6 replies to this topic

[BearFlag]

No, I don't mean the proposed Skill Tree is another Long Tom.

What I mean, after watching the podcast, is that PGI clearly circled the wagons around their idea and offered arguments to keep the Indians at bay. It reminded me of LT at the July Round Table. There a player panel kept on topic and rejected PGI's argument that LT has to have a big in-game effect while pointing out that LT was having deleterious effect.

This time, all four participants defended in unison the current versus a cleaner linear system. We got uncountered monologues of why linear is bad.

In fact, the goals of the current system are either not achieved or can be duplicated in a linear system.

First, clarification. What exactly is being "traded off?" It is high value skills against other high value skills with the mechanism being SP exhaustion. Both systems can reach this.

Some of the defenses offered:

#1) A linear system allows "cherry picking." That first skill or two, yes, it does. But with 91 SPs, so does the current system. If I want full Speed Tweak, I get it - along with numerous other buffs wanted or not. Any way, isn't picking what you want the very purpose of choice?

#2) A linear system allows players to get everything important. No, it doesn't. Tree depths and SP exhaustion prevent this.

#3) The current system allows role specialization. There are no roles so, no, it doesn't. Alternatively, it does allow for role-like specialization. Take your pick. But, as a comparative argument, it does this no better than a linear system and arguably less efficiently.

#4) A lot players like the PTS system and there's a group that doesn't. If the forums and forum polls are any indication, this is a serious misread of the community. It looks like 70-80% prefer a linear and/or tiered system. Not quite LT unanimity, but high.

#5) It makes no sense to dislike the gate nodes because they're still buffs. Point missed. SPs to skill a mech create an SP budget for that mech. Having unwanted and mandatory expenditure to reach something wanted "feels" wasteful, if not gamey. Not the end of world, but it is a little irritating.

What does a linear system seek to correct?

#1) Size. It would seek to reduce the total number of nodes and reduce SPs. This makes it more manageable.

#2) Simplicity. The divisions and progression are clear. Paths are straight.

#3) Intuitive. Because of #2 there's no scanning through jumbled tree of unrelated nodes. The Range Tree is the Range Tree is the Range Tree. Not other things.

#4) Less gating. The current ST uses gate/buffer nodes to compell expenditures which in turn eventually reach a point of SP paucity. The is wasteful and unnecessary since the same can be achieved without gate nodes.

Basic design concepts of a linear system:

A linear system achieves virtually identical end results as the proposed system. Highly desirable skills have deep trees (and possibly ramped). Less desirable skills have shallow trees. SPs are reduced. Doesn't get much simpler than that. If Speed Tweak is six or eight nodes deep and there are only 40 SPs available, you have some hard choices ahead. Maxing something out starves other potentials. If Hill Climb is one or two nodes deep, it may be worth grabbing for a mech focusing mobility.

Like the PTS system, a linear system seeks to create trade-off through SP exhaustion. Properly done, it accomplishes this with greater clarity and efficiency.

[soapyfrog]

Well said.

Even if PGI wanted to keep this structure the same thing could be reproduced with 1/4 the nodes and no filler. A lot of the buffs that are in the tree don't even really need to be there, and no one really cares about them. Uninteresting nodes should just be automatically included in the chassis.

Tighter trees with interesting bonuses on offer would force hard decisions and maybe real sacrifice. As it is now I can max out agility and survival, get the majority of the firepower bonuses for my boat (because that is what this tree rewards), and still have points leftover for sensors, ops and misc.

[Skribs]

What got me is they said the first skill tree on PTS was a linear tree, and unless there was a previous PTS I wasn't part of, Skill Tree PTS1 was definitely not linear.

A linear tree with increasing cost per node, and more expensive 1-off skills for things like Advanced Zoom would be much better. I've done a mockup like this, but the basic idea is:
1) You convert Mech XP into Skill Points to use, up to a skill point cap
2) Abilities like Fast Fire would be 5 points that would cost 4, 5, 6, 7, and 8 skill points for Ranks 1-5. This way focusing on a specific boost comes with additional cost. (As opposed to 1-5 points, in which it is much easier to grab all the rank 1 and 2 abilities)
3) Things like Hill Climb, Advanced Zoom, or Radar Derp would cost something like 4, 15, or 35 points (respectively) to showcase the power of that particular ability
4) Things of less power or use, i.e. jump jet abilities, may be 3-tiered instead of 5.

You might end up with *more* skill points this way, but it would be easier to use them all. It wouldn't let players just "cherry-pick" everything, because you have to invest more heavily to get what you want. It would give players more choice in how they level up their Mech.

As it is right now, it feels like PGI is giving us "freedom" in letting us pick what path we choose. The problem is they're following the mold of older MMOs like WoW's old talent tree, where you might slog through filler talents to get to something juicy like Mortal Strike or Bladestorm. If we were getting something like a cooldown where we generate no heat for a period of time in the operations tree, a 50% damage cooldown in the firepower tree, a defensive shield that activates when we lose a component in the defensive tree, or a teleport in the mobility tree, then the filler skills would feel worth it. We're not building up to something big and game-changing, though, we're building up to something that is better maths or a slightly more relevant buff. Yay.

Then there's the sensory tree. If you go down that tree you pretty much get Seismic, Radar Derp, Target Decay, and a few others (I think 360 targeting, sensor range, and target info gathering), and have easy access to Advanced Zoom if you want it. That means just in that tree alone, you could have access to a lot of modules. LRM boats are going to have easy access to all of those, as are snipers, so how exactly will Radar Derp and Target Decay work with each other?

It seems like PGI has created roles in their mind without understanding how the modules work to counter each other in-game, without understanding their own meta, and without understanding what the players actually want. This system has some improvements over live (removal of rule of 3 is the big one), but comes with a lot of negatives. I agree I got a circle-the-wagons vibe from the podcast, but the misinformation in it makes me question the skill tree even more.

[Skribs]

The thing is, they could put an appropriate cost for the various skills, i.e. hill climb could be cheap in SP, while seismic or radar derp could be expensive. A linear tree could be built such that it benefits multiple weapons at once (i.e. you could have cooldown, cooling, heat threshold, and then have one for laser duration + projectile velocity, and another for spread + jam chance, which would be more all-encompassing).

While there are different roles for different Mechs, I do not feel like the trees as-is really give you the option for much of a role, any more than the base equipment you pick in the MechLab.

[BearFlag]

Skribs, on 13 March 2017 - 07:23 AM, said:

A linear tree with increasing cost per node, and more expensive 1-off skills for things like Advanced Zoom would be much better. I've done a mockup like this, but the basic idea is:

Yeah. This would make for a cleaner tree than just depth and one point per node. However, in any case where increments are needed, depth could be brought back.

[BearFlag]

Count Jester, on 13 March 2017 - 09:11 AM, said:

One thing not addressed by the OP is the desire for some players (and apparently PGI) to make it so that taking multiple weapon system types isn't made stupid by needing to invest lots more SP than boating a single weapon type. Thus, they think they're doing something useful by giving you additional nodes you wouldn't otherwise get if you were just paying more per node, so that you might take other weapons as well.

You're right. I was aiming for a general defense of linear. There are several proposals for linear out there now and I wanted to defend them all as viable (with refinements).

PGI took a step with the boating/diverse problem by combining cooldown, range, heat gen. They left duration and velocity separated - presumably because they're not universal like, say, range. They could combine these two into a single progression. That is, duration/velocity would share nodes and thus buff all weapons (except flamer). This would give you four sets of universal buffs and leave the question of how to handle further weapon or weapon-type specializations.

Edited by BearFlag, 13 March 2017 - 07:42 PM.