7 replies to this topic

[Anatidaephobia]

Hi, much like last time, I’m going to provide some feedback here regarding the current iteration of the skill tree.

First of all, I want to make clear that I’m not against an implementation of a skill tree and would really enjoy the addition of being able to customise/specialise mechs further. I may sound harsh at some points or objecting certain parts, this is directed at the current concept of a skill tree, not the skill tree idea itself.

With that said, let’s jump into the meat of the skill tree; I will divide this up in the following sections :

• News post observations and initial thoughts
  
• The Good : Items which are in a good enough condition and need minor or no improvements The Bad : Items which require changes for improvement
  
• The Ugly : Items which are best described as a “Disaster”
  
• Neutral / No Opinion
  
• Personal thoughts and conclusions

News post observations and initial thoughts

Quote

The Skill Tree marks a major shift from the previous Skill system. At its core, the current Tree you'll see in this build aims to address the following main goals:

Now, the biggest change I can see is more regarding the economic model which includes the keeping of progress, refunds, etc… The skill itself is relatively unchanged, which is part of where the issue lays.

Previously I’ve mentioned that I’ve noticed the “baby syndrome” in this skill tree, this observation still holds true as the main issues were the tree itself were not tackled, with only minor changes on a superficial level.

To start;

Quote

Greater overall give-and-take system for 'Mech skills and specializations.

There was no change compared to previously, once more we have a skill tree system which is designed around filler-choices combined with a flawed structural layer.

Quote

Heavy initial investments in high value nodes, with reduced costs for subsequent investments within the same Tree.

I can’t see where this design goal has been fulfilled. There is no high initial investment, nor is there a reduced cost; everything is split by mostly filler nodes making either the investment not worth the effort, or said investment is a full-on requirement for gameplay.

There is no choice present in the system on offer. It remains a “Find-the-best-general-path”-meta, someone will find the optimal way to fill out this skill tree, it will be copy-pasted from mech to mech with slight adjustments where necessary.

Quote

Addressing 'progression walls' seen under the previous Skill system.

Now, a lot of people will disagree as I’ve seen in some other threads; however I do feel like there has been significant improvement in the addressing of this issue that was previously present, and personally I am content with the work put into the system. My Basic-mechs will remain somewhat basic’d, my master-mechs will remain fully skilled out without an extra price tag added to this.

No longer do I feel as if the entire work I’ve put into this game is completely being taken away, yes, it is different, but it’s not as much of a disaster as the previous iteration was.

The next part of this thread will order items in 4 sections, these will not be chronological, but each sub-title will hopefully be a searchable term in the news post, so hopefully won’t be an issue for readability.



The Good : Items which are in a good enough condition and need minor or no improvements

Quote

Refund Ledger

I. Love. It.
It shows me exactly what I wanted to know, what I’m getting back, how many cbills, MC, global xp, modules, consumables, mech mastery returns, …

We need more of these full-open displays of data! Can we get a statistics screen like this, ingame? So I can see all my weapon accuracy, … statistics in a neat and accessible format?

Refunding consumables, both CB and MC? Amazing.

Refund from modules, cbills/skill points depending on data purchased? I can live with this, I’ve heard concerns from other people, but for me I don’t have an issue. [doesn’t mean it’s perfect, /I/ just don’t have an issue with it].

‘Mech Mastery Breakdown? Yes, I like it.

Quote

Skill tree colours and identification of nodes

Finally, one good thing about this skill tree- It is finally clear which nodes are active, which nodes are available for purchase and which nodes are locked. That said, that’s the only good thing about it.


The Bad : Items which require changes for improvement

Quote

Base Consumable Slots

I’m neutral on the lowering to 1 consumable slot standard on a mech, I’d prefer more testing in a full environment to form an opinion on this. However, I’m opposed to the new maximum number of 5 consumable slots. Yes, this requires a skill point investment, but 5 consumables will have a massive effect upon the game as it is played, and I doubt it will it will be a positive light. This also makes it possible to spend around 200K cbills in a single match on consumables.

Quote

ECM

Now, it appears that ECM is considered a low investment/high reward equipment; I disagree. Not all mechs can equip ECM, it takes a slot, it takes 1 ton for clan; it takes 2 slots and 1.5 tons for IS.

