140 replies to this topic

[Ultimax]

Cato Phoenix, on 28 April 2017 - 12:29 PM, said:

Which is the problem Solahma's tree fixes. Allows you to individually customize your weapons to your liking.

Yes, his tree not only lets you customize your weapons to your liking.

It also allows Mixed Weapon loadouts to not lose a single thing cost-wise in the tree vs. boats - allowing a build with Ballistics, Lasers & Missiles to put 10 skill points into each of those trees - where the boat can only add 10 points to the one weapon it boats.

Even better, it allows PGI granular control over mech balance in this area by allowing a KDK to say, only have 5 "Firepower" points and an Atlas to have 10 - while not affecting the rest of the trees at all.


It's such a simple, clean solution - with a superior UI. https://www.reddit.c...ockup_proposal/

Edited by Ultimax, 28 April 2017 - 01:34 PM.

[Shifty McSwift]

So suddenly it is not about choice anymore, but what you personally consider to be the best selection of choices, your meta ideal, that would so automatically kick my butt all day. Is that right? Sorry it's hard to keep up with a changing discussion base. Maybe it will be because its too ugly to you next, and how can anyone say you are wrong in thinking it ugly, you are the beholder right?

[Clownwarlord]

[AphexTwin11]

Lukoi Banacek, on 28 April 2017 - 01:03 PM, said:

What does it remove?

I'm not going to defend specialized builds, because 1) I've never said they were a good idea, good thing or healthy for the game (Direstars are specialized crap too) and 2) because the Tree isn't designed to make that easy. It's designed to make you PAY for the hyper-specialization.

What this adds is granularity to things however, that can be useful. The faster a mech I'm running, the less I'm going to ramp up Radar Derp for example....don't need it as much imo. Before you were either all RD or no RD. This offers more customization in that regard.

So, I'm not seeing what it "removes." You can literally play something that's almost identical to what you have now (with the caveat that there are some universal nerfs in this that impact everyone the same so it's a wash), with some probably unintended ancillary benefits. Or you can tweak the fine tuning of %'s as your mech/playstyle role differ based on the situation.

That's not removing anything. It might be really anticlimactic to re-click 91 times to achieve largely the same thing BUT you have the option to do things much differently. But if you want to get away from what we're all used to, you have to pay a trade off for it. Don't see the problem with that part of the ST really.

^ That is probably the best, most sober way to think about it. I agree that hyper-focused builds are not good for the game, especially with so many complaints (legit or otherwise) about low TTK, Power Creep, and imbalance. Making you pay in SP for quirks you might not necessarily want in order to make a mech hyperfocused, appears to be a way of baking in some opportunity costs to min/maxing. I think the granularity and increased customization of the skill tree is a great addition to the game, especially considering you don't need to purchase (with C-bills or $$) two additional mech variants that you may or may not want. I can totally understand the concerns of IS pilots with their baseline quirks being removed as well, and I definitely think that is something PGI should address, or at a minimum, explain the rationale for.

To some extent I can see SC's point, but stomping your feet and wailing at the top of your lungs about this seems pretty short-sighted and selfish. Not saying the skill tree is PERFECT for EVERYONE, but to some extent I feel you need to look more at the forest and not just the tree.

Edited by AphexTwin11, 28 April 2017 - 01:39 PM.

[Lukoi Banacek]

MischiefSC, on 28 April 2017 - 01:30 PM, said:

Currently it removes the ability to have derp and sensor stuff as modules to get the weapon module performance (or close to), which as I said I'm okay with. We're rolling back power creep.

This is where I'm having a hard time following you, as demonstrably, this is not true. I have the same sensor modules, better mobility and generally same heat management/weapon bonuses as I do currently. With this system, clunky as it is, I can reduce the RD and save a few points for something else, whereas on live, I can either take RD or not have it at all. No granular options.

[VanillaG]

MischiefSC, on 28 April 2017 - 01:02 PM, said:

Go build something other than a new version of what's on live.

Go ahead.

Then compare it to just buying the new version of what's on live.

Then you'll see.

There are no 'tradeoffs' any more than putting 6 MPLs in a Dire and filling the rest with a TC and DHS is a tradeoff vs taking any loadout that doesn't suck on a Dire. That's not 'tradeoffs'. It's not 'customization'. It's being presented with good choices, that get you exactly what we have minus modules and some very slight changes, or you make a mech that plays like a live mech without basics unlocked but gets a few sensor quirks and some extra armor.

