65 replies to this topic

[Shifty McSwift]

MischiefSC, on 29 April 2017 - 08:30 AM, said:

UAV and situational awareness > seismic and derp which, at best, give a small advantage maybe 1x in a match. Mobility gives you an advantage in every shift of positioning, trade and in a brawl.

UAV and situational awareness is really not greater than seismic with decent range, seismic is frigging amazing at ensuring certain things like not having to predict movements ever, your guesses, no matter how educated and a limited use/time/health consumable can't really top the given nature of seismic.

Derp is different because we have come to weigh module values, and the defenses it provided outweighed that of most other modules except in specific situations/particular builds, it has defensive value that varies between builds.

But yeah I agree about mobility, it is of critical importance, mostly to (ironically) survivability

Edited by Shifty McSwift, 29 April 2017 - 08:38 AM.

[Wintersdark]

MischiefSC, on 29 April 2017 - 08:30 AM, said:

Then there's no advantage to not maxing the best stuff and never touching the rest. Ideally you want a system that gives me a reason to take anything but max agility, Ops and 19 pts in weapons on a laservomit mech. There's nothing I. The other trees worth the cost of giving that up for. If I could scale some of that down in manageable pieces to get other things I would. This system doesn't allow that. Gating means 1 pt if a useful sensor node costs me 5 useful points from mobility or OPs tree to get. Bads will do that. They'll give up 50% of their weapon performance and mobility to get 10 total pts of health on the location that matters. So they'll be worse on every trade, every exchange of weapon fire and every position shift so that 1x they can soak 2 ML of damage.

So they'll be bad, spending cbills to play worse than they do in live already. That's not a tradeoff it's an invitation to rage. In a sick way I appreciate that PGI is trying to nerf 75% of the games players while buffing 25% (my Roughneck 3A is better on PTS than live for example because I'm not making bad skill tree choices) but it's bad for the game.

I'll bet you a mech pack right now that if this goes live then less than 30 days after that we'll have people raging over the metamech equivalent of skill tree and how it's ruining the game because those evil tryhards are making smart skill tree choices and smart mech choices and the people in non-meta mechs (which are even worse in the new system) and bad skill tree choices (like spending any points in sensor and 99% of the time survivability) are losing more than they were in the old system.

Give your head a shake. No matter what PGI does, what you warn about above is going to happen. Even if there's much greater balance between "good skills" and "bad skills", there will ALWAYS BE "the best skills" and there will ALWAYS BE "the best builds".

ALWAYS. It's inevitable.

If you allow players choice, players will make bad choices.



The ONLY way to avoid your complaint is to not allow choice at all, in which case we'd be better off without skills at all and simply have NO skill tree. I can see that as an argument, but it's less fun as progress adds reason to play, and reason to play adds content for other players.


A way better way to approach this is:

Does "everyone" [read, everyone who's being successful] take the same skills? If so, increase value in unused skill trees to make them more competitive. Maybe make Seismic work better, or work at higher speeds. Increase the mid-level skills value, to make Hill Climb actually do something.

In short, buff the trees nobody takes to make them more compelling choices.


But to be honest, I think you're overstating the value of some of the trees. The mobility gain, for example, from the agility tree is noticable but NOT one that's going to let you safely circle an identical mech without it. It WILL give you an edge, but then the mech without that has other advantages - Maybe he's got 2 more UAV's and Seismic, so while once in a brawl you've got a slight edge (again, it's not that huge), he's in a better position to get the first strike, or to avoid someone else getting the drop on him.

Maybe at the very top end of play (well, not maybe) the relative value of trees will be starkly different, but at "normal" play - and particularly in solo queue play - things like Sensors become a lot more valuable, so you need to think of this across all levels of play, not just at top end play.



Ultimately, though, you know this damn well: No matter how well balanced it is, there will ALWAYS be an optimal build, there will ALWAYS be "meta" skill builds.

Always.

[MischiefSC]

Shifty McSwift, on 29 April 2017 - 08:37 AM, said:

UAV and situational awareness is really not greater than seismic with decent range, seismic is frigging amazing at ensuring certain things like not having to predict movements ever, your guesses, no matter how educated and a limited use/time/health consumable can't really top the given nature of seismic.

Derp is different because we have come to weigh module values, and the defenses it provided outweighed that of most other modules except in specific situations/particular builds, it has defensive value that varies between builds.

