Whole talent system rework

[yabluchko]

wall of text approaching

Current talent system recap:
a lot of talent trees, that could be locked, 4 talents in each tree, 5 levels in each talent.
Trees are divided into class and generic.

I see some problems in current system.

• Class skills vs generic skills. Long ago there was some difference between them, but is gone now:
  There are some generic skills available to single class. There are pure utility class skills, and pure damage generics.
  I see no point in talent separation into two categories at all.
• Every tree has exactly 4 talents in it. It leads to existence of a whole lot of useless talents used only to fill trees.
• Every talent have exactly 5 levels. This is fine for some talents, but it leads to scaling problems, crappy diminishing returns (where you put point in talent and gain almost no boost). Having levels is fine for talents where it gives duration/cooldown reduction/range/some extra effects. But on pure damage talents it just feels awkward.

My proposal.

Useless talents
All useless talents that are not appealing to anybody must be removed, and if they give some interesting effect it should be incorporated into other talents.

Generics
The only generic talents left must be generic. also no trees, just talents:

• Talent to use heavy armor
• Talent to use light armor and get additional avoidance
• Talent to use robes and get negative fatigue
• Talent to use two-handed strength weapons
• Talent to use staves and get magic boost from it
• Talent to use mindstars and get mindpower boost from it
• Some talents to increase some resources regen
• Racials

and so on
talents for equipment proficiency mus be one-pointers: one should either have two-handed weapon proficiency (+100% danage)or not. no 5% increases.
two generic points must be given at birth to choose preferred armor and weapon, and then one every 5 levels.

Class talents
Trees must not be exactly 4 talents each, and all remainig skills must be somehow rearranged into
fewer trees.
Almost all class talent scaling must be re-adjusted, to make investing points interesting,
so each invested point would give some noticeable effect.
Some examples to show my point of view:

• Firebolt - two levels, same damage. First level - fires a bolt, second level - a beam.
• Phase door - First level allows to teleport to teleport self or other somewhere, second allows to specify
  target location
• Mutually exclusive talents like celestial/hymns or psionic shields should be incorporated into one talent
  giving multiple abilities
• Gloom - four levels: first level confusion/stun/slow, second level adds weakness, third adds dismay, fourth - sanctuary.

Also amount of class points must be adjusted to make some point distribution decisions necessary.

Category points
I am not sure how category points must work with this system.


I want do discuss it, and if I am not only one who feels that way, think it through more deeply and, may be, implement it

[The Revanchist]

Two questions right at the start.

What merit is there in this new system?
What level of experience do you have writing addons?

I'd be interested to see how it plays out... Before Darkgod reworks the entire class system.

[Crim, The Red Thunder]

wall of text approaching

I see some problems in current system.

• Class skills vs generic skills. Long ago there was some difference between them, but is gone now:
  There are some generic skills available to single class. There are pure utility class skills, and pure damage generics.
  I see no point in talent separation into two categories at all.
• Every tree has exactly 4 talents in it. It leads to existence of a whole lot of useless talents used only to fill trees.
• Every talent have exactly 5 levels. This is fine for some talents, but it leads to scaling problems, crappy diminishing returns (where you put point in talent and gain almost no boost). Having levels is fine for talents where it gives duration/cooldown reduction/range/some extra effects. But on pure damage talents it just feels awkward.

I'm sorry, but what? There aren't 'useless' talents, name some. Everything has at least some use, in one build or another.

Why is 4 talents a tree bad?

Every talent tree has infinite levels, scaling was changed specifically to provide for long term scaling across a great many levels. This is frequently seen late-game in enemies, and by players in the ID. And you can't apply extra effects to everything. For instance, how would you balance flurry if you were giving it cooldown reduction? Or the basic tribeam abilities on an archmage.

Finally, the class/generic split is designed FOR balance. Moving desirable trees from one to the other to force tougher choices on builds is a method that has been used to balance some classes. (See: Solipsist and Dream Forge)

[SageAcrin]

Some talents do still have issues, but they're all worked on to be useful to some degree. Any talents that currently have issues are more of a product of people not having gotten around to fixing them than any intended design, and I feel the vast majority of them are useful right now.

As Revanchist said; There's flaws in the current system, but are the benefits to overhauling it notable?

You suggest every single level of every talent give strong impact; How, without massively upping the power scale of the game, and potentially making enemies(who, remember, get levels over 5 of talents) much more capable of surprise oneshots?

