

The Skill Tree (A General Discussion Review): Too Expensive, Too Grindy, Too Much Waste, Not Enough Customization.
#1
Posted 09 February 2017 - 06:03 PM
So, after playing on the PTS for the afternoon, these are my conclusions.
The Skill Tree system is not a bad idea in theory.... but it's execution is lazy, sloppy and the Devs are being pretty dang disingenuous in how they present it, as it certainly will cost us a hell of a lot more, across the board, than previously.
I like the idea of there being Various Skill Trees, and you not being able to just max out everything. But.... they do it in as generic and lazy a manner possible.
-They do nothing to promote the actual Roles of individual Mechs
-It's the same options for everyone... to the point where your mechs with NO LOWER ARM ACTUATORS... are still required to waste Skill Points on Nodes dedicated to Arm Yaw. WTF?
-The Node Trees are not remotely Linear, and require you to cross pollinate on Skill Nodes you may have ZERO desire to spend your points on.
-91 Skill Points seems like a lot... it's not. Almost everyone is going to want to open up 90% of the Mobility and Operational Skill Nodes... and there goes 1/2 to 2/3 of your Points there. Then you get to decide between spending on super itemized weapon systems, most will want 15 of the 17 Defensive Nodes.... and that is without even touching InfoWar.
- Too expensive, requiring XP, GXP and Copious CBills expenditures, well beyond what one has to spend now, even with the 3 way XP Trees.
-The New User Experience is going to Hell in a Handbasket. The Time, XP and Cost needed to grind out a new chassis... is going to leave the Noobs farther in the dust than ever before.
-A lot of the actual percentages are still completely ludicrous when one maxes out.
Mind you, I'm not saying all is lost. Please don't do like you did with the InfoWar PTS, get sulky and chuck the whole thing because of a few butthurt Cryhards. Take the critiques and ADJUST THE DAMN THING.
-Diversify the Skill Trees, to actually promote the intended ROLES of each Mech. Scout, Sniper, Skirmisher, Juggernaut, etc.
http://www.sarna.net...attleMech_roles
-Streamline the Node Trees themselves so we are not forced into purchasing Nodes we do not want, unless they directly impact the next level of a Node. I should not have to unlock Arm Yaw to affect Torso Speed, etc. And don't force Mechs without Lower Arms to waste already scarce skill points on bloody Arm Yaw to get to torso Quirks.
-Re-evaluate the Prices. Realize there is a hell of a difference between the Smurfy Cost of a Maxed Out Mech, and the REAL cost, since most of us have not and will not EVER buy an individual Seismic, Radar Derp, etc, for each and every one of our 200 Gundams, no matter how you try to dress it up. Please for once, respect our intelligence.
PGI... you have a chance to do things real well here, or to ship it out halfassed. The idea is not bad, the execution, as seen in the PTS? Is execrable. (And that's not even getting into the real poor UI)
#2
Posted 09 February 2017 - 06:08 PM
25-30+ skills ought to do it.
I think 91 is too few too, maybe 110 to 120 would be good.
I'm with you on the useless skills. We need actual separations of roles. I don't need Fall Damage reduction or AMS Overload in my other mechs but from the feedback I learned that you're forced to waste a few skill on barely useful skills.
#3
Posted 09 February 2017 - 06:11 PM
If the tree switches over to linear, parallel branches, I would expect the number of points per skill to vary relative to the effectiveness of that skill.
I'd like to think I address the other problem aspects of this system in my post here: https://mwomercs.com...99#entry5609499
#4
Posted 09 February 2017 - 06:11 PM
I skipped the upper chassis section because there's no way I'd bother getting arm quirks (torso twisting is important, but mechs with torso mounts don't give a damn about arm quirks).
I know Paul acknowledged the tree section with the consumables... but how did that even get there in the first place? Forgetting what MC consumables do compared to their C-bill equivalents is akin to saying "the left hand doesn't know what the right hand is doing". I don't even touch MC consumables, and that was forgotten so quickly?
There is some level of integration with the modules noone uses.. but that says a lot about those modules (360 target retention, hill climb, speed retention) that they aren't really balanced amongst themselves (they are simply not "equally good options" compared to the rest).
Oh well, it'll fall on deaf ears.
#5
Posted 09 February 2017 - 06:12 PM
Here are some numbers if you want to reference costs for things (I didn't get to every weapon tree, but they are a mostly all designed the same so it's easy to eyeball the costs for a particular).
https://docs.google....yzpQ/edit#gid=0
If anyone wants to add those trees I skipped (Missiles, LB ACs, Auxiliary) - PM me your e-mail and I'll give you access.
Edited by Ultimax, 09 February 2017 - 06:48 PM.
#6
Posted 09 February 2017 - 06:14 PM
Edit - Great post Bish, