7 replies to this topic

[banana peel]

An overwhelming amount of economy related complains kinda overshadowed the core aspects and goals of the upcoming Skill Tree revealed by Chris and Russ. And that is dangerous. Being released at March, 21st, the Skill Tree will affect the gameplay in the long run a lot more than economy flaws that are easily fixed. To say it shortly, if you have problems with game mechanics, you would not enjoy the game regardless of how many mechs you are able to master.

So moving to Chris's pitch (and partly comments of Russ). Some aspects of the tree we saw when we were leveling our first mechs on PTS shaped up to be the pillars of the new system that PGI is not ready to abandon:

• Give-and-take concept featuring the rule "progress through secondary nodes to get the best nodes"
  
• Pathway progression through nodes that rewards a player with cheaper "side" nodes and - in theory - encourages to collect several related skill sets which will result into a "role" skillbuild
  
• Not linear or tier execution of the task, but tree-looking one

Whether some of us like it or not, these pillars are not going anywhere. And it partly explains why the huge amount of constructive feedback (including Solahma's tree) was not considered. I've also spent a week building a workable example of the Skill Tree which, as i see now, is not even close to what PGI wants.

But here is the deal.
The current state of the Skill Tree has flaws and doesnt really embrace goals stated by PGI. Two main goals shared by the community and the company are: provide customization and roles of mechs. And for now regarding these two roles:

THE NEW SKILL TREE ADDS NOTHING TO THE GAME

Simple example. If i want to build a damage dealer support mech that relies on uacs, i can heavily invest into the firepower and get: -7% cooldown, 10% velocity, 15% range, +16 magazine capacity, -5% uac jam chance, -8% heat gen for 40(!) nodes, NOT including nodes inbetween. Any sane player will trade all these bonuses for seismic+radar deprivation and some more points to progress to the speed tweak.

You can build these examples for every weapon specialization or sensor specialization. Tiny bonuses for a huge cost.

The Tree is HIGHLY disbalanced. Most of numbers are too low. Minmaxing it out will be a very easy task. No customization or specialization so far.

In fact, as for now, the Tree even takes some customization away: many underperformers are seeing further reduction of their quirks, which leads to a further gap between them and go to mech variants.

So what can we do now? We can continue to moan about the economy, we can dream of our perfect skill tree, we can quit quietly and shelve the game until better days...

Or we can try to provide reasoned suggestions that do not contradict PGI's pillar items and at the same time help to achieve our shared goals.

Some quick suggestions on top of my head, to start with:

• To make the presented tree balanced and also provide customization and specialization:
  1) Most of node values can be dramatically increased, including multiplying most of firepower branch values by 2 (laser duration and range by 1.5), vastly bigger values for secondary nodes across the tree (such as hill climb) and significant buffs to target info and sensor range for sensor specialists.
  2) Maximum number of nodes can be reduced to 65-70 including free of charge 10-15 points for each new mech (to lower the gap and make every new mech unique from the start). This will provide specialization and embrace the give-and-take concept.
  3) Node cost can be respectively increased to get the overall cost of a mastered mech in line with the current skill tree state.
  
• Severely cut ingame values of arm pitch and heavily increase arm pitch node values to make these nodes a noticeable addition to mechs with arms. Otherwise they can be just replaced by hollow nodes, that would look more fair.
  
• To balance the underperforming mechs, their quirks must be restored AND they can get more maximum skillpoints based on how many quirks they have already.
  
• To maintain the build experimenting part of the game, respec must not be charged with XP. Instead, a minor c-bills price (150k for full respec or 10k for one node) would work. Also there can be created a tool of saved skillbuilds: each mech can buy up to 2 additional slots for saved skillbuilds for c-bills and even more but for mc price.
  
• I would love to see more links and nodes rearrangend for the Mobility, Survival and Operations branches to provide more possible paths. Sensors branch looks a lot better for me in that sense.

Looking forward to your suggestions, guys. I know, many of you are pissed off and exhausted. Me too. But we heard the stated goals during the podcast. We were asked of constructive feedback and now we know what basis we must build it on. We might as well try this one last time.

https://www.reddit.c..._summary_where/

[soapyfrog]

I like 95% of what you say it makes good sense. I agree I don't think PGI will change their mind but they could maybe be shepherded towards a more manageable version of what they have proposed.

banana peel, on 11 March 2017 - 11:06 AM, said:

• Severely cut ingame values of arm pitch and heavily increase arm pitch node values to make these nodes a noticeable addition to mechs with arms. Otherwise they can be just replaced by hollow nodes, that would look more fair.