Does ECM require some sort of nerf? Perhaps, I don’t know for sure, but this just seems a bit over-the-top; OR it is because the ECM boost nodes are too much of a hassle to actually get to.

Quote

’Mech Mobility

As I’ve stated in previous feedback threads, this is something which should be handled separately from the skill tree implementation; there are too many variables to account for, making the entire impression rather muddled.

I do agree that this can be another method of balance, which I support, but there are so many changes coming with this skill tree that it’s difficult to pinpoint issues with the desync itself because the entire concept of changes as a whole is too large.


The Ugly : Items which are best described as a “Disaster”

Quote

Quirks

… You are completely and utterly nerfing the underperformers who previously relied on their quirks to have even a slight impact;

I shall include a quick image to detail what I mean here.

There is a complete lack of understanding the role that quirks play for underperforming and bad mechs; there is absolutely no reason to nerf the Vindicator or Trebuchet or Victor, these mechs are already bottom of the barrel.

If you’re nerfing ahead of time for the new technology coming into the game, WHY? You’re going to make these already terrible mechs almost completely unplayable for the duration between the skill tree and the new technology being released.

And most of these mechs have issues that new technology won’t even solve, I have absolutely no idea what kind of logic was used to come to the conclusion to nerf almost every quirked mech; while top performing mechs such as the Marauder IIC and Kodiak will be gaining extra advantages from the skill point spending. Good mechs are becoming better, bad mechs comparatively even worse than they were before. Just take a look at the two listed PDFs, aside from the engine desync quirk removals, almost EVERY weapon-based quirk is getting nerfed. Victor, Vindicator, Dragon, Centurion, …

Quote

The Skill Tree

Yeah, it pains me to put this here yet again, but it’s just not up to snuff; it remains bloated and oversized, barely impactful and a mess to traverse. The same issues still remain from my previous feedback session

The design of the tree does not conform to any of the design goals that were stated; there is no meaningful choice, there are too many nodes with too many skill points with too many small numbers which don’t matter, and there is too much filler.

Having 242 Skill Nodes where you can use 91, is not a good number. Bigger numbers aren’t better, it just needlessly complicates things.

And, once more, any system which has a number such as “91” as an “optimal” number, is fundamentally flawed in its design and no matter what you try with said design, it won’t solve the glaring issues present with it.

Trying to make this work, is both pointless and a waste of time and effort; please go back to the drawing board, it doesn’t need to be complex, it needs to be functional where it matters. Heck, if I were living anywhere close to Canada, I’d offer to do it myself..

The UI is… messy? It’s easy to lose track of things, there are some bug and issues; as an example, let’s play a game : In the following screenshot, how many historical skillpoints do I own on this mech, and how many skill points do I need to buy to get exactly 91? Leave your answer in the comments, because I sure as hell don’t know.

Album detailing the issue is here.


Neutral / No Opinion / Not enough information to work with

Quote

Target Info Gathering

Not enough data to work with, I don't know if this is good, bad, no effect or superficial.


Personal thoughts and conclusions


While I can see the positive effort that was put into the entire refund system, which was one of the main issues of the previous iteration; the main glaring issue that is the skill tree design itself, is still present and is as bad if not worse than before.

Regarding quirks, I don’t think anyone currently in charge of balancing actually realises what role these quirks serve on bad mechs, what the quirk accomplishes and what should be done before removing said quirk. In some cases, removing the quirk is not an option, nor is lessening it because all it will do is make a bad mech into a terrible mech.

Making the quirks matter worse, is that quirkless top performing mechs suddenly gaining boosts. If everything gets the same boost, you can talk that things will remain similar to eachother, however in this case, the result is that the great mechs become amazing, widening the gap between the bad<->underperformers<->good mechs even further. I’m sure that this is not something that is actually intended, atleast, I hope it isn’t.

Tinkering is punished; if you want to swap from a PPC to an LPL, it will cost you around 6-10K experience each time; experimenting is still punished.

The Skill Tree UI, aside from node colouring, is terrible. It’s a mess, it’s overly complex, far too many numbers and repeats and … Yeah, there is not more I can really say.

_________________________________________________


I’m honestly starting to feel burned out on this tree, it’s not changing, the majority of my feedback is still the same as it was for PTS1 and I doubt what I’m saying will actually get considered or reach the right ears.