In which case the new skill tree is pointless.

Remove Derp/Seismic, weapon modules, refund everyone.

Decouple engines, etc. changes.

Remove the hill climb mobility stuff added years ago.

Boom. We've now got the game the new Skill Tree will create, minus a bunch of traps for new/bad players to make terrible, terrible choices on skilling up a mech.

Comparing to live makes no sense because the game is changing. All of the specific cooldown modules are going away, at best you get back the Cooldown skill from the old tree. Doubled basics for heat management is going away making heat management more important. They are adding ways to increase the amount of armor and structure on your mech, make your mechs fly again, carry more than one of specific type of consumable, and add specific quirks for certain types of weapons. The generic range, heat, and velocity quirks actually promote multiple weapon types in a build and having to take nodes for a weapon type that you do not have to max them punishes boating.

There are real trade offs under the new game, comparing them to live is foolish because the game is changing. Just like when the game changed when Endo, XL Engines, and DHS were added. Just like when the game changed when the Clans were added. Just like the game will change when the Civil War tech gets introduced. This tree and engine decoupling fixes some of the issues with power creep that occurred over the two previous changes and hopefully mitigates some that will occur with the next.

[Flak Kannon]

I agree with original post.

Skill Tre,e in concept is good.


In reality, as its currently coded, it will 'massively change' the game.

I do not say this in jest.


BACK TO THE DRAWING BOARD>


Enjoi

[Ced Riggs]

There is a reason why I suggested this, but I am also aware that it's not going to be taken into account, nor implemented. The current PTS #2 Skilltree is, with mild variation, what we are going to get. Piggy won't throw out what they made so far.

[Lukoi Banacek]

Ced Riggs, on 28 April 2017 - 01:46 PM, said:

There is a reason why I suggested this, but I am also aware that it's not going to be taken into account, nor implemented. The current PTS #2 Skilltree is, with mild variation, what we are going to get. Piggy won't throw out what they made so far.

Too much sweat invested, I agree. We're getting some semblance of this.

That's why I offered up a handful of I think reasonable, quick-fixable items and considerations that I hope they listen to:

- rework portions of the tree where a node interrupts progression with a purely useless feature; a missile node on a mech with no missiles etc. I don't mean "disliked/unwanted" node btw...I don't WANT gyro stability, but I'll take it as an ancillary benefit along the way if it means I get to the node I want...it remains a positive effect for me, even if I wasn't going to go for it. I understand that gatekeeping mentality and it makes sense. But completely useless nodes shouldn't happen. If it's a node that cannot possibly benefit the mech as built, it should be greyed out/reworked imo.

- streamline it a bit. Some of this is just unnecessarily convoluted. You can keep the gatekeeping but the UI could be easier to see, less cluttered and more intuitive.

- reconsider what you're doing with quirks. I understand that newtech is coming...but until it does, alot of these mechs really need to keep some more of these quirks. Especially with the added variable of desynching engines from mobility issues.

Edited by Lukoi Banacek, 28 April 2017 - 01:53 PM.

[Shifty McSwift]

Lukoi Banacek, on 28 April 2017 - 01:52 PM, said:

Too much sweat invested, I agree. We're getting some semblance of this.

That's why I offered up a handful of I think reasonable, quick-fixable items and considerations that I hope they listen to:

- rework portions of the tree where a node interrupts progression with a purely useless feature; a missile node on a mech with no missiles etc. I don't mean "disliked/unwanted" node btw...I don't WANT gyro stability, but I'll take it as an ancillary benefit along the way if it means I get to the node I want...it remains a positive effect for me, even if I wasn't going to go for it. I understand that gatekeeping mentality and it makes sense. But completely useless nodes shouldn't happen. If it's a node that cannot possibly benefit the mech as built, it should be greyed out/reworked imo.

- streamline it a bit. Some of this is just unnecessarily convoluted. You can keep the gatekeeping but the UI could be easier to see, less cluttered and more intuitive.

- reconsider what you're doing with quirks. I understand that newtech is coming...but until it does, alot of these mechs really need to keep some more of these quirks. Especially with the added variable of desynching engines from mobility issues.

I pretty much agree and respect you for consistently remaining reasonable.

I would think the idea of making the skilltree dynamic to mechs or mech loadouts might be something to work on in further advancing the system in the future, the idea of new mechs and tech means the skill tree is probably always going to be subject to further change so it can't really be final in this release of it anyway.

And with quirks, I am just in the mindset that while some quirks might be making some non meta mechs capable of competing in the meta, that wasn't really the purpose of them, the scaling back of the huge lists of quirks specifically is not a concern IF the devs will be willing to monitor and identify the mechs that suffer the most/bottom meta mechs and work to boost them in other ways, basically the same as quirking but just direct buffs to base stats or whatnot, on a case by case basis for the mechs it applies to (while at the same time nerf batting the top end guys a bit).

Obviously that is way too much trust in PGI for some to accept, and/or just naivete on my part. There are plenty of things I am critical of and advocate change for, or have opinions on, but it doesn't make me automatically always assume the worst. Not saying this applies to you, maybe it does maybe it doesn't, just my current thoughts.

[VanillaG]

Ultimax, on 28 April 2017 - 01:33 PM, said:

Yes, his tree not only lets you customize your weapons to your liking.

It also allows Mixed Weapon loadouts to not lose a single thing cost-wise in the tree vs. boats - allowing a build with Ballistics, Lasers & Missiles to put 10 skill points into each of those trees - where the boat can only add 10 points to the one weapon it boats.

Even better, it allows PGI granular control over mech balance in this area by allowing a KDK to say, only have 5 "Firepower" points and an Atlas to have 10 - while not affecting the rest of the trees at all.


It's such a simple, clean solution - with a superior UI. https://www.reddit.c...ockup_proposal/

There are two major flaws with his design. The first it assumes that all nodes are of equal value, which in the current system is not the case. Some nodes are more valuable but their nature and with the skill tree we have right now you have to pay extra for them because they are at the bottom of the tree. All this does is promote cookie cutter builds because you don't have to invest more to get the max benefit.

The second flaw is the per weapon tree. It encourages boating early in the game because it takes more XP to unlock different weapons and discourages experimentation because you need to grind out each weapon type. This would be really apparent when all the new tech drops in July. With the PTS tree the new tech should just work and not need any additional nodes if there are no new weapon attributes added. You still might want to tweak your trees but you don't need to grind to use the new weapons.

[Khobai]

PGI was originally going to give us unique skill trees for each variant

I think that could work because then each skill tree is catered to the strengths and weaknesses of each mech. It lets them make sure that each mech stays different.

Rather than getting rid of quirks that make mechs different and replacing them with a generic skill tree that makes all mechs the same.

[Shifty McSwift]

Then again even if they did people would still complain about its complexity, or the amount of clicks and reading involved, or complain that the relevant and exact quirk to ST changes weren't exactly perfect enough and they will threaten ragequit because of wasted money on a now different mech.

You literally can't make a choice in game development that doesn't annoy someone in some way.

[Bud Crue]

Lukoi Banacek, on 28 April 2017 - 01:40 PM, said:

This is where I'm having a hard time following you, as demonstrably, this is not true. I have the same sensor modules, better mobility and generally same heat management/weapon bonuses as I do currently. With this system, clunky as it is, I can reduce the RD and save a few points for something else, whereas on live, I can either take RD or not have it at all. No granular options.

Could you share your node paths? I don't think I am alone in thinking that if you have come up with an optimized set of branch pathway that provide Derp, Seismic, the same heat and mobility characteristics, and weapons bonuses, as a given mech has on the live serve you would be doing the community, especially new players, a service if you shared them. Obviously we would need to modify them for specific mechs but even having yours as a starting point would be a great benefit.

[Marquis De Lafayette]

MischiefSC, on 28 April 2017 - 12:21 PM, said:

It is an utter waste. If I have no energy hardpoints or am not going to use energy hardpoints on a mech then 4 SP on laser duration is a total waste. If I have only torso mounts on a mech then arm flex nodes are a total waste.

An utter waste. 100% pointless. 0% return for the value spent.



