GURPS Power-Ups 10: Skill Trees

[ericthered]

Originally Posted by Donny Brook

I do hope they do Techniques soon. That's one of the best innovations the rules have had in 4e, very elegant and useful, though it could use some tweaks to make them more attractive for the CP cost. I would definitely want to see a bunch of new ones and/or building guidelines for non-combat-skill Techniques.

I know psi-wars basically halves all the technique costs and people seem to love that.

[Donny Brook]

Originally Posted by ericthered

I know psi-wars basically halves all the technique costs and people seem to love that.

Interesting. In my games I've implemented the low-hanging fruit of eliminating "Hard" techniques. I also let characters purchase Techniques after they reach campaign skill caps.

[Varyon]

Originally Posted by Donny Brook

to produce a CP cost structure I find too steep

This is the big sticking point for me, at least as an immediate knee-jerk reaction. But, as it turns out, for a rules-hacker like myself (and make no mistake, the rules-hackers are explicitly the intended audience for this supplement), it's actually rather easy to adapt the concept but abandon the default static costs to have the pricing function more like the default skill system (where the first few points in a skill have a larger impact on skill level than later ones - Default+4 for [1], Default+5 for [2], Default+6 for [4], and each additional +1 is [+4]). I'm hesitant to actually post the adaptation, however, as it kinda steps on the toes of the book itself.

Originally Posted by Donny Brook

likely to reduce skill diversity on character sheets, pushing the rules toward character 'classes' (which were one of the things I despised about TOG).

I don't think it really pushes toward character classes, but it probably is likely to reduce diversity a bit (in favor of characters all having much wider ranges of capability). Essentially, it encourages picking up several Trunks related to your character concept and building those up to an acceptable level, then potentially specializing in a single Branch or Twig of those Trunks (you can technically have two Branches at a lower cost than increasing that of the Trunk, but it's so close that I'd say boosting the Trunk is a better option), and maybe picking up one or two Leaves for that Branch/Twig (the only exception is if you have Leaves that can stack).

[Kallatari]

Originally Posted by Varyon

This is the big sticking point for me, at least as an immediate knee-jerk reaction. But, as it turns out, for a rules-hacker like myself (and make no mistake, the rules-hackers are explicitly the intended audience for this supplement), it's actually rather easy to adapt the concept but abandon the default static costs to have the pricing function more like the default skill system (where the first few points in a skill have a larger impact on skill level than later ones - Default+4 for [1], Default+5 for [2], Default+6 for [4], and each additional +1 is [+4]). I'm hesitant to actually post the adaptation, however, as it kinda steps on the toes of the book itself.

I'm with Varyon on this. The apparent steep CP cost to get equivalent levels of skills if you simply convert is, at the moment, a knee-jerk reaction. I haven't run all the numbers, and haven't redone all of my templates to see how it actually change all of the costs On to be honest, I expect the variances to be across the board with possibly a small number of templates actually being cheaper, most having a cost increase, and another small number having a large cost increase.

To me, that's just affects figuring out how to set up the starting parameters of the game. The actual in play application of the Skill Tree is, in my opinion, super easy to implement. I love its flexibility is deciding the scope of your skills.

So GURPS Skill Trees is playable out the gate, but it is just not "backwards compatible" in that you can't convert an existing campaign (well, not without a lot of work that probably wouldn't be worth it). You just need to start with it, and see how the point costs go.

As a rule-hacker myself, I'm also looking at, among other options, increasing but levelling off cost, probably something like [2], [4], +[8 x level] for Trunks (or what I prefer to call Skill Groupings) and [1], [2], +[4 x level] for Branches (i.e., Skills), while starting at Attribute -2 as nudj suggested as a possible house rule. I'd leave Twigs/Specializations at [2 x level] and Leaves/Techniques at [1 x level]. I think this would significantly narrow the changes in cost of existing templates and make them manageable with minimal tweaks. There will still be some changes in costs, but I think it will be close enough to not matter too much.

In addition, with that proposed change, existing NPC and such only uses Skills (or Branches) would end up having the same cost and thus be fully convertible back and forth... they just wouldn't be optimized in possible point savings (because they could merge multiple Branches/Skills into a single Trunk/Skill Grouping for less points). But there are plenty of NPC who aren't optimized (i.e., dropping skills to increase attributes), so that's not a problem. So this houserule option is currently leading as my favored one... but I'll wait until I see all the cost implications before deciding.

And that's not to "fix" how Skill Trees work in of itself. That's just to make it more backward compatible with the current skill structure.

Finally, I also truly appreciate the math that's provided, as it lets me understand the impacts of any changes I make in my own houserules.

[DouglasCole]

One thing about this method that combined with extensive use of proper meta-traits for commonly grouped clusters of advantages/disadvantages (like "Flora" that we used in Garden of Evil) is that it could potentially vastly reduce the required size of character/monster stat blocks.

GURPS labors some what under the burden of what I call over-traited actors. The entry for Zombie, for example, has 24 different traits as presented in DFRPG Monsters.

What you really want to do is be able to say:

Zombie
Animated Dead Thing

ST 20 DX 12 IQ 7 HT 12
Will-10 Per-12 FP N/A
High Pain Threshold
Damage Resistance 2
High Pain Threshold
Vulnerability (Fire)
Melee Combat-12

and more or less call it a day.

The Animated Dead Thing meta-trait would subsume

Appearance (Monstrous)
Bad Smell
Doesn't Breathe, Eat, Drink, or Sleep
Immunity to Mind Control and Poison
No blood, brain, eyes, or vitals
No sense of smell/taste
Temperature Tolerance 5
Unhealing
Unliving

While one could argue "yes, this" or "no, that other thing" on a nitpick basis, the key bit is to be able to present a creature simply, with the understanding that a lot of the details are going to be only important off-screen.

There's a lot buried in the "No this or that" in Animated Dead Thing, but what it really means here is that there's really nothing here but to swing for the Hit Points, with some exceptions for lopping off a leg to have a mobilty kill.

The other part of having a trait list two dozen strong is that it means some really key bits get washed away in the Big Lists of Everything. I've seen important traits that are buried in the list get overlooked, and being able to only highlight the really important stuff (Altered Time Rate, Extra Attacks, Injury Reduction, Supernatural Durability are all things that spring to mind as radically changing the threat level in combat come to mind) is handy when writing things for public consumption.

So while I might cast a few nits to the wind on the specific implementation of Skill Trees ... the overall end result for me is going to be super worth it.

Mission X, currently in development, is going to use this one way or another. I've been talking about Skill Families, Skill Clusters, Skills, and Specializations/Techniques/Perks for a bit, and this provides an already-published stepping off point.

[Donny Brook]

Originally Posted by DouglasCole

...
What you really want to do is be able to say:

Zombie
Animated Dead Thing
...

and more or less call it a day.

The Animated Dead Thing meta-trait would subsume

Appearance (Monstrous)
Bad Smell
Doesn't Breathe, Eat, Drink, or Sleep
Immunity to Mind Control and Poison
No blood, brain, eyes, or vitals
No sense of smell/taste
Temperature Tolerance 5
Unhealing
Unliving

While one could argue "yes, this" or "no, that other thing" on a nitpick basis, the key bit is to be able to present a creature simply, with the understanding that a lot of the details are going to be only important off-screen.

I have recently been reminded of how much I really dislike stat blocks using Metatraits instead of listing out the contents. It puts me to the effort of looking up the metatrait separately from the creature which is an extra flip through the rule books I would prefer to avoid.

[dataweaver]

Originally Posted by Professor Phobos

EDIT: Also, it occurs to me that if you wanted to reintroduce hyper-competence Wildcard-style super skills, you could call them "Forests" as an optional level above the Branches...

That would be "Trunks".

[Professor Phobos]

Originally Posted by dataweaver

That would be "Trunks".

Ah, fixed. Yes.

[dataweaver]

Frankly, I think Trunks are already the right level for Wildcard-level super-skills. So does PU10.

[jackcelso]

25 pages not a single full example, with cost and levels and correlations to attribute.

if spend X points in that Y trunk with an IQ of 12 will have a skill of Z