You suggest that all "generic" categories must be totally class-universal, but what is class-universal? I've ran Celestial/Light on a Bulwark, a magical category, with extremely good effect to the point of suggesting it as a primary build option. Would this be something you'd remove? If so, why? What is the gain?

(Also, there is actual distinction between Class and Generic; Generic skill categories can be unlocked in the world, but Class never can. It's more of a thematic distinction, I admit, but it still matters for gameplay enough to count.)

You suggest "no useless talents", but how many of those are there? Who decides what is useless? You don't even know what to do with Category points yet?

You suggest an overall massive reduction in point requirements for the game in general. Would the Class/Generic points gained by the character be lowered to compensate? There would be no long term choices if everyone could simply buy every talent their class has and have piles of points left over-an effect I consider detrimental. If there is a lowering of point quantity, to what values?

Don't get me wrong, I'm not entirely adverse to the concept of an overhaul-though it's pretty late in the game's development for that. But I don't see what this gains; It reads too much like "Make the Talent system better" as a suggestion.

All well and good, I'm not going to say there's no way to do it or no room for improvement, but how? Like Revanchist said; This is something that making an addon and showing your first draft would work better to convince people on, rather than a wall of text.

[yabluchko]

Two questions right at the start.What merit is there in this new system?

I'd be interested to see how it plays out... Before Darkgod reworks the entire class system.[/quote]
I want to have noticeable increase in character power with each invested talent point. Not just 5% damage
on some skill. And to make some decisions like "do i want an shield ability or attack ability, or augment some ability with cool effect?" questions when leveling up, not like "do i want to add increase shield power from 39 dam to 43 dam, or will i add 10% damage to this attack ability"

What level of experience do you have writing addons?

I am experienced in programming in general, but as for addons, i just glanced at it few days ago, wanting to change some mindslayer skills, and haven't seen seen anything extraordinary difficult down there.

I'm sorry, but what? There aren't 'useless' talents, name some. Everything has at least some use, in one build or another.

they are not completely useless, but they are just bad, as they are created not to add some cool ability, but created to fill 4 talent trees.
Few examples from classes i played

• Two-handed maiming/(Sunder Armor/Sunder Arms/Blood frenzy)
• Warcries/battle cry
• Hexes/empathic hex
• Sun/Sun flare and Sunburst - what is the difference between this talents?
• Mobility - whole category was made to put mobile defence somewhere
• Torment/Blood lock
• Light/Bathe in Light

ok, you are right, there are not that many useless talents, and they can be useless just in my opinion.

But there comes the second problem - scaling

Every talent tree has infinite levels, scaling was changed specifically to provide for long term scaling across a great many levels. This is frequently seen late-game in enemies, and by players in the ID. And you can't apply extra effects to everything. For instance, how would you balance flurry if you were giving it cooldown reduction? Or the basic tribeam abilities on an archmage.

The problem is that you cannot add good scaling to pure damage skill. If you will add linear scaling it would be too bad with low level or too god on high level, and with logarithmic scaling is just too depressing to invest talent points for 5% damage increase.
My opinion is that such talents must not scale at all.
Talents like manathrust must just have two levels - first as a bolt, and second as a beam. with same damage. Or first as lighting and second as dazing lighting. Or something like this.
Of course number of talent points received for leveling up must be reduced.

Finally, the class/generic split is designed FOR balance. Moving desirable trees from one to the other to force tougher choices on builds is a method that has been used to balance some classes. (See: Solipsist and Dream Forge)

There are other ways to do balancing. Generic points just feel unfair for some classes.
I played corruptor and alchemist - there are no generics except escorts and racials, to put talents in, and I played cursed, who have to put 20+ generics, almost a half (armor mastery/weapon mastery/accuracy/stun resist) just to be able damage enemies in melee. I am not sure if this somehow balances them.

[Faeryan]

How I see this: Someone claims there is a problem. He then sets the rules on how things should be fixed, yet doesn't present the actual fix.

That actually sounds a whole lot like what the politicians all over the world are saying.

Is there really a problem? That can be argued but for the sake of conversation let's answer yes.
Does the existing problem need a fix? Yes.
These are the rules which we must abide to fix things, okay? Uhm, we still need the actual fix, not just the rules on to fix it.

Yes, yes, yes, now give me the new improved talent tree so I can tell if it really works better.

