Jump to content

Constructive Criticism On The Tree Layout.


14 replies to this topic

#1 Excalibaard

    Member

  • PipPipPipPipPip
  • Overlord
  • Overlord
  • 169 posts

Posted 26 April 2017 - 05:18 AM

Hi PGI,

I want to give feedback on a few key points that are holding back the appeal of the new skill tree, regardless of individual node balance or economy:

  • Clutter and Complexity
  • The inverse law of choice: Too many skill points!
  • Mech identity: Quirks in another way.

TL;DR:
1. Change nodes into something more compact, where people can easily find the bonuses they're looking for.
2. Don't have too many nodes, as it doesn't actually improve the min-max capabilities or enhance the choices between stats in a meaningful way on such a small increment between each point.
3. Preserve Mech Identity and balance across geometries (currently governed by quirks) by making separate trees easier or more difficult to spend skill points in for certain mech chassis or variants, or having some skill nodes unlocked by default.

1. Clutter:

The current skill tree (and the ones before it) takes up too much space. This makes it very hard to plan your build, and an already high-information game even more difficult to grasp for new and returning players.

Instead of the web, you can have clear blocks each representing a different stat. Some stats are more powerful or specific than others, and can therefore 'hidden' behind an investment wall, where you first need to spend a certain amount of points in other stats before you get access. This has the same philosophy, but is much more comprehensible as people can find what they want easier.

Here's a quick example of the firepower tree with such a system (not balanced to the current amount of points, just as indication). You can mold it into a nice shape related to the Skill Class if you like, but it's a lot better to comprehend:
Spoiler

Even if this isn't what you're looking for, there have been many suggestions regarding the layout, so please listen to them and don't stubbornly implement an excess of hexes.

2. The inverse law of choice.

There is a thing as too much choice. The more steps you include, the less significant a certain 'choice' becomes and the lesser the feeling of actually making a choice becomes.

Instead of giving incremental 1% increases that people won't notice, it would probably be better to have a few distinct choices: you get Cooldown OR Heat reduction, no middle grounds. That makes people think about their builds: On this PPC boat, do I want less DHS and go for the heat reduction skill, so I can invest in a better engine? Do I mount more DHS, and then take the cooldown option for more DPS? Separating these choices into tiny 1% increments doesn't change the overall choice, but DOES increase the clutter (though, slightly mitigated with a more space-efficient skill system)

An example of this can be seen in 99% of League of Legends Masteries. The option to divide 5 points between two distinct skills will end up with all 5 in one of the two, never split across both, as it makes no sense from the mix-maxing standpoint that many players here use. One of the two skill is better than the other for build X, so it's maximized for build X. This is of course far from an optimal system, but it serves as a nice example as to why I think tiny increments will not work in this system.
Spoiler


More comparable to the current system is the extensive Path of Exile tree. This is extremely complex and huge, but there the significant nodes are clustered into clear super-nodes, or split into two distinct choices across a certain path. The larger path length also makes it easier to recognize where each point leads. This makes it MUCH easier to navigate than your current system for MWO. While this is a system based on increments, they're much more impactful than they can be in MWO as it will eventually add up into something like 400% weapon damage, therefore their influence is much more noticeable.
Spoiler


3. Mech Identity: Quirks in another way

While it's good to see quirks in their current form go, it's also a little bit sad. Some mechs were only really worth running for their quirks, with otherwise poor hardpoints (amount or locations). Also, some mechs had great lore-background for their quirks, which I'd hate to see removed.

It's weird that a Victor is as mobile as a barn-door Awesome for example under ther new decoupling and unquirkening. Or that the extremely mobile Dragon will barely be more maneuverable than the Cataphract. The Mech Identity inherently suffers from this change, and will make the different designs even more similar to pilot and put even more focus on designs with high hardpoints.

As mech-specific quirks are on the way out, I suggest that instead you have some skill points already unlocked and equipped for certain variants. With the suggested system above you may even increase their maximum in a certain skill to highlight specific strengths, that people can use, IF they want.

Another way to increase mech identity is to give each mech only certain skill classes of the global skill tree that they 'specialize' in, and can spend only 1 skill point per node. Other trees that fall outside their expertise will cost 2 skillpoints to equip a single node.

For example, the Stalker could be 'specialized' in firepower to become a very good weapons platform, but have to invest significantly to get more mobility. The Kodiak and Dire Wolf could also be made into mechs with specific different roles, without the Kodiak simply outperforming the Dire Wolf. The Kodiak could focus on Mobility while the Dire Wolf enjoys cheaper Armor and Firepower nodes. These specializations could even be variant-specific, where a more counter-intelligence focused variant could receive bonuses to auxiliary trees, or a jump jet capable variant can invest more easily into jump jets, if they want.


I hope that you'll listen to these ideas, or give some constructive feedback.

Edited by Excalibaard, 27 April 2017 - 04:56 AM.


#2 Excalibaard

    Member

  • PipPipPipPipPip
  • Overlord
  • Overlord
  • 169 posts

Posted 26 April 2017 - 08:49 AM

Anyone else with feedback on my views here, please do leave a comment :) I appreciate all views.

#3 Mycroft000

    Member

  • PipPipPipPipPipPipPip
  • The Partisan
  • The Partisan
  • 511 posts
  • LocationArizona

Posted 26 April 2017 - 08:53 AM

You're completely correct. I've gone through all of the trees and I think the optimal choices use up 87/91 before you even start considering what to choose as "electives".

#4 Jikil

    Member

  • PipPipPip
  • The Covert
  • The Covert
  • 83 posts

Posted 26 April 2017 - 09:43 AM

I think your spot on but I don't like your idea regarding quirks. Typically the meta will push a lot of mechs out of the way and if there is a certain best variant or chassis people will use those and the others will be neglected. Look at the cataphract, in its day it was one of the best ppc jump snipers but now it gets benched for warhammers and night gyrs.

I do like the idea of skill points automatically unlocked so it can kind of guide you as to what is going to work best.

At the end of the day though any kind of quirk changes that aren't straight up buffs are going to hurt someone's feelings because they are attached to that mech.

#5 Excalibaard

    Member

  • PipPipPipPipPip
  • Overlord
  • Overlord
  • 169 posts

Posted 26 April 2017 - 10:21 AM

View PostJikil, on 26 April 2017 - 09:43 AM, said:

I think your spot on but I don't like your idea regarding quirks. Typically the meta will push a lot of mechs out of the way and if there is a certain best variant or chassis people will use those and the others will be neglected. Look at the cataphract, in its day it was one of the best ppc jump snipers but now it gets benched for warhammers and night gyrs.

I do like the idea of skill points automatically unlocked so it can kind of guide you as to what is going to work best.

At the end of the day though any kind of quirk changes that aren't straight up buffs are going to hurt someone's feelings because they are attached to that mech.


There will always be a 'most meta' mech, quirks or nah. I don't think there's harm behind lore-supported (or performance supported) advantages to certain mechs. Not necessarily to make them compete with meta mechs, rather to retain a sense of individuality for the chassis while it's running a build that is available across the board.

For example, Firestarters and Wolfhounds are very similar. With quirks to their performance in any way, or changing their 'specialized skill trees', they can be separated a little bit further, for example making the firestarter fast and the wolfhound tanky (both relatively). Or giving the Firestarter some flamer points for free, and the wolfhound some large laser points.

Edited by Excalibaard, 26 April 2017 - 10:22 AM.


#6 Lazor Sharp

    Member

  • PipPipPipPipPipPip
  • Stone Cold
  • Stone Cold
  • 353 posts

Posted 26 April 2017 - 10:50 AM

While I realize the previous skill tree/modules was a place holder, and you always meant to revamp it at some point, I don't like how you implemented the skill trees at all!

PGI, You had to hide Speed Tweak at the very bottom of the Agility tree, and to me, Speed Tweak is the most important mobility skill! Everything else comes after that..... Some nodes are not even applicable to some mechs, like a mech without arms, or weapons in the arms, don't need arm quirks, a mech with JJ's, don't need hill climb, etc.... Very, very bad design, to force me to have nodes that I have no need for/want, and will not help my mech in the least, just to get to the nodes do I need/want...... The unneeded/unwanted nodes, should be able to be used elsewhere in the trees, to help get the skill feature combination I want!!! Its hard to get any other skills up to snuff, when you have to use so many nodes JUST to get to Speed Tweak........ Plz fix this before it goes live.....

Edited by Lazor Sharp, 26 April 2017 - 11:33 AM.


#7 Cato Zilks

    Member

  • PipPipPipPipPipPipPip
  • Hero of Marik
  • Hero of Marik
  • 698 posts
  • Twitter: Link
  • LocationPrinceton, NJ

Posted 26 April 2017 - 11:08 AM

Points 1 and 2, awesome.

# 3 not so much. So, Russ has explicitly ruled out mech specific trees. It takes too much time on their end. While quirks aren't elegant, most are needed.
A victor needs mobility buffs, awesome needs ppc buffs, etc. These don't need to be huge, but they cannot go.

#8 Cementi

    Member

  • PipPipPipPipPipPipPip
  • Urban Commando
  • Urban Commando
  • 779 posts

Posted 26 April 2017 - 11:43 AM

I am ok with the current system however I would much prefer a linear skill tree. I feel it would also be far easier to balance as well.

Jump jet tree is still far to weak. Needs to be made far more potent. If your worried about poptarting make the recticle shake both going up and down and then put a few nodes in the jump jet tree to reduce the shake when the jets are not on.

I think quirks to give mechs identity need to stick to durability, sensors and mobility of all types. Weapon group or weapon specific quirks just end up pigeonholing mechs into certain weapon systems which I am not a fan of.

Bottom line is I am ok with it as is, though there are areas that could be improved. I also realize that this system is going to have to have ongoing adjustments. I hope that those are done monthly and incrementally instead of for months and then nerfing into oblivion. Nerfs are nessesary but I would prefer they apply them with a tack hammer and not a sledge hammer.

