Vapourware Wich of these progression models to a Skill Tree do you favor?

[DJOGamer PT]

Since this is a specific question I'll give a specific example.

Say your ARPG has a sword/axe/etc. skill tree that with each node/level it expands the range of attacks the characters can perform with the respective weapon said skill has control over.

Would it be better for the Skill Tree to have:
> Model A - fewer nodes/levels but needing more requeriments to unlock (be they of any kind like Stats, more Skill Points, XP tresholds, Items Quests, etc...)
> Model B - more nodes with less requeriments

I know you might just look at them and say: "Dude... both those models in the image require the same number of skill points to master. So that are you on about?".

But seriously think of the implications of each Model.

A:
> the fewer nodes exist, the more important they become and has such there's less likelihood of a node being superflous
> it becomes easier to balance each level
> slower progression as attaining new levels is more demaning
> can however lead to a sense of stagnation and dullness as there's big time intervals between each level with anything new to "spice things up"
> progression can feel "unnatural" since the player spends alot of time without seeing any improvements, and when he does level-up helearns a vast amount of new actions
> another and even worse consequence of this is that such sudden quantity of information can very well "overwhelm" 1st time players leaving them confused about the mechanics
> finally, the early game can easily lose it's challenge as the player possess a good range of actions he can perform
> overall a more "rigid" system for character progression but also more robust and demaning

B:
> the more nodes the less impact they have and some can even become unnecessary to obtain
> this opens up the door for min-maxing
> it more difficult to achieve a proper balance between nodes
> progression is however faster
> there's a nice sense of improvement
> player learn the mechanics smoothly
> early game retains it's challenge but because of the possibility of min-maxing the late can become a joke
> overall a more "flexible" system for character progression but more fragile and trivial

 

[Nortar]

It's not an easy question to answer, so I'll use a comparison diagram.

Model O:
> lots of vitamin C
> requires peeling, so it leads to delayed gratification
> great look and texture, but limited color palette

Model A:
> rich in iron
> instant gratification - no need to peel
> comes in great variety of sizes and colors

The bottom line is, either can taste like shit or like nectar of the gods.
And in every case it depends on the qualities of each particular thing.

But anyways, with all things being equal, I would prefer a system with larger number nodes in hopes that it would allow more flexible character generation and progression.

 

[Silly Germans]

Since this is a specific question I'll give a specific example.
...

You sound as if A or B are exclusive, but you can use both at the same time. You can have a group of skills/perks/abilities/whatever that can be improved with every level up and others that improve only after gaining x levels or so. And most of the points that you address go for both systems and are mostly a question of implementation details. You can have a game that gives you something only every second level but level ups themselves might come very fast. On the other hand you can have a game that gives you something on every level up but every single level takes very long.

 

[Tigranes]

I'm completely puzzled as to why the two models have the implications that you wrote down. I also don't know if A vs B is even the most relevant decision.

What I'd want to know about any system is, what are the qualitative changes in the gameplay enacted by the different skills? Is the meat and potatoes of this system about starting with basic attack animations and unlocking new movesets as the key components of progression? Or is it primarily about giving you all the tools to start with and increasing their numbers? Or it's about different "categories" of abilities (e.g. basic attacks, special ultimate attacks, combo attacks) that get unlocked over time?

Once that is settled, you can look at that and say what kind of tree or whatever might be the best model.

 

[luj1]

Both are retarded but B I guess

 

[Lurker47]

 

[Kalin]

I like hot chicks with big tits.

 

[Trashos]

Best solution: Two different types of gain on levelup, say Perks and Skill Points. You get Skill Points every level, no level up is boring. You get a Perk every few levels.

Skill Points are incremental, make a gradual difference, and do not break balance even if taken every level.
Perks make a bigger difference in style or power, and as such they are more dangerous to the balance. But they are rarer, so the developer has more breathing room for the consequences.

Of course, I am merely describing what Fallout did. But now numbers are the fucking devil because the little brains of our retarded players cannot handle looking at them. So instead, here is a huge picture with a square hole in it. Would you put the the square object in the hole, Timmy, or maybe the round object?

 

[Butter]

I prefer less granular progression in general. There's no point in having level-ups that barely change anything except to give the player some dopamine.