But yeah I agree about mobility, it is of critical importance, mostly to (ironically) survivability

Wallhack is great but not as universally critical as the mobility and cooling from agility and ops.

Once you find the enemy and engage then sensor stuff declines in value. Mobility, heat management is 100% useful in every part of engagement. I haven't put seismic on most my mechs in a while, most my team doesn't. Derp is convenient but not critical. Weapon modules help you win trades and brawls which is what really wins matches. Mobility and ops are the difference between a mech without even full basics and a mastered mech. It's telling and critical.

[DAYLEET]

MischiefSC, on 29 April 2017 - 08:30 AM, said:

UAV and situational awareness > seismic and derp which, at best, give a small advantage maybe 1x in a match. Mobility gives you an advantage in every shift of positioning, trade and in a brawl.

uav is the one thats a one time use, not seism or derp, they just keep on giving. About mobility, id rather have agility, speed do affect turning but not as much as pure agility skills. Maybe ill use both, depends the mech im gona use and how people adjust. Thats the point of the tree, im gona make different choices. Maybe you wont but thats like people who only play meta mech, its their choices too.

accell and decell are great but easy to access, speed is at the end and not worth the points for me.

Edited by DAYLEET, 29 April 2017 - 08:48 AM.

[MischiefSC]

Wintersdark, on 29 April 2017 - 08:41 AM, said:

Give your head a shake. No matter what PGI does, what you warn about above is going to happen. Even if there's much greater balance between "good skills" and "bad skills", there will ALWAYS BE "the best skills" and there will ALWAYS BE "the best builds".

ALWAYS. It's inevitable.

If you allow players choice, players will make bad choices.



The ONLY way to avoid your complaint is to not allow choice at all, in which case we'd be better off without skills at all and simply have NO skill tree. I can see that as an argument, but it's less fun as progress adds reason to play, and reason to play adds content for other players.


A way better way to approach this is:

Does "everyone" [read, everyone who's being successful] take the same skills? If so, increase value in unused skill trees to make them more competitive. Maybe make Seismic work better, or work at higher speeds. Increase the mid-level skills value, to make Hill Climb actually do something.

In short, buff the trees nobody takes to make them more compelling choices.


But to be honest, I think you're overstating the value of some of the trees. The mobility gain, for example, from the agility tree is noticable but NOT one that's going to let you safely circle an identical mech without it. It WILL give you an edge, but then the mech without that has other advantages - Maybe he's got 2 more UAV's and Seismic, so while once in a brawl you've got a slight edge (again, it's not that huge), he's in a better position to get the first strike, or to avoid someone else getting the drop on him.

Maybe at the very top end of play (well, not maybe) the relative value of trees will be starkly different, but at "normal" play - and particularly in solo queue play - things like Sensors become a lot more valuable, so you need to think of this across all levels of play, not just at top end play.



Ultimately, though, you know this damn well: No matter how well balanced it is, there will ALWAYS be an optimal build, there will ALWAYS be "meta" skill builds.

Always.

I agree with everything you said but sensors. They are nice, sure, but what you can do when you identify a target is way more useful than identifying to begin with. In solo queue if I'm just slightly more cautious on the approach and pay a bit more attention to who's where I eliminate my need for seismic entirely. Identify the targets location and position to engage. Check for teammates position (his and yours) , check exposure and then move to engage or trade.

All the new system does is reward me more for staying at longer range and trading more. Even if it pushes to a brawl I'm still keeping a performance edge.

[Shifty McSwift]

MischiefSC, on 29 April 2017 - 08:47 AM, said:

I haven't put seismic on most my mechs in a while, most my team doesn't.

Well that is the other aspect to it, if you have a coordinated team all it takes is one or two guys to equip it to get the full advantage for your team too, which just makes it even better really.

[MischiefSC]

Speed is control of positioning, time and pacing of engagement and disengagement. Sensors but inferior speed are just going to give you a bit more info about what I'm forcing you to do because I'll be in position first. It's the difference between being action or reaction. In pug queue it's the difference between keeping up and left behind. Speed is life in a lot of ways.