This I think would be a mistake. I don't think a baseline unskilled mech should be so unwieldy, and to gimp unskilled mechs is to make the NPE even more miserable.

Skill node costs are still too high. In any case, if they must charge they should charge for skill POINTS not for skill NODES. First of all it would fix the respec cost problem, second of all it would provide a consistent cost for the skill tree across all mechs, instead of wildly variable one based on how much experimentation you like to do.

Lastly I wholeheartedly agree that XP cost for respecs must go. A nominal c-bill cost is preferable to xp, but really respecs should be free and it makes no sense at all why it would cost anything.

[QuePan]

yeah the main flaw to the skill tree outside of the costs is the major unbalance between a unskilled mech and a skilled mech , compared to a unskilled mech and a skilled , moduled mech we have now .


they really need to make the Base values more , and more diminishing returns on the tree to balance it out .from my experience the new system doesn't create roles , experimentation or any form of diversity. it locks you into a build in any given mech it more promotes boating to get the best out of the system . and certain mechs will suffer if trying to use this system

[Dracol]

Ok... everytime someone says "give underperforming mechs more skill points" I facepalm.

Y'all want to impose an xp/cbill cost onto the least desired mechs to bring them up to equivalent power level. Why? What will it acieve by giving a Vindicator a higher xp/cbill cost to fully skill out achieve? I'll tell, it makes it even less desirable than before.

Quirks haven't been eliminated. Underperformers are still recieving quirks to improve viability. Yes, some of em have been toned down / removed / moved to base stats. But the doom and gloom squad blow it out of proportion how much they have been affected.

[banana peel]

Dracol, on 12 March 2017 - 08:39 AM, said:

Increasing number of skillpoints is not the best way, as it increases the "cost" to fully skill the mech, yes. Better way is to increase values of nodes, of course. But that would require a lot more balance work regarding every tree of every mech variant. PGI made it clear they are not ready for this amount of additional work.

Second paragraph i cannot really comment, as both my hands are busy facepalming. For everyone that understands the game mechanics you sounded like "1+1=3".

Edited by banana peel, 12 March 2017 - 02:23 PM.

[Znail]

I disagree with that the economics is secondary to the design of the skill tree. Even if we assume that the new skill tree doesn't add anything of worth to the game so is that still fine as that is far better then taking away from the game. The economics of the skill tree is such a major issue that it's quite able to cause the game to end in a near future. Compared with that so is the merits or demerits of the actual skill tree not very significant.

[Dracol]

banana peel, on 12 March 2017 - 02:23 PM, said:

Increasing number of skillpoints is not the best way, as it increases the "cost" to fully skill the mech, yes. Better way is to increase values of nodes, of course. But that would require a lot more balance work regarding every tree of every mech variant. PGI made it clear they are not ready for this amount of additional work.

I completely disagree that increasing values of nodes is a proper way to balance mechs. Its simple math really. If a mech requires investment into specific skills in order to be on par with other mechs, then there is a cost associated with reducing ones disadvantage just like the additional node proposal.

Quirks are still there. They are still being used for balancing and just like live, all mechs get access to the same skills (exception being the IS/Clan node bonuses). The skill tree should have zero input on balancing individual chassis. It really shouldn't have been to use to help balance IS v Clan either.

Edited by Dracol, 12 March 2017 - 04:30 PM.

[banana peel]

Dracol, on 12 March 2017 - 04:28 PM, said:

I completely disagree that increasing values of nodes is a proper way to balance mechs. Its simple math really. If a mech requires investment into specific skills in order to be on par with other mechs, then there is a cost associated with reducing ones disadvantage just like the additional node proposal.

Quirks are still there. They are still being used for balancing and just like live, all mechs get access to the same skills (exception being the IS/Clan node bonuses). The skill tree should have zero input on balancing individual chassis. It really shouldn't have been to use to help balance IS v Clan either.

Ok, you are doing it again: one sane paragraph and the other one - crazy. Yes, bigger node values for one variant imply a "cost to match" the better variant. It's not perfect, but in the end mastered mechs will spend equal amount of c-bills and have equal power. What's wrong with that?

Quirks are there? Quirks are reduced, the most hard cuts - cooldowns on mechs that DIRECTLY rely on a high cooldown to be SOMEWHAT viable. Why in the world the Skill Tree shouldnt be used to help the balance?

Sorry, but i have to stop replying to you until you get some reason behind your words.