 

[Trashos]

With incremental progression (you say "granular", is that how it's called?), you are actually letting the player create his own skill tree for skills that only lead to changes when certain numbers are reached. Then you combine that with skills that have a continuous progression (eg, damage goes from 14% to 15%). The result is a system that is very rich in content, full of trade offs, and not necessarily too punishing to mistakes.

I personally consider it a genius system. We are going to spend the rest of the days of civilization trying to improve it, only to realiise that we had hit jackpot very early on.

 

[mondblut]

Skill trees and perks are gay. Only incremental skills are the way to go. If you want flashy shit, add it at thresholds values.

 

[Mastermind]

B and anyone who picks A should be castrated.

 

[Mastermind]

Skill trees and perks are gay. Only incremental skills are the way to go. If you want flashy shit, add it at thresholds values.

Incremental skills only are for women and sodomites.

That said, so are skill trees with arbitrary skill restrictions, a cancer propagated primarily by Diablo 2. Incremental skill gain + a variety of passive and active abilities unlocked to train/buy as your skill increases is the way to go.

 

[Darkzone]

B and anyone who picks A should be castrated.

Anyone who picks A or B without thinking carefully about the implications of one against other should be castrated. This way we can lower the amount of some retard genes in the next generations.

A:
> the fewer nodes exist, the more important they become and has such there's less likelihood of a node being superflous
> it becomes easier to balance each level
> slower progression as attaining new levels is more demaning
> can however lead to a sense of stagnation and dullness as there's big time intervals between each level
> progression can feel "unnatural" since when the player levels-up he learns a vast amount of new actions
> another and even worse consequence of this is that such sudden quantity of information can very well "overwhelm" 1st time players leaving confused about the mechanics
> finally, the early game can easily lose it's challenge as the player possess a good range of actions he can perform
> overall a more "rigid" system for character progression but also more robust and demaning

B:
> the more nodes the less impact they have and some can even become unnecessary to obtain
> this opens up the door for min-maxing
> it more difficult to achieve a proper balance between nodes
> progression is however faster
> there's a nice sense of improvement
> player learn the mechanics smoothly
> early game retains it's challenge but because of the possibility of min-maxing the late can become a joke
> overall a more "flexible" system for character progression but more fragile and trivial

I would add that A can feel more rewarding on leveling up, like the feeling in DnD if you reach level 2 or 3. But B can feel more like a more continuous development, while the perceived reward factor declines. Sorry but i cannot decide on this one, therefore i take the Kingcommerad option.

 

[Eisen]

So, where is KingComrade option?

 

[DJOGamer PT]

Sorry but i cannot decide on this one, therefore i take the Kingcommerad option.

So, where is KingComrade option?

Just added it

 

[DJOGamer PT]

You sound as if A or B are exclusive, but you can use both at the same time.

Well because it's a spectrum...

Spoiler

In all seriousness

If there's few skill nodes to attain then it's natural that there should be higher requirements to make sure players don't became legendary sword masters in the first levels
And obviously if there's already a good amount of levels forming the skill tree, you should lax the requirements so the game doesn't become a time-wasting grindfest

It's true that by trying to shoot for the middle I might end up getting the positives of both approaches, but only on a small quantity. But I'll also get the negatives of both (even if again on a small)
So while there might not be anything particulary wrong with the progression, there also won't be anything remarkable about it
That's why I think it's better to go for one side of the scale or the other

I'm completely puzzled as to why the two models have the implications that you wrote down. I also don't know if A vs B is even the most relevant decision.

What I'd want to know about any system is, what are the qualitative changes in the gameplay enacted by the different skills?

It's exactely what I said in the OP's example.

The skills determine the range of possible actions the game world's characters (both the PC and NPC's) can perform
Since the example, for purposes of simplicity, deals with melee weapons those actions are attacks that can be done with the weapons that said skill controls

Expanding on, and even altering a bit, the example in the OP
The skills in question are the sword and pyromancy skill (those 2 branches on the image)
The Sword moveset has 15 attacks while the Pyromancy grants you 10 spells
That means that in Model A, each skill node/level unlocks 5 attacks and 3 spells (one of the nodes unlocks 4 for obvious reasons)
While for Model B, each node/level unlocks 3 attacks and 2 axe spells