[Shifty McSwift]

MischiefSC, on 29 April 2017 - 09:40 AM, said:

Speed is control of positioning, time and pacing of engagement and disengagement. Sensors but inferior speed are just going to give you a bit more info about what I'm forcing you to do because I'll be in position first. It's the difference between being action or reaction. In pug queue it's the difference between keeping up and left behind. Speed is life in a lot of ways.

Yeah man there is no argument here on the value of speed, I am there with you.

I guess, for the sake of argument though, the fact you can fill the speed related trees and have points to spare means that you don't ever really need to compare speed directly with seismic, at that stage you would be weighing it against other choices.

[Cy Mitchell]

I wish that the skill tree was about 30% smaller and that you had about 30% less points to invest. I understand and emmpathize with what PGI is trying to do with the Skill Tree. They are trying to roll back some of the years of power creep and they are trying to limit the min/maxing. I support both goals. They way they have chosen to do this is a less than elegant design but it does manage to accomplish some of those goals IMO.

The "gating" is used to raise the prices of the desirable skills. This has been explained to us by Chris Lowery who is one of the Devs working on the Tree and quirks. Picking intermediate nodes in between a specific line of nodes raises the prices of those subsequent nodes. In the meantime, you receive small benfits from those intermediate nodes too. You may not have wanted to chosse Hillclimb nodes but it sure is nice when you come to an incline on Canyon and you just roll right up it instead of slowing to an almost complete stop.

No one "forces" you to take "worthless" nodes. You can chose to pursue other paths that will give you something that will benefit you. But if you are in the mindset that you must have those last two heat gen nodes then you may have to pay a higher price to max out the heat gen line by choosing a couple nodes that will not benefit your build. You may not like it but it makes sense if you consider what the devs are trying to accomplish.

I wish the whole system was leaner. However, at this point I do not think it will happen. Everything, including the refund calculations, are built around the 91 Skill point model. Changing that changes everything and moves the ST back at least another couple months. I do not think that a delay that is an option with what is coming this summer. Besides if they did what I want and reduced the tree by 1/3 then they would have to reduce SP by 1/3 also and people would go nuts about that too.

I Mastered 11 Mechs on PTS. I can adjust to the changes. I personally love the GSP compensation module as it is great benefit to my account. I am glad that it looks like the ST will go LIVE on May 16th at long last. I want to get this in the game where it can be refined and tuned along with the quirks on the LIVE server where there is a population to give some decent data that can be used to accomplish those task. I am tired of all the delays and cancellations.

[Insanity09]

My understanding was that one of the reasons for replacing the skill tree was to stop forcing people to buy unwanted/useless skills. (e.g. pinpoint, or arm reflex for mechs w/ no arms).

Instead of fixing the problem, PGI doubled down, and there are forced choices of useless/unwanted skills all over the place (except for the jump jet tree). Not only are you forced to buy skills you don't want or need, how you do it feels haphazard, like being nibbled to death, or at least annoyance, by ducks.

How many people actually bought and used improved gyros or hill climb modules? Target decay for non-LRM (& sSRM) mechs? Sensor Range?
Crit rate reduction, previously an oddball quirk on a handful of mechs is now required for anyone who wants to delve into survivability?

Honestly, some of these issues sound like choices made by prideful and vengeful designers.
"Darn it, I designed this imp gyro module, it's good, but nobody wants to use it! I'll show them, I'll force them to use it."
I think the designers should be honest with themselves, and us, about the motivations here.
(on the subject of imp gyros, if they were enhanced to also reduce screen shake from jj's, which they should do, I think a lot more people would be interested, but anyhow...)

I understand about forcing people to make trade-offs. I'm highly in favor. But, requiring people to make purchases they do NOT want, multiple times in a single build, and repeated for every single mech. That will be a constant pain point.

So, an idea. The devs want gating. Fine. Determine what number of less desirable nodes are required in each tree and array the choices in tiers. Set it up more like the traditional (hey, it works) skill/spec trees in mmos.
As an example, in agility, let tier one be a choice between 5 points of hill climb or gyros. Each player chooses which they would rather have. You must spend 5 points in tier one before you can buy anything in tier two. Second tier is an array of torso twist/pitch/speed and arm related stuff. Again spend 'x' in tier two before you can reach tier three. Tier three could be turning/braking/accell. Spend 'x' in three to get to tier four, which is all about speed tweak.

You've still gated entry to the "better" nodes, but now you've allowed people to choose where they want their gating, a vast improvement.

