How to make better looking (good and simple) talent trees

[Val the Moofia Boss]

(This is a thread I have been thinking of for a while but I finally made it due to the talk in the WoW thread)

I have always been baffled seeing people on forums or in youtube videos talk about "great" the vanilla WoW talent trees were or FF10's Sphere Grid or Path of Exile's infamous skill forest and so on. It looks impressive, but then when you actually unfurl the paths or examine the design of the nodes, you are usually being railroaded and not being offered the true freedom you desire.

In the current WoW revamp skill trees, you will either stick to the left or the right sides of the talent tree to optimize for single target for raid or AoE for M+. You might pick a few different things for flavor or to modulate your APM a bit, but overall there is nowhere near as much actual customization going on here as the screen or its fanboys would want you to believe.




Here is the FF10 Sphere Grid. It is dazzling at first, but then once you unfurl the tree, you find that it's pretty much just straight lines in which you are getting +2 STR and stuff, and ocassionally a spell or ability.








Path of Exile promises you the illusion of deep customization. But like the FF10 Sphere Grid it is modelled after, it is mostly balooney. The rapidly scaling nature of the game means that you are forced to minmax. That means you are almost taking the shortest path to the same few nodes that really matter of and over.




I could go on.

I am convinced that people just like the appearance of these skill trees. The functional MoP talent trees that just present you with three different meaningful options per tier got the job done and did not waste your time with +1 STR nonsense or pathing to the actually relevant nodes. But the UI was not made to look sexy.




I think these talent menus that only display the actually meaningful abilities/passives just need a cosmetic revamp to look cool, and people would like them more. This BG3 traits screen is perfectly functional and just shows you the stuff that actually matters like increasing your polearm reach by 2 or causing you to automatically counterattack enemies that come within range. It just doesn't look sexy.




This also overlaps with a thread about UI design last year, but the Asians do great job at making cooler looking menus. More games need to invest more art into their UI and talent screens. If we could get a synthesize of great aesthetics and simplicity, we would have the perfect talent screens.


FF7 Remake Materia screen


FF13 Crystarium


Trails Through Daybreak Xipha quartz screen

Arknights talent screen

[rusty_shackleford]

The problem with talent trees is it's often just a question of optimization combined with punishing people for not knowing things they couldn't possibly know at decision time without outside assistance.

Picking between a spec in WoW is often, but not always, picking between entirely different play styles.

[rusty_shackleford]

And to answer your question, I don't know. But I really hate the massive ones that are filled with tiny bonuses. Just an excel spreadsheet

[Oyster Sauce]

The only players that matter are the ones who don't look up meta build orders and instead pick haphazardly. Players who are going to use the strictly best talents won't care how complicated or customizable it is, so you don't need to design your game around them. Ret paladin is always popular no matter how bad it is, and people who pick bad specs are going to have a very large overlap with people who treasure the original WoW talent trees.

If you cater only to the competitive meta-pick crowd you end up killing your genre. See: RTSs and MMOs.

[Stack of Turtles]

I immediately lose interest in any game the moment I'm expected to "optimize my damage output" or anything gay like that. I rarely even look at any talent options further ahead than the immediate current choices available to me. The idea that I would "either stick to the left or the right sides of the talent tree to optimize for" whatever or try to chart the quickest path to some bonus is silly. Norfleet can do that if he wants but I have better things to worry about.

[Stack of Turtles]

The only players that matter are the ones who don't look up meta build orders and instead pick haphazardly. Players who are going to use the strictly best talents won't care how complicated or customizable it is, so you don't need to design your game around them. Ret paladin is always popular no matter how bad it is, and people who pick bad specs are going to have a very large overlap with people who treasure the original WoW talent trees.

If you cater only to the competitive meta-pick crowd you end up killing your genre. See: RTSs and MMOs.

This is also a problem that D&D ended up having, btw. They were too afraid people would get dogpiled by minmaxers for making "bad builds" so they made character advancement almost completely linear.

[Norfleet]

This is also a problem that D&D ended up having, btw. They were too afraid people would get dogpiled by minmaxers for making "bad builds" so they made character advancement almost completely linear.

Character advancement was linear in the beginning, too. Returning to linear character advancement is just a rejection of modernity and an embrace of tradition. Besides, limiting options by action resource economy is more interesting than limiting options by buildspace limitations. One leads to interesting gameplay decisions made during the actual game, the other leads to buildfaggotry. If I receive a class with a wide variety of options and must choose the best option for the moment based on action and resource economy, it makes for a more interesting game than when I customize my class before the gameplay even starts, narrowing my actual in-game responses to following a boring rotation or spamming the most buffed action, being incapable of doing anything else or otherwise making interesting choices because all of those alternatives were optimized out through buildfaggotry.