I really do think they have to give up on the skill web. I know they are grasping at holding on to it because they like the idea but I feel they are expending to much effort into trying to make it work. If you feel Radar Derp needs to expend 15 points or whatever it is then make a nice straight line with 15 points to get the max Radar Derp (I still think max should be only 80% but thats me). Less desired things like hill climb, which is still somewhat usefull can have only a few points. This would also make it FAR more easy to balance as they can easily tell which things are being taken because they are wanted rather than trying to figure out what is being taken only because it is linked to something people want. Then they can buff the values of under used nodes to make them viable options. With the current system if you do that then you are giving the path to something that is highly sought after an indirect buff and that is the last thing something like Radar Derp needs.

#9 Excalibaard

    Member

  • PipPipPipPipPip
  • Overlord
  • Overlord
  • 169 posts

Posted 26 April 2017 - 03:49 PM

View PostCato Zilks, on 26 April 2017 - 11:08 AM, said:

Points 1 and 2, awesome.

# 3 not so much. So, Russ has explicitly ruled out mech specific trees. It takes too much time on their end. While quirks aren't elegant, most are needed.
A victor needs mobility buffs, awesome needs ppc buffs, etc. These don't need to be huge, but they cannot go.


These are not mech-specific trees. They're the same trees, but some mechs need to spend 2 points for a node instead of 1 point in a node, depending on which trees they specialize in. Basically their archetype makes it easier to get deep in some trees, while not ruling out other builds. It's less work than quirks per variant.

I agree that the quirks and/or some form of nudging chassis into archetypes outside their weight (class) is necessary to keep the game balanced and fresh.

Edited by Excalibaard, 27 April 2017 - 03:53 AM.


#10 Excalibaard

    Member

  • PipPipPipPipPip
  • Overlord
  • Overlord
  • 169 posts

Posted 27 April 2017 - 03:54 AM

View PostCementi, on 26 April 2017 - 11:43 AM, said:

-Snip-


Yes, web is definitely not flexible enough.
Regarding Radar Derp: As long as it can be properly acted against with Target Decay, that would already make it much more balanced than it is currently. Instead of a percentage, it should reduce the decay time by fractions of seconds.

#11 KevinRuddPM

    Rookie

  • Bridesmaid
  • 8 posts

Posted 27 April 2017 - 04:10 AM

I'm really glad to see some actual constructive criticism on this topic, rather than just a bunch of whiners that have no idea how the tree will play out on the battlefield.

I've only just had a go at the PTS and my feedback is similar to the OP. I found myself doing 2 things. I spent a while on the first mech figuring out which way I wanted to go etc, but every mech after that was just a wash of clicking without much thought which brings me to 2 points.
1. I feel like most nodes will just be clicked because they are there or in the way, which negates the choice and planning aspect of the new skill tree.
2. As all the nodes are so incremental, there doesn't seem very little real world difference between switching a few nodes around. This makes me think that perhaps the nodes themselves need to carry more weight and be more meaningful, as well as not having to invest in a completely useless node (like missile crit on a mech with no missile hard points) in order to get something else.

I completely understand the idea behind this new skill tree, and I'm offering my support in trying it out and leaving feedback.

#12 Excalibaard

    Member

  • PipPipPipPipPip
  • Overlord
  • Overlord
  • 169 posts

Posted 27 April 2017 - 03:08 PM

Good point, but as unskilled mechs still need a decent performance on the field, the advantage from skills can't be too big, sadly. What could be done is roughly have a half or a third of the current amount of skill points, and then make the nodes of actual meaningful size (say 5% heat gen reduction).

#13 Jack Shayu Walker

    Member

  • PipPipPipPipPipPipPipPip
  • The God
  • The God
  • 1,451 posts

Posted 27 April 2017 - 03:10 PM

Stop being right! PGI needs to think they thought of it first, otherwise they won't do it >:|

#14 ANOM O MECH

    Member

  • PipPipPipPipPipPipPip
  • Survivor
  • 993 posts

Posted 27 April 2017 - 03:51 PM

It is still way too many nodes. Still don't want to 'pay' or have a consequence by getting garbage nodes that I don't want or need. Would rather pay more for valuable nodes or have choices reduced through the branching off of nodes.

Just too busy and too much.

#15 Excalibaard

    Member

  • PipPipPipPipPip
  • Overlord
  • Overlord
  • 169 posts

Posted 28 April 2017 - 06:17 AM

View PostJack Shayu Walker, on 27 April 2017 - 03:10 PM, said:

Stop being right! PGI needs to think they thought of it first, otherwise they won't do it >:|


I got this information from PGI logic center, i swear!

View Posttker 669, on 27 April 2017 - 03:51 PM, said:

It is still way too many nodes. Still don't want to 'pay' or have a consequence by getting garbage nodes that I don't want or need. Would rather pay more for valuable nodes or have choices reduced through the branching off of nodes.

Just too busy and too much.


Agreed. If people would even pay more in a new non-web system than in the current system, that says something :)





1 user(s) are reading this topic

0 members, 1 guests, 0 anonymous users