The problem now is laid bare

With Model A, each level has big impact on the character performance, none of them can be deemed inconsequential
If are going to make a warrior/mage you have to master the whole tree to make sure the character is effiecient in area of those skills
This Model of progression is obviously more "solid" and "consistent" as balancing becomes easier, there's less "holes" in the system, and best of all the game's can have a nice challenge curve throught the whole experience (with some hick ups in the first levels as the player gets a good array of tools that early)
However due to the small quantity of improvements the player goes through the sense of progression suffers alot
As there will be long chunks of time where the player is essentially stagnant
Then when the player does level up he will receives alot of new actions very abruptly, which not only feels "unnatural" means that 1st time players with those skills will probably be "overwhelmed" with that amount of new mechanics

Model B is essentially the reverse
Each node is of much less importance because it's rewards equally diminished
So unlike with Model A, in B you can very much just get the majority of sword nodes for your character to become proficient in that art
And then use those remaning skill points in the pyromancer tree to cover, even if a little, your warrior's magical shortcomings
So while the system is more flexible on the builds it allows you to craft, it's also more easily broken and worst of all some of it's skill levels are pretty much superflous
In fact, players that aren't out to metagame can very well end up with sub-par builds not because of their own fault but because the system is filled with useless shit
Besides since min-maxing is possible, that means the late can become a complete joke in terms of challenge
On the progression side of things
Because the player levels-up more often, he'll have a more fun fulfilling and natural sense of evolution
And since he learns few things at a time, there's a much smaller chance of not learning the mechanics correctly

 

[Tigranes]

Okay, so you have a very particular setup in mind, then.

If there are 15 unlockable attack moves and all 15 are distinct, why aren't they all separate skills you can choose to unlock, which may or may not have prereqs? Is there a design reason for putting them into packages?

A lot of it also depends on presentation - if it feels like I only go from Level 1 to 5, that's different from thinking of it in terms of, let's say, having two sub-classes or prestige classes to unlock along the way.

 

[laclongquan]

A is more like BG/IWD system. While B is more like Fallout system (the nodes in fallout system are perks that has skill requirement).

A has more advantage is that you can plan for which nodes appear in which stage of a game. Like, you only has grandmaster of axe at minimum level 8 for example, and level 8 is generally late chapter 3 or early chapter 4. A depend more on equipment than B, and if a middle stage weapon appear in early stage it can ruin balance.
B has more advantage is that you can flex the time period of which happen when. Very useful if you want high level skill appear in early stages, or vice versa. In a way, B depends less on equipment and more on character plan.

In short, apple and orange. It entirely depend on what you want for where and when.

 

[KeighnMcDeath]

Fuck skill trees. % to learn anything and everything. Some classes can learn easier and others are hard as iton ballz and nigh impossible but still attainable if you're autistic to train it.

Perhaps paths in the tree if it makes something easier to learn. I did like some of the mods for Diablo 2 that enhanced magic and skills though so maybe all trees aren't too fucky. I can't even recall the first time i knew of skill trees. Before it was all vancian magic or mana/magic points and unlockable shit as you levelled up. Maybe skill trees were there but unseen.

bah.. wtf do i know?

 

[Lhynn]

None, skill trees are utter shit. Go for lists like fallout.
And never make weapon specific skillt trees, those are even more retarded, dont lock your players into a single repetitive playstyle.

 

[Ol' Willy]

Weapons skills with incremental progression - like Fallout; from 0 to 100 and overmaxing if makes sense.
Skill affects your primary offensive characteristics - to hit chance, block chance, crit chance, etc. Weapon skill requirements as well.
Skill value also works as a threshold for learning new moves, which you do learn from trainers/books.

 

[Funposter]

every interaction in the game-world should be governed by its own separate skill, all of which range from 0-100.

 

[Bricker]

"B" of course!

 

[DJOGamer PT]

Guys the question wasn't to discuss if lists are better, but wich of those 2 progression models you prefer to find in skill trees

Is there a design reason for putting them into packages?

C'mon man seriously...
What's the reason for the sword skill to control and improve sword abillities?
You can go ask that question about any other RPG.

Anyway the question here isn't about the whatever or not skill trees are the best design choice for a RPG system, it's between those 2 choices what do kind of progression/design for a skill tree you prefer.
A more demanding one, where your choices carry more weight, and the system itself is more "solid". But at the cost of build flexibility and weird sense of progression.
Or one that gives you more freedom in how you craft your PC and a more pleasent feel of improvement. But at the cost of the game's "integrity" (both in it's challenge and mechanics) and some nodes/choices becoming meaningless.

So which one is it for you. A or B?
If, for any reason, you can't decide then just hit the KC option.