Other possible factors:

• set a maximum points spent in a given tier
  
• have an actual tree mechanism, where once you go down one branch of the tree, you can't spend point in another (i.e., in firepower, pick one to spend in of e/b/m, for jj's pick forward or vertical, etc.)
  
• spending points in one area can lower the maximum spent in another (more survivable = less jj/mobility; more firepower = less sensors) (I fully expect that would be unpopular, but hey, choices matter, and should have consequences)

Edited to correct skill to quirk for crit reduction.

Edited by Insanity09, 30 April 2017 - 10:25 AM.

[Deathlike]

The funny thing is that the PTS "functionally" works best on "PGI Time", so there's no reason to take your useful feedback when the weekend is ideal for most players.

[Deathlike]

Another way of looking at the problems of the tree is quirks... specifically meta unquirked Clan mechs vs quirk-dependent IS (and some Clan) mechs (that are getting nerfs to their quirks)... that still has been unresolved.

Edited by Deathlike, 29 April 2017 - 12:01 PM.

[Satan n stuff]

KodiakGW, on 29 April 2017 - 07:03 AM, said:

Another thread like this. I'll reiterate what I posted in another one.



They should put whoever made the Jump Jet tree in charge of redesigning the rest, and keep everyone else out of that process. EVERYONE ELSE, INCLUDING MANAGEMENT. Then release that to test. We'll have a Skill Tree that a lot more will be happy with, or at lease content.

Hell no, the jump jet tree is awful. The ability to jump a little bit higher shouldn't require me to buy nearly the entire damn tree. Out of 20 nodes only 10 contribute to "jumping higher" but you have to buy an additional 7 to actually be able to buy all 10 of them, that's the worst ratio of useless to useful nodes I've seen yet.

[Insanity09]

Compared to most of the other trees/webs, the jump jet tree is efficient and non-punitive.

(since I can't look at it right now, I'm working from blurry memory).
Iirc, all of the vent calibration nodes are in a direct line with each other. All those contribute directly to higher jumps (longer jump time = higher). You just need to buy one heat reduction to get into the vent cal path.
Again, those 5 nodes are the ones you want if you want to jump higher.

So, if you just want to maximize your jump height , that's five nodes all in a direct line with each other, plus one heat reduction node to unlock the chain. (6 points total)
Done. Quick, simple, and easy.

The lift speed nodes merely get you to your maximums faster, they don't add to your height, per se. They would effectively allow you to hover for longer at your full height (you got there faster, so you can stay there for longer on a single jump).

Every node in the jump jet tree does something useful for a jumping mech. You might not care about jumping heat reduction or forward vector, but at least they are beneficial and you are given the actual choice. Other trees/webs are not so kind. Not by a long shot.
By comparison most other trees are harsh; almost every other node is one you probably wouldn't take if you didn't have to, and in some cases you must buy nodes that are absolutely worthless to you. (e.g., arm speed for a no arm-weapon mech)

Edited by Insanity09, 29 April 2017 - 01:36 PM.

[KodiakGW]

Satan n stuff, on 29 April 2017 - 12:38 PM, said:

Hell no, the jump jet tree is awful. The ability to jump a little bit higher shouldn't require me to buy nearly the entire damn tree. Out of 20 nodes only 10 contribute to "jumping higher" but you have to buy an additional 7 to actually be able to buy all 10 of them, that's the worst ratio of useless to useful nodes I've seen yet.

I was looking at it more by the fact that it is much more linier than any of the other ones, which have similar nodes spread all over the place. I'm sure management had them split Lift Speed for the same reason why they locked nodes behind other nodes like Reinforced Casing (don't get me started about RNG in an FPS shooter), Arm Pitch (on a BNC-3M?), Improved Gyros (I can shoot fine while under dakka fire, thank you), and Hill Climb (which should be much higher percents to be useful when fully unlocked). I agree that it should be an offshoot on one side or the other.

That is what I meant. If we can get other trees closer to that, I think we will be better shape since it looks like this is going to get rammed into live next patch.

[JC Daxion]

chucklesMuch, on 29 April 2017 - 12:24 AM, said:

Hill climb is less of an issue for me since it offers something (I guess)... the arm ones or weapon/ams buffs that effect nothing equiped on my mech... seriously?