[yabluchko]

About generics and celestial/light on bulwark.

there should be no generic/class talents. All talents should be same, and all "generic" talent trees should be available to any class. Relying on specific escort to get access to specific talent tree is not something I am fond of. I am sure i am not the only one who feels this way.
So bulwark would just use category point to unlock Light tree, and spend some talent points in it. Or get 1
point in Healing light in locked tree from escort, like now.

Of course talent point amount should be reduced to make taking all talents impossible.

And about making addon first. This is ideas forums, it is to discuss some ideas, so do not demand implementation right now

[Faeryan]

And about making addon first. This is ideas forums, it is to discuss some ideas, so do not demand implementation right now

Then you need to speak more as so far what we've read doesn't really open us all.

[ghostbuster]

Nobody think the present talent system is perfect, but it is the result a very large effort by many coders and testers for at least 3 years. Making such a game is difficult, and balance is a fundamental issue. It is easy to break the game and to make it either undoable or so easy it will be boring. And it seems that people are rather happy with the global balance presently.

So, yes sunder arms is less useful than rush. So what? Not all the talents can be very powerful, otherwise you can easily break the game balance.
And yes gaining 10% in an ability is not useless. Most ennemies cannot be one shoted but if you tailor your talents carefully this addition of 10% gains will allows you to be the winner. ToME is not a game on fighting, but rather on the build of a character with talents and luck (items, rewards, etc). What makes this game so interesting is the diversity of the potential build. Presently, for most classes, it is not possible to invest in all the talents and you have to make choices. It make replaying a similar character a new experience.
Massively reducing the number of talents as you suggest would maybe render most build similars.
Maybe, maybe not. But just to know if this change is interesting requires monthes of tests and tweaks. I am not convinced that the potential advantages of this rework are worth it.

[Grakor456]

Relying on specific escort to get access to specific talent tree is not something I am fond of. I am sure i am not the only one who feels this way.

I agree with this in principle, but I don't think there needs to be an overhaul of the talent system to make this happen. I just feel that some more trees need to be made available in reliable ways, so that escorts don't just end up radically changing builds. I'm cool with random chance dictating items (it's not my favorite system but I understand most people like it) but I don't want to feel like I'm rolling the dice on my character's abilities.

Though, just making more reliable generic tree options would be helpful. Especially for non-willpower classes, since Antimagic really skews this aspect in their favor.

[HousePet]

The suggested system would be difficult to balance and would also mean you would have everything you could use/need within about 10 points.

I can't see it working.

[Grey]

I agree with several of the points, but no system is perfect really. And it would work better to write a new game from scratch than to try and implement all of these suggestions on top of ToME's existing system.

ToME comes with a great roguelike engine by the way :)

[Sirrocco]

And about making addon first. This is ideas forums, it is to discuss some ideas, so do not demand implementation right now

You probably didn't intend this, but the way you described your idea (using phrases like "must" and "must not") sounded a lot like demands to me.

My personal impression (and I believe, but do not know, that this is shared by a number of people here) is that this idea will cause more damage than it fixes, in a few different ways. We am willing to be convinced, but arguments of the variety you have given thus far are unlikely to convince us. If you wish to convince us, someone will need to produce a mod, in which to demonstrate these changes in action. If you don't care enough to do it yourself, you are unlikely to find someone else who cares enough to do it in your stead - it looks like a significant mount of work. Thus, if you want to push this idea forward, you are probably going to have to write a mod.

[Doctornull]

Thus, if you want to push this idea forward, you are probably going to have to write a mod.

This doesn't sound like an addon that would take a lot of special expertise to start, either: it'd just be a lot of tedious work.

You could do part of what you want using superload for every talent tree's .lua file, something like:

superload/data/talents/spells/arcane.lua

Code: Select all

local _M = loadPrevious(...)

Talents.talents_def.T_MANATHRUST.require = spells_req1
Talents.talents_def.T_ARCANE_VORTEX.require = spells_req1
-- ... etc.

Good luck!

[qui]

I think total overhaul is unnecessary. On the other hand, I agree with some of the points.

I see no reason for the trees to have exactly 4 talents each, wouldn't have any problem with some having more or some having less. I don't think that "hey I have 5 good talent ideas for this tree, need to scrap one" or "I have 3 good talent ideas for this tree, but can't release it unless I get another one" should be issues.

But the current system is what it is, and probably a better approach would be to suggest improvements/replacements for the talents that are "useless". After making sure that they really are. I know that some are for me, but that's not to say someone didn't find a way to benefit from them.