I get the concept. However the current skill tree does nothing but provide some blanket nerfs and put us exactly where we are now. There's no tradeoff, no point in a 'skill tree' at all really. Just remove Derp and Seismic and Zoom and all weapon Cooldown modules, cut the value of range modules in half and nerf mobility a little. Boom. Done.

If the specific skills in the skill tree were linear then I would have real tradeoffs.

So like this:

Heat Generation Hill Climb Extra Ammo Consumable Perk Arm Elevation
0.5%......................7.5%.................+10.............................10% ...........................5%
1.0%.......................15%.................+20..............................20%..........................10%
1.5%.......................22.5%..............+30..............................30%...........................15%
2.0%.......................40%.................+40..............................40%............................30%


So the advantage of progressive quirks in the same tree is usually linear for useful quirks but could be graduated for less useful ones. However if I wanted to give up 0.5% of heat gen to get 7.5% hill climb or bonus consumable or ammo quirks, I can.

That would be a real set of tradeoffs. It makes the less useful (generally) quirks suddenly worth considering a somewhat larger investment because at the upper tiers they give a steeper benefit.

Make sense? That would be real customization and a real skill tree. Mechs with arm mounted weapons you could logically decide to invest in arm quirks because those quirks would be useful to that mech, making it play different than their torso weapon mechs. Ballistic mechs could drill down on velocity and ammo at the expense of heat gen.

Do you get what I mean? That would be real tradeoffs. That would make for meaningful customization. It means that on my PHX I could turn an otherwise bad mech into a good one by drilling down on what it does best to make it exceptional instead of having to give it the same general stuff everything else has just to keep it from being drastically inferior.

I think is possible to totally agree that how you are proposing would be better for customizing your mech the way that makes sense and to also be willing (not enthusiastic) to let PGI go a different direction in how they are viewing trade-offs.

Don't get me wrong....I like your way better. I can just see that PGI is also forcing trade-offs and more painful ones at that. If I want all the speed tweak, I am going to have to take things I don't want to get to the highest level. In effect it's costing me more than 1 node to get that extra level I actually want....as I have to take some other node that is a "waste".

However, if we are all forced into those kind of choices it doesn't hurt balance. All our mechs suffer. Now it might reduce TTK as people can twist as well, etc. I expect "unintended consequences"...that is where the game could be "broken"

Maybe PGI has just finally wore me down on this subject. the current skill tree makes me take skills I don't need on some mechs to unlock the x2 bonus to basic skills....so maybe I didn't expect that I wouldn't have "wasted nodes" here. Idk. Of course I prefer to not have them...but it isn't making me crazy to have them do it to me again.

[Insanity09]

I have now tried to wade through the skill tree.
All in all, this feels very much like an effort from game designers who have next to no design experience and/or have not played any mmo or game where customization of a character or similar was possible. Sad.
Apologies for the length herein, I wanted the critique to be fairly comprehensive.

Things I like.

• The armor/structure values are higher for lights and lowest for assaults, scaling along the way.
  
• related to that, you no longer seem to spend tonnage for the boosted armor (yay!)
  
• The jump jet web kinda works. (mostly because the main lines of the web are direct links to each other, rather than a weave through extraneous stuff)
  
• Mechs that had skills unlocked in the old system seem to get a certain number of unlocks for free (no xp/gxp and cbill cost). All those mechs I had basic'd (only 1 chassis variant) still start with some free nodes. (However....)
  
• Some quirks are still present, though typically reduced in value
  
• Weapon range, cooldown, and heat gen are all now generic. (Nice, but there is a downside)

That's pretty much it.








Things I dislike (longer list, and please read with greater vehemence, many of the things I didn't like, I REALLY didn't like)

• I still have to unlock things I don't want or need (hill climb, gyro, shock absorbance, speed retention, target info, crit chance reduction, target decay). I'm not saying these things aren't useful, I'm saying I don't want to be forced into spending points on them. Wasn't avoiding unneeded/undesirable unlocks a goal of the new system?
  