The arm ones I really don't like, as I would prefer to reduce how they flap around not increase it. (In pts I default to locking them, in live I don't). So I'm 'paying' for an upgrade that is reducing my enjoyment :/

Well the easy fix on this is to really slow down torso twisting, to make mechs with ONLY weapons in the torso MUCH harder to track lights/fast moving mechs.. Not basically making torso only mechs work almost better in many cases because of shield arms and weapon grouping.

That is the biggest issue with the arms, the way the game is, it does not make arm weapon mounts better outside of sometimes being able to reach targets above and bellow. very rarely does the arm flex sideways help in hitting a mech for the most part.


But if you think having to use a couple of nodes in arms to get what you want is an outrage.. Can you imagine if they made it harder for torso only mounted mechs to track lights? the HORROR!


Fact is, You can't just let people with torso only weapons to skip out on points, and make arm mounted weapon mechs, NEED to spend extra points to get said bonuses. You are effectively make any mech with arms have to spend extra points, that is not a good balance feature..

The other thing PGI needs to do, is Get ride of ARM-lock, AND slow down arm tracking to make these skills actually mean something. which could and should be done...



Hill climb is actually usefull on live if you play fast mechs without JJ's, but it is a 10% boost. I think i read on test 7.5 is the cap on that. I'd actually like to see the cap increased, and put a few extra nodes, or swap a bit, so if i spend a couple more points i can actually boost it further. It could really help mechs like the Stormcrow, cicada, commando and others get around much better.. Maybe even those 80-90 KPH non-JJ mechs as well.

But again, to make those few mechs need to spend extra points again would be bad for balance if they are not included in the base tree.

It is a movement tree.... NOT a torso twist tree. arms, hills, twist, ext that is all movement.


The over all effectiveness of certain trees, is all about balancing the numbers to make different builds viable. and for that they need the tree to go live to see how people are using them, and how the numbers all stack up in live 12 v 12..


Quirks can also be added to under performing mechs, which is again a balance thing, which needs live data..

Edited by JC Daxion, 29 April 2017 - 02:06 PM.

[MauttyKoray]

MischiefSC, on 28 April 2017 - 07:00 PM, said:

Okay. I really appreciate that PGI is continuing to iterate on this. I do.

I also recognize that some of my prior responses were a bit heated. However those concerns were ones voiced universally from the release of the first PTS about it. I know I suggested Tina actually slap people across the face with a wet rag and I realize that was inappropriate, Tina seems like a very nice person and not the 'slap them with a wet rag' sort.

I'm totally cool with that being done by Matt Newman. Nobody who posts regularly on the forums is going to have too many objections to 'Wet Rag of Community Disapproval' duty.

Critically though -

Stop with the gating.

Stop it.

Don't do it.

Nobody is going to like it and it's not going to do anything for you that another method won't do better.

It flat out prevents us from making meaningful compromises in balancing skills. It forces us to pay all the way up to and past the gate (and all the unwanted 'Universal' nodes) to get the ones that are worthwhile.

The only thing this will accomplish is to expand the problems already created by 'good mechs' (what people call metamechs) playing in matches vs 'bad mechs' via introducing a system that more or less requires you to stick to a pretty narrow path to get into anything good.

It does not create a balanced set of tradeoffs to decide between - in fact it prevents useful tradeoffs.

You want to keep people from min-maxing to get max bonus on weapons? Cap weapon bonuses at 7.5% or whatever. Establish a relative value between them and make the otherwise 'less valuable' stuff worth more per point than the 'valuable' stuff. Have speed tweak only 0.5% per point up to 5% but Hill Climb 7.5% up to to 30% - so there's 10 Speed Tweak nodes but only 4 Hill Climb, so it's worth it for me to only do 8 or even 6 Speed Tweak (as the cost is small) to get some very useful mobility in a lot of situations.

In the current gated system I can't do that. I have to give up heat management stuff (which is a gotta have on most mechs) in order to give up the speed tweak, but I can't spend that on other mobility stuff, I have to spend it on sensors or the like - and I'd need to back out 5SP from, say, Mobility, just to get 4 minimally useful sensor quirks and 1 pt of Derp.

That's never going to be a smart trade. Ever.

Stop gating PGI. Don't do it. You want to gate weapon specific stuff behind universals? That's fine. However gating is stopping us from doing with the skill tree what the skill tree should literally be there to do.

