weapons that they augment are objectively shit.

Maybe. That sounds good on paper, but I think that risks content bloat. Then again, I suppose things like skill trees, classes, or anything that helps refine a character's "identity" is prone to content bloat to accommodate for various identities.
Ultimately, I think it's about having a gameplay mechanic that makes sense as a gameplay mechanic but also serves as a narrative of progression.

It would also be interesting to have general purpose skills at the root of the tree and more and more specializations as you travel upwards.

And, like in the Borderlands example, any of the lower level skills allow you to add to the prerequisites to the higher levels.
So, for example, you get all the defensive buffs at the bottom and then the "truncate damage" skill that is somewhere towards the end and skip everything else.
This forces you to make a balanced build (because you can't just cherry pick your overly specialized glass cannon by taking all lategame perks) but gives you mostly full freedom over how your character is expressed.

The only issue is that there needs to be some kind of branching limit. Otherwise you'll just cherry pick after you invested enough into basic skills, which makes branches pointless.
Or you ignore branches from the get-go and just make it a linear system with levels that are reached after a point investment threshold, where each level has a bunch of skills that are similar in power but not thematics.
If you want to limit it, you can then force the player to choose one skill per level. And maybe there could be skills that just allow you to take another skill on a level of your choice, (or overload an already gained skill to make it more powerful) instead of getting something on the current level.

Personally, I'd just give players "modules" which can be assembled into different builds.
Instead of getting a new skill via point investment, you buy yourself a new module from a list that is limited to your circumstances.

How the modules are assembled depends on taste. One possibility is just doing something like Materia in Final Fantasy, where a piece of equipment can get modules added to it and modified through that.
Personally, I'm currently working on a system for an FPS where you have a Diablo-like 2D grid for your inventory and modules are interconnected geometrically. So you put your gun there, then the fire element next to it and you get a fire gun. Then, if you are clever, you put it such that the fire element also touches the muscle module, allowing you to imbue both your muscles and gun with fire.

So a player could just buy a bunch of fire modules to get a build with fire powers. Or defensive modules that add defensive properties.
If done correctly, you can do something like a plug and play skill tree where players are able to experiment between encounters.

Been playing around with an idea sort of similar to that. You'd have "classes" that grant passives that define "who you are". Each class would have access to modules of skills that one can invest in that equate to "what you can do".
Classes, in this supposed system, are more of a narrative role instead of an actual identity.

This thread is full of cool ideas. Thank you Veeky Forums you saved the day!

>Nothing says "I am playing a game" quite like a skill tree.
In reality, you'd have inverse skill trees.
Pick a second martial art and it's easier to learn thanks to your first martial art.
And both make it easier to learn anything related to using your body.
This reaches a point where some skills become freebies.

This is true and it ends up being more three dimensional than you'd think. That said, I'm wondering if class trees wouldn't be better tools for conveying progression and specialization.

>it ends up being more three dimensional than you'd think.
I heard bad things about doing competitive judo and then taking dancing lessons.
I was told it's a good idea to learn dancing *before* you learn how to throw anyone who randomly touches you by reflex.

>tl;dr - skill cloud
I like the idea of a skill tree in that it can show an immediate and visceral evolution of skill without a training montage and I see each new node as a flash of inspiration more than a grind. Like melee 1 would be how to throw a proper punch so you don't break your hand

I think a skill web is a better way to go, or actually more of a cloud. It can solve the issue of buying a skill you don't need or want to get to the one you do. The problem I see with that though is it gets messy quickly; Ranged Weapons and Alchemy might not be near each other or related, but you could have grenades at their intersection. Similarly, strength kind of strength based skill would be a key ingredient in both throwing punches harder and shooting arrows farther

The Ranged+Alchemy thing actually reminds me more of Doodle God. Connecting skills you have and following the path around the surface of a cloud seems more fun (maybe even more realistic) than refining them to the finite edge of a tree. It basically offers infinite trees to follow.

the FFG SW biggest advantage is how tempered a skill is.
beggingforxp.com/wp-content/uploads/2016/01/StarWars-FaD-TalentTree-Color-v2.pdf
While the bread and butter of quite a lot of those trees are still 'get blue dice to x', where blues are the smallest dice, they are rarely skills, but narrative qualifiers. The draw of a specialisation comes from doing stuff in a different way.

If your party Soldier and your party Assassin both have massive dice pools in ranged(heavy) - the rifle skill - that is fine. Their talents mean they are going to be rolling that dice pool in different ways - the Assassin might be decreasing difficulty at long range, where the soldier might be giving foes penalties when he shoots.
I think talent trees are useful when they take a narrative idea and group together the interesting and different ways in which different members of the party might accomplish the same dice pool task. They aren't as exciting when it is about strict bonuses.

Say you have a sword skill in your fantasy game. You have three fighter classes: A dashing rogue swashbuckler, a grizzled military man and a knight in shining armour.
Don't give any of your fighter classes 'better with sword'. All the fighter classes are meant to be able to do that, so place it outside the tree, as a skill or generic ability.
The Swashbuckler probably wants a talent that makes him better at using props and scenery in his fighting, so give him a talent to reflect that in his talent tree.
You also want the Swashbuckler to be good at taunting and drawing attention, but you realise that you also had a Knight's challenge idea for Mr Shining Armour. So maybe give them both that in their trees. The knight and Merc might share talents to do with dealing with military folk.
If you find yourself sharing too much, consider multiple, smaller trees (Like fighting style, background, history, etc)

Also, strict upgrades feel like wasted XP. When weak punch -> Strong punch was mentioned earlier, it feels lame because once you have strong punch, there is no need to ever use weak punch. That level can't help but feel a bit wasted, and a little like a tax, even if unlocking the pre-req was needed.
So if you have to have pre-reqs in the tree, make stun punch the pre-req for strong punch. Even late in the game you have a choice then of punching for more damage, or punch to disorientate and help an ally. The level doesn't feel wasted.

That's something that I noticed, but wasn't able to put into words like you just did- and I agree, even just from a cursory look: the mechanical benefits are being informed by a greater narrative at play and that's pretty fantastic.

I see what you mean and I'm starting to get a clearer idea of how these systems ideally work in tabletop settings. I think at this point it's a matter of reducing content bloat and making it accessible. Nothing wrong with a lot of content, but it can't end up looking like a hot mess.

For GURPS that's a great idea, considering its progression is mind numbingly boring

It's a god point but how do you differentiate them,
say the differences shouldn't be in that particular tree but in others?

They shouldn't exist as a specifically designed part of the game, but can be created afterwards as quick reference.