[Stack of Turtles]

This is also a problem that D&D ended up having, btw. They were too afraid people would get dogpiled by minmaxers for making "bad builds" so they made character advancement almost completely linear.

Character advancement was linear in the beginning, too. Returning to linear character advancement is just a rejection of modernity and an embrace of tradition. Besides, limiting options by action resource economy is more interesting than limiting options by buildspace limitations. One leads to interesting gameplay decisions made during the actual game, the other leads to buildfaggotry. If I receive a class with a wide variety of options and must choose the best option for the moment based on action and resource economy, it makes for a more interesting game than when I customize my class before the gameplay even starts, narrowing my actual in-game responses to following a boring rotation or spamming the most buffed action, being incapable of doing anything else or otherwise making interesting choices because all of those alternatives were optimized out through buildfaggotry.

But linear character advancement MAKES the entire decision point happen before the gameplay starts, because after that your character just follows the predefined path except for like one decision at level 2ish. That's exactly what I'm saying: there should be (meaningful) choices occurring throughout the game instead of just one choice at the beginning.

[Norfleet]

But linear character advancement MAKES the entire decision point happen before the gameplay starts, because after that your character just follows the predefined path except for like one decision at level 2ish. That's exactly what I'm saying: there should be (meaningful) choices occurring throughout the game instead of just one choice at the beginning.

Well, let us consider the trivial case: Linear character advancement in the form of NO character advancement: The Doom Guy (of the originals, I don't know anything about nuslop) does not level up in any way. He doesn't even have the option to pick a class other than "Space Marine". Without any form of buildfaggotry whatsoever, ALL of your decisions are made in gameplay. What weapon do you use? How do you choose to use it? All of these decisions are moment-to-moment, made in the actual game. You don't really get to outsource much gameplay, which is why you're just told "to kill the Cyberdemon, shoot it until it dies". No one's going to publish a multi-page treatise on buildfaggotry in Doom. Cuz there ain't any.

Now compare that to various buildfaggotry games: Most of your game decisions have already been made before you even log in to create your character. The act of picking a skill at level up is merely committing to a decision made before you even began playing.

You can also see this in a very much more direct apples-to-apples comparison: Starfreak Command I/II vs. III. In I and II, you picked a ship and then you fought, using the toolkit you were given. In III, you get to engage in buildfaggotry, and this tends to result in you optimizing away everything and gameplay devolving down into spamming your one attack. It's an absolutely crystal clear demonstration of how buildfaggotry shifts the decision space out of the actual gameplay.

[Stack of Turtles]

Well, let us consider the trivial case: Linear character advancement in the form of NO character advancement:

Hold on, I don't agree that a point is a line. You need to work on your geometry here.

Now compare that to various buildfaggotry games: Most of your game decisions have already been made before you even log in to create your character. The act of picking a skill at level up is merely committing to a decision made before you even began playing.

You can also see this in a very much more direct apples-to-apples comparison: Starfreak Command I/II vs. III. In I and II, you picked a ship and then you fought, using the toolkit you were given. In III, you get to engage in buildfaggotry, and this tends to result in you optimizing away everything and gameplay devolving down into spamming your one attack. It's an absolutely crystal clear demonstration of how buildfaggotry shifts the decision space out of the actual gameplay.

Right, that's linear. That's what I don't want.

[Norfleet]

Hold on, I don't agree that a point is a line. You need to work on your geometry here.

NO U! I was talking about a line with a slope of 0.

Now compare that to various buildfaggotry games: Most of your game decisions have already been made before you even log in to create your character. The act of picking a skill at level up is merely committing to a decision made before you even began playing.

You can also see this in a very much more direct apples-to-apples comparison: Starfreak Command I/II vs. III. In I and II, you picked a ship and then you fought, using the toolkit you were given. In III, you get to engage in buildfaggotry, and this tends to result in you optimizing away everything and gameplay devolving down into spamming your one attack. It's an absolutely crystal clear demonstration of how buildfaggotry shifts the decision space out of the actual gameplay.

Right, that's linear. That's what I don't want.

Which one is linear? Freeform buildfaggotry of your ship wasn't linear. But it resulted in greater linearity than just a canned ship class, because the player would optimize away everything that wasn't specifically part of the target playstyle due to buildspace resource constraints. Then, when the action actually started, he had only one card to play, having optimized away everything else. Having a nonlinear branching build tree doesn't matter when you are ultimately just forced to pick a line anyway, except now that line is even straighter and more linear than before.

[Stack of Turtles]

Which one is linear? Freeform buildfaggotry of your ship wasn't linear. But it resulted in greater linearity than just a canned ship class, because the player would optimize away everything that wasn't specifically part of the target playstyle due to buildspace resource constraints. Then, when the action actually started, he had only one card to play, having optimized away everything else. Having a nonlinear branching build tree doesn't matter when you are ultimately just forced to pick a line anyway, except now that line is even straighter and more linear than before.