Edited by Anatidaephobia, 26 April 2017 - 05:20 AM.

[Duilliath]

I find it amazing that PGI has picked out one of the prior complaints from the last PTS and made a significant improvement to it(the transition/refund), though it is no way perfect at this point, while completely ignoring all other complaints about the new system as well as all suggestions for improvement. I have to wonder how many people will be asking for refunds on various pre-orders at this point, and how many more will once PGI forces this horribly broken system out next month.

[Anatidaephobia]

Duilliath, on 26 April 2017 - 06:47 AM, said:

I find it amazing that PGI has picked out one of the prior complaints from the last PTS and made a significant improvement to it(the transition/refund), though it is no way perfect at this point, while completely ignoring all other complaints about the new system as well as all suggestions for improvement. I have to wonder how many people will be asking for refunds on various pre-orders at this point, and how many more will once PGI forces this horribly broken system out next month.

Pretty much, yeah.

I like what they've done to try and tackle the "MY PROGRESSSSS"-problem, but there are also other problems which have just been left by; some having grown even worse than they were. :/

[C E Dwyer]

I was going to write my own opinion of the skill tree, and what is right what is ok, what has failed to be adressed and what is utterly terrible.

I don't have to

This post pretty much says it all, with very few minor variations, all the major issues which haven't been touched that will make the Skill tree a disaster, are mentioned here.

I came back from hiatus to test this, it had potential, but it had lots of flaws that are going to ruin the game tech balance and the lower performing mechs would suffer greatly.

When P.G.I cancelled the roll out for further improvements, I felt hey they've listened, I felt good about it, ordered a lot of things because I felt we were getting somewhere and P.G.I would come back with a better skill tree.

The sad fact is they have done nothing worth while with it, all the flaws, and damage to balance and mech performance still remain.

They are totally not getting that this skill tree is not going to balance mechs it is not going to stop the need for Offensive quirks, slashing the mobility quirks is also a very bad idea.

I would probably feel o.k about this if they let quirks go untouched and then nerfed the ones that need nerfing because there would be about 10 mechs most of them IS that would over perform quite hard,but many many mechs in the game remain even with their current level of quirks serious under performers.

It would take a few months to nerf the obvious over performers because it would be so obvious, it will take years again for bad mechs to be even worth considering, dropping with.

Sorry, but after all the positive vibes I don't feel I'll be supporting the game if this goes live

[Dee Eight]

Duilliath, on 26 April 2017 - 06:47 AM, said:

I find it amazing that PGI has picked out one of the prior complaints from the last PTS and made a significant improvement to it(the transition/refund), though it is no way perfect at this point, while completely ignoring all other complaints about the new system as well as all suggestions for improvement. I have to wonder how many people will be asking for refunds on various pre-orders at this point, and how many more will once PGI forces this horribly broken system out next month.

They did that deliberately. They have no intention of changing how they wish to handle nodes with the best ones being spread around the tree and requiring players to go thru other nodes to get to them. Just because other game designers may have gone a different route on skill trees doesn't mean they're obligated to do the same. If you want the other game's skill system...go play that other game.

[Virlutris]

Dee Eight, on 26 April 2017 - 09:17 AM, said:

They did that deliberately. They have no intention of changing how they wish to handle nodes with the best ones being spread around the tree and requiring players to go thru other nodes to get to them. Just because other game designers may have gone a different route on skill trees doesn't mean they're obligated to do the same. If you want the other game's skill system...go play that other game.

Non-sequitur much?

Nothing in the quoted post said anything about another game.

Something interesting is that the devs are actually using a system of buffer/filler nodes that's common in other games.

What some folks are saying is that they don't want that style of system the other games are using. It's not about going to other games where they have the desired style of system. It's about not bringing the other games' style of system here.

Ed: clarity

Edited by Virlutris, 26 April 2017 - 09:30 AM.

[Virlutris]

Anatidaephobia, on 26 April 2017 - 05:09 AM, said:

Hi, much like last time, I’m going to provide some feedback here regarding the current iteration of the skill tree.

First of all, I want to make clear that I’m not against an implementation of a skill tree and would really enjoy the addition of being able to customise/specialise mechs further. I may sound harsh at some points or objecting certain parts, this is directed at the current concept of a skill tree, not the skill tree idea itself.