• don't even get me started on the consumable changes
  
• the armor/structure bonuses feel insignificant. E.g. after getting every single armor bonus (8 nodes? 10? PTS crashed me out and won't reload, so I can't count atm) for a 70 ton arc-5r, I got 7-12 armor in each location (arms 7, ct 12, etc., ok, only 2 hd, but w/e)
  
• many of the new values often do not achieve what the old skill system did (e.g. getting every cool run node yields 10% faster cooling, vs 15% in the old/current system)
  
• with the exception of the weight class number differences, the trees do not seem to vary between variants or chassis. (so, the all energy black knight must spend points in ballistic/missile nodes to get to some of the range/heat gen/cooldown nodes). (you do get warnings that some nodes might not apply to your mech)
  
• the interface is atrocious
  
• HSP grants seem to be based on skills you had actually unlocked. So, even though my pxh-2 was fully basic'd with an extra 60k xp sitting on it (enough to allow me full mastery in the event I ever bought and basic'd two more variants), I only got 23 HSP nodes for that variant. The extra xp is just gone into the ether, apparently. Which sucks.
  
• The values for a given node, esp. considering how many unwanted nodes must be purchased, are very small, and thus feel of limited value, at best. -1.5% laser duration per node? -1% crit chance per node? You gotta stack lots of them to start feeling the benefit.
  
• In the firepower web, I'm sad that I can't pick nodes that enhance only those weapons I actually use on a mech
  
• Many (all?) of the supposed enhancements are effectively blind purchases, I can't see what the changes actually get me (that 1.5% range does what for me? or 2% torso yaw?) without constantly tabbing back and forth pre- and post- change. And in some cases, even that is a little nebulous. (in fairness, the old system was guilty of this as well, but with boatloads of enhancement numbers now...)

A few suggestions...

• Condense the webs. (example, in the firepower tree, just have a single range node that can be clicked up to x times)
  
• Simplify the prerequisites, rather than weaving through the tree to get to all the nodes for a given enhancement, apply the simplified web (above), but have each level of a node require certain prior point expenditures and/or points spent in other enhancements (a couple examples: say to get the 4th armor bonus 4, you must have spent 5 points in the survival tree with one at least structure and one crit reduction bonus; in the sensor area to get seismic 1, you must spend 8 points on sensors, with at least 3 points in sensor range; etc.) mousing over the node you want to purchase should tell you what pre-reqs you need to achieve. Also, please note that simply adding the pre-reqs in such a way that they effectively duplicate the existing webs doesn't seem like a reasonable path forward.
  
• In addition to generic nodes, add nodes for specific weapon type enhancements in the firepower tree (e/b/m only), give them higher pre-reqs(?) or greater cost (iffy) if you must, but they should offer a higher bonus (4% instead of 1.5%?). Adjust numbers so that a maximum bonus can only be achieved by buying both generic and specific nodes. however...
  
• Make each node more significant (at least 2% each?)
  
• Offer special nodes/webs specific for variants/weight classes. E.g.: perhaps lights could get to 15% speed tweak instead of just 7.5%, perhaps the cyclops could have greater sensor boosts; etc.
  
• It is okay (imo) to require more sp for the unlocking of some nodes. That's a trade-off. Seismic 1 and 2 might cost 2 points each, for example.
  
• when it comes to listing the enhancement numbers, since there are so many now, please, for sanity's sake, categorize them.
  
• For bonus points, actually show the effects when mousing over a node (eg. so mousing over the range node, show the medium laser range before and after unlocking that range node)

[MischiefSC]

Shifty McSwift, on 28 April 2017 - 01:33 PM, said:

So suddenly it is not about choice anymore, but what you personally consider to be the best selection of choices, your meta ideal, that would so automatically kick my butt all day. Is that right? Sorry it's hard to keep up with a changing discussion base. Maybe it will be because its too ugly to you next, and how can anyone say you are wrong in thinking it ugly, you are the beholder right?

Way to try and shift the goal post.

It's really simple. Customization that makes a bad mech isn't customization, it's just the ability to make poor choices. Good customization would be to have a variety of good choices to choose from.

What works best isn't my opinion in MWO. It's what statistically and mathematically and historically, both through application and observation, works best. I hate poptarting. Not a big fan of laservomit. However those have their specific values and the circumstances where they work best.

My opinion or yours doesn't change that germs not evil humors cause sickness or that the earth orbits the sun. There's no opinion here. Making bad choices and the ability to design something that works worse with no real useful tradeoff isn't real customization any more than the ability to put a STD engine in a Clan mech is a real option.