The question I'm talking about is whether what HAPPENS IN THE GAMEPLAY is linear. Freeform buildfaggotry is completely linear if your choices are then locked in (whether de jure or de facto) and nothing you do afterward means anything. I never said anything about branching build trees.

[Norfleet]

The question I'm talking about is whether what HAPPENS IN THE GAMEPLAY is linear. Freeform buildfaggotry is completely linear if your choices are then locked in (whether de jure or de facto) and nothing you do afterward means anything.

I agree. So you see how I view buildfaggotry as a more linearizing influence than even fixed, linear progression classes or even no-progression-at-all. Because buildfaggotry involves removing functionality from the character. You're de facto noncompetitive if you simply spread your build resources across many options, and thus you're forced to converge on a single set of skills and then stacking bonii to those skills. And this decision is made before the gameplay even starts. At best, you can reduce lock-in through inexpensive respeccing. But now your entire buildspace has become nothing more than a glorified loadout.

[Tangerine]

This also overlaps with a thread about UI design last year, but the Asians do great job at making cooler looking menus.

This is flash to cover for lack of substance.

[rusty_shackleford]

The asian examples are overdesigned. Out of all the examples listed, I think the first WoW one looks adequate, but it's not the same as e.g., the BG3 feats selection, which would be represented much more flatly rather than a tree structure.
Feats that do have prerequisites could just be child nodes, but that would make it difficult to represent feats that have multiple prerequisites.

I suspect this is less of an issue with 5E which IMO butchered feats. 3E may have had too many Timmy cards, but the solution was not to gut the system.

Apparently this is a (fanmade, obviously) feat tree for Pathfinder: Kingmaker. It's very hard to understand, mixes character ability scores with feats as prereqs, etc.,

---

Regarding the concept of skill trees altogether, I just don't like them. EverQuest had Alternative Advancement(AAs) added in the Kunark expansion, which are sorta like talents, but you were technically capable of unlocking all of them if you put in the effort. This would probably suck in WoW raiding where people would therefore only want those that unlocked them all, but in a game like RuneScape it would be great.
I think it's because WoW is a bad RPG, as played by 'serious people, to be fair. When it's played by people like this:

The only players that matter are the ones who don't look up meta build orders and instead pick haphazardly. Players who are going to use the strictly best talents won't care how complicated or customizable it is, so you don't need to design your game around them. Ret paladin is always popular no matter how bad it is, and people who pick bad specs are going to have a very large overlap with people who treasure the original WoW talent trees.

It's a great and fun game.

[Kalarion]

EverQuest had Alternative Advancement(AAs) added in the Kunark expansion, which are sorta like talents, but you were technically capable of unlocking all of them if you put in the effort.

Minor nitpick (well, major if you're a big EQ fan I guess), AAs weren't implemented until Luclin.

Even with the caveats that come with AAs, I think they're the best option. Save leveling for major vertical power spikes, and use AAs as the grind method to open up new ways to play, or to improve on your chosen playstyle. That way it's fine to make them take lots of time in total (because you're probably only doing them while waiting on new content anyways).

[rusty_shackleford]

EverQuest had Alternative Advancement(AAs) added in the Kunark expansion, which are sorta like talents, but you were technically capable of unlocking all of them if you put in the effort.

Minor nitpick (well, major if you're a big EQ fan I guess), AAs weren't implemented until Luclin.

Even with the caveats that come with AAs, I think they're the best option. Save leveling for major vertical power spikes, and use AAs as the grind method to open up new ways to play, or to improve on your chosen playstyle. That way it's fine to make them take lots of time in total (because you're probably only doing them while waiting on new content anyways).

I have no idea why I put Kunark, I meant Luclin. Thanks.

[Cipher]

The sphere grid has a lot of unnecessary backtracking and also Kimari has nowhere to go but to follow someone else's path.

If you remove the unnecessary backtracking that's just a lineal progression, like one of your images show. So... not sure why someone would heap so much praise for it. In the context of the game it is marginally worse than just a straight up level system as it is linear for all the characters until the post game/very late game. Outside the context of the game is just convoluted and again, a worse version of a linear path.

EDIT: Spelling

[rusty_shackleford]

I am convinced that people just like the appearance of these skill trees.

Well, there's nothing wrong with this, is there? It's alright to enjoy something that you just think looks aesthetically pleasing.

---

Some of these are also just not equal comparisons. The MoP talents aren't a tree, none of the talents have any prerequisites.
One of the strengths of the trees with a lot of minor 'talents' is you can use prerequisites to balance an otherwise unbalanced talent. This could also be achieved in the MoP scenario by having talents that require more than one point.

I think it's a weakness of how "competitive" these games are that your character is arbitrarily locked from learning things that are relevant to him. Going back to my AAs example, allowing characters to continue to learn at increasing cost is more interesting to me.