With that said, let’s jump into the meat of the skill tree; I will divide this up in the following sections :

• News post observations and initial thoughts
  
• The Good : Items which are in a good enough condition and need minor or no improvements The Bad : Items which require changes for improvement
  
• The Ugly : Items which are best described as a “Disaster”
  
• Neutral / No Opinion
  
• Personal thoughts and conclusions

News post observations and initial thoughts





Now, the biggest change I can see is more regarding the economic model which includes the keeping of progress, refunds, etc… The skill itself is relatively unchanged, which is part of where the issue lays.

Previously I’ve mentioned that I’ve noticed the “baby syndrome” in this skill tree, this observation still holds true as the main issues were the tree itself were not tackled, with only minor changes on a superficial level.

To start;
There was no change compared to previously, once more we have a skill tree system which is designed around filler-choices combined with a flawed structural layer.


I can’t see where this design goal has been fulfilled. There is no high initial investment, nor is there a reduced cost; everything is split by mostly filler nodes making either the investment not worth the effort, or said investment is a full-on requirement for gameplay.

There is no choice present in the system on offer. It remains a “Find-the-best-general-path”-meta, someone will find the optimal way to fill out this skill tree, it will be copy-pasted from mech to mech with slight adjustments where necessary.

[i]
Now, a lot of people will disagree as I’ve seen in some other threads; however I do feel like there has been significant improvement in the addressing of this issue that was previously present, and personally I am content with the work put into the system. My Basic-mechs will remain somewhat basic’d, my master-mechs will remain fully skilled out without an extra price tag added to this.

No longer do I feel as if the entire work I’ve put into this game is completely being taken away, yes, it is different, but it’s not as much of a disaster as the previous iteration was.

The next part of this thread will order items in 4 sections, these will not be chronological, but each sub-title will hopefully be a searchable term in the news post, so hopefully won’t be an issue for readability.



The Good : Items which are in a good enough condition and need minor or no improvements



[i]
I. Love. It.
It shows me exactly what I wanted to know, what I’m getting back, how many cbills, MC, global xp, modules, consumables, mech mastery returns, …

We need more of these full-open displays of data! Can we get a statistics screen like this, ingame? So I can see all my weapon accuracy, … statistics in a neat and accessible format?

Refunding consumables, both CB and MC? Amazing.

Refund from modules, cbills/skill points depending on data purchased? I can live with this, I’ve heard concerns from other people, but for me I don’t have an issue. [doesn’t mean it’s perfect, /I/ just don’t have an issue with it].

‘Mech Mastery Breakdown? Yes, I like it.

[i]
Finally, one good thing about this skill tree- It is finally clear which nodes are active, which nodes are available for purchase and which nodes are locked. That said, that’s the only good thing about it.


The Bad : Items which require changes for improvement




I’m neutral on the lowering to 1 consumable slot standard on a mech, I’d prefer more testing in a full environment to form an opinion on this. However, I’m opposed to the new maximum number of 5 consumable slots. Yes, this requires a skill point investment, but 5 consumables will have a massive effect upon the game as it is played, and I doubt it will it will be a positive light. This also makes it possible to spend around 200K cbills in a single match on consumables.
[i]
Now, it appears that ECM is considered a low investment/high reward equipment; I disagree. Not all mechs can equip ECM, it takes a slot, it takes 1 ton for clan; it takes 2 slots and 1.5 tons for IS.

Does ECM require some sort of nerf? Perhaps, I don’t know for sure, but this just seems a bit over-the-top; OR it is because the ECM boost nodes are too much of a hassle to actually get to.

[i]
As I’ve stated in previous feedback threads, this is something which should be handled separately from the skill tree implementation; there are too many variables to account for, making the entire impression rather muddled.

I do agree that this can be another method of balance, which I support, but there are so many changes coming with this skill tree that it’s difficult to pinpoint issues with the desync itself because the entire concept of changes as a whole is too large.


The Ugly : Items which are best described as a “Disaster”



[i]
… You are completely and utterly nerfing the underperformers who previously relied on their quirks to have even a slight impact;

I shall include a quick image to detail what I mean here.

There is a complete lack of understanding the role that quirks play for underperforming and bad mechs; there is absolutely no reason to nerf the Vindicator or Trebuchet or Victor, these mechs are already bottom of the barrel.

If you’re nerfing ahead of time for the new technology coming into the game, WHY? You’re going to make these already terrible mechs almost completely unplayable for the duration between the skill tree and the new technology being released.

And most of these mechs have issues that new technology won’t even solve, I have absolutely no idea what kind of logic was used to come to the conclusion to nerf almost every quirked mech; while top performing mechs such as the Marauder IIC and Kodiak will be gaining extra advantages from the skill point spending. Good mechs are becoming better, bad mechs comparatively even worse than they were before. Just take a look at the two listed PDFs, aside from the engine desync quirk removals, almost EVERY weapon-based quirk is getting nerfed. Victor, Vindicator, Dragon, Centurion, …
[i]
Yeah, it pains me to put this here yet again, but it’s just not up to snuff; it remains bloated and oversized, barely impactful and a mess to traverse. The same issues still remain from my previous feedback session

The design of the tree does not conform to any of the design goals that were stated; there is no meaningful choice, there are too many nodes with too many skill points with too many small numbers which don’t matter, and there is too much filler.

Having 242 Skill Nodes where you can use 91, is not a good number. Bigger numbers aren’t better, it just needlessly complicates things.

And, once more, any system which has a number such as “91” as an “optimal” number, is fundamentally flawed in its design and no matter what you try with said design, it won’t solve the glaring issues present with it.

Trying to make this work, is both pointless and a waste of time and effort; please go back to the drawing board, it doesn’t need to be complex, it needs to be functional where it matters. Heck, if I were living anywhere close to Canada, I’d offer to do it myself..

The UI is… messy? It’s easy to lose track of things, there are some bug and issues; as an example, let’s play a game : In the following screenshot, how many historical skillpoints do I own on this mech, and how many skill points do I need to buy to get exactly 91? Leave your answer in the comments, because I sure as hell don’t know.

Album detailing the issue is here.


Neutral / No Opinion / Not enough information to work with


[i]
Not enough data to work with, I don't know if this is good, bad, no effect or superficial.


Personal thoughts and conclusions


While I can see the positive effort that was put into the entire refund system, which was one of the main issues of the previous iteration; the main glaring issue that is the skill tree design itself, is still present and is as bad if not worse than before.

Regarding quirks, I don’t think anyone currently in charge of balancing actually realises what role these quirks serve on bad mechs, what the quirk accomplishes and what should be done before removing said quirk. In some cases, removing the quirk is not an option, nor is lessening it because all it will do is make a bad mech into a terrible mech.

Making the quirks matter worse, is that quirkless top performing mechs suddenly gaining boosts. If everything gets the same boost, you can talk that things will remain similar to eachother, however in this case, the result is that the great mechs become amazing, widening the gap between the bad&lt;-&gt;underperformers&lt;-&gt;good mechs even further. I’m sure that this is not something that is actually intended, atleast, I hope it isn’t.

Tinkering is punished; if you want to swap from a PPC to an LPL, it will cost you around 6-10K experience each time; experimenting is still punished.

The Skill Tree UI, aside from node colouring, is terrible. It’s a mess, it’s overly complex, far too many numbers and repeats and … Yeah, there is not more I can really say.

_________________________________________________


I’m honestly starting to feel burned out on this tree, it’s not changing, the majority of my feedback is still the same as it was for PTS1 and I doubt what I’m saying will actually get considered or reach the right ears.

Similar to Cathy's statement, you've covered most of the points I've been chewing on.

Thanks for carefully any thoughtfully articulating your thoughts.

I think your treatment of the subject on the outreachhpg subreddit did a great job detailing a positive and constructive proposal for the tree, and addressing the reasoning behind it. Well done.

I might add a thread of my own that proposes a couple ways to handle particular branches, but you've covered a lot of my thinking. Thanks again for that.

Edited by Virlutris, 26 April 2017 - 04:43 PM.

[Hotrob]

This basically sums up what I said in my post and more. Very good and thoughtful breakdown.

The only thing I want to add is that I looked at the Shadowcat Prime and Blackjack one, and after all the nerfs, the shadowcat is left with an additional 85 nodes worth of buffs over what a current baseline shadowcat would do. The blackjack 1 gets ten.

This is the kind of imbalance we're talking about here.