Some gating has to be done otherwise we'll just get a min/max tree of garbage. The PTS shut down notes however are exactly what I wanted to see. We'll get LESS of the gating and be able to invest further into those general skills without having to make the sacrifices. The gating isn't the problem, it HAS to be done in order to prevent Min/Max trees, but the amount of it especially early on needed to be reworked.

I would honestly like to see this a bit more applied to the other trees however and allow you to make some investment progress in other areas without having to sacrifice points right away. Specifically the movement and survivability aspects.

As for the whole 'XP sink'...yeah and? Its a F2P game, it first of all needs a sink, and the amount of time it takes to XP a mech should mean you've become 'proficient' with that mech in play which is then reflected in additional rewards/boosts allowing you to further tweak your mech's performance to fit your play style.

The old skill system was horrid, and the 2x basic bonus is honestly the thing that broke the game the worst in hand with the Engine-Agility coupling. The game's balance has become so inflated that people are gonna whine but it needs to be broken down to the basics and started over. The skill tree, removal of many quirks, and engine-agility decoupling are what look like the first steps towards this process.

Edited by MauttyKoray, 29 April 2017 - 02:25 PM.

[MischiefSC]

MauttyKoray, on 29 April 2017 - 02:24 PM, said:

Some gating has to be done otherwise we'll just get a min/max tree of garbage. The PTS shut down notes however are exactly what I wanted to see. We'll get LESS of the gating and be able to invest further into those general skills without having to make the sacrifices. The gating isn't the problem, it HAS to be done in order to prevent Min/Max trees, but the amount of it especially early on needed to be reworked.

I would honestly like to see this a bit more applied to the other trees however and allow you to make some investment progress in other areas without having to sacrifice points right away. Specifically the movement and survivability aspects.

As for the whole 'XP sink'...yeah and? Its a F2P game, it first of all needs a sink, and the amount of time it takes to XP a mech should mean you've become 'proficient' with that mech in play which is then reflected in additional rewards/boosts allowing you to further tweak your mech's performance to fit your play style.

The old skill system was horrid, and the 2x basic bonus is honestly the thing that broke the game the worst in hand with the Engine-Agility coupling. The game's balance has become so inflated that people are gonna whine but it needs to be broken down to the basics and started over. The skill tree, removal of many quirks, and engine-agility decoupling are what look like the first steps towards this process.

any and all useful "tradeoffs" *are* min/maxing. That's why you would take one thing over another. The lack of reason to trade one for another because of gating is why it doesn't work.

Tradeoffs happen when I can give a bit of something for a bit of something else and they're of equal value. Giving up 5 points of Speed Tweak and useful mobility quirks to get 4 useless sensor quirks and 1 pt of Radar Derp isn't that. It's a trick to sucker bads into bumping their mechs so they're at a bigger disadvantage.

[Insanity09]

Yes, Mautty, it is true that balance is currently quite a mess. IS < Clan. Many mechs (both clan and IS) are awful compared to others (both IS and clan). I'd don't think anyone could sensibly argue that wasn't the case.

You say quirks are part of the problem, not the solution. I disagree, but would like to read your specific arguments for that.
I will, possibly, agree that quirks did not solve the problem completely, but whether that was the nature of the quirk system, or their specific implementation... data is needed.

However, what many folks, including myself are currently advocating for is not necessarily that engine decoupling not happen, or that quirks not be adjusted, or even that the skill#2 not be implemented (though many, including myself, believe it needs more work to be a reasonable change).

We are saying do each of those things one at a time and assess the results.

Suppose all those things are done and balance gets (much) worse. Then what? Which change made the situation worse? All of the above? Which contributed how much to the problem?
Suppose (and here is it hard to type due to shaking my head in disbelief) balance is improved, but there are still obvious issues? The same questions pertain.

Better to implement one change group at a time, assess, adjust, and then add the other changes in over time, assessing and adjusting each as they come. That has a much higher likelihood of long term success.

[Napoleon_Blownapart]

how about skill trees that are mech specific (i only got to spec 1 mech) like a mech with no missle hardpoints doesnt see missile skills in its tree at all?or a mech with no ballistics doesnt have to take ballistic nodes.

Edited by Gorantir, 30 April 2017 - 08:58 AM.