As I said before. The goal isn't to make just 1 good path - that's my absolute core complaint with the current Skill Tree, that's what it really does and that path is the exact same thing we've got right now, which is so pointless. The goal is to make real tradeoffs; The granular ability to adjust one value directly for another. Have you seen Solahams example? It would let you really make useful variations. I could make 2 of the same mech with different paths that are both approximately equal in value but very different performance profiles. It also reduces the value in boating while giving a lot of opportunity for value in mixed loadouts.

This current PTS has none of that. You go in, click through the same stuff you've got to have to be baseline viable in ops and agility, spend your 19 points on the right or left side depending on you being a laser boat or not and away you go.

Modules give me more useful decisions to make on a mech.

My point is that the new skill tree just puts us exactly where we are now. All this **** storm over refunds, irrelevant. Why are we doing it? It's a lot of change, time, cost and energy to get back into the same spot.

Add a cbill cost to the XP cost on existing skills. Remove modules. Refund people. Build hill-climb into all existing mechs. Boom. Better than the skill tree because it doesn't create yet another way for terribads to make their mechs even worse than they are now compared to everyone else.

Or make the skill tree something worthwhile. That's the irritation with the pointless unlocks. Their very existence reminds, every time you're unlocking a mech (and many of us have a lot) of how bad the skill tree is compared to what it should be. I don't want to have to look bad choices in the face 144 times. If I wanted constant reminders of other peoples bad choices and how they negatively impact me I'd friend my crazy ex on Facebook.

[Shifty McSwift]

MischiefSC, on 28 April 2017 - 03:15 PM, said:

Way to try and shift the goal post.

What? You said, "go and try it for yourself", which I already have and explained quite simply that with the choices available I built my mech in a way that was different to its previous configuration. To that you basically responded that my terribad choices of not selecting the exact buffs you would makes that point inert, I proved that there was choice involved, you then made it about being "bad or good" and that there is only one true choice for all mechs within the skill tree.

There will be a meta that develops with the ST as with almost everything in the gaming world and beyond, there is no way to avoid that really with a system of choices, there will always be ideals of "the best/most efficient choices" that applies to different people in different ways.

The system we have now is quite horrible at providing real choice, to me it seems, MUCH more so than the currently advertised version running on the PTS.

[Mister Blastman]

I wouldn't mind them reworking the skill tree itself. You make a good point, Mischief, we shouldn't need to grab useless crap to get the good stuff, as it defeats the purpose of targeted builds.

If some of us want our 'mechs to stub their toes on rocks in order to armor up, let us. When that robit is stuck in the bottom of a canyon, much good that armor will do when they can't move to cover.

The skill tree should be about tradeoffs, not forced normalcy. There's too much emphasis on requiring players to use skills they don't want in order to get the stuff they do.

So PGI can do two different things:
1. They can change the skill tree and make it better.
2. They can make the useless skills far more valuable. Make them do something worthwhile.

[MischiefSC]

Shifty McSwift, on 28 April 2017 - 03:26 PM, said:

What? You said, "go and try it for yourself", which I already have and explained quite simply that with the choices available I built my mech in a way that was different to its previous configuration. To that you basically responded that my terribad choices of not selecting the exact buffs you would makes that point inert, I proved that there was choice involved, you then made it about being "bad or good" and that there is only one true choice for all mechs within the skill tree.

There will be a meta that develops with the ST as with almost everything in the gaming world and beyond, there is no way to avoid that really with a system of choices, there will always be ideals of "the best/most efficient choices" that applies to different people in different ways.

The system we have now is quite horrible at providing real choice, to me it seems, MUCH more so than the currently advertised version running on the PTS.

Except the choices are between something exactly like we have now and something worse. That's the choices and that's no choice.

There's no 'one true choice' any more than there's one true set of modules to play. However everyone maxes the existing skill tree the same. If there were options in the skill tree we have now to get Speed Tweak AND Cool Running but get a 5% UAC cooldown bonus, would you say that's a good tradeoff? Because, to be clear, it's not. For any build. That's not opinion that's an observation of the math involved.

Again. Real choices have comparable value. Bad choice vs good choice isn't a real choice. Giving someone the choice to build a mech that would lose against someone who chose to make a good mech isn't a 'choice'. It's just a skill someone learns to do well or not, make good choices.

Hence it eliminates the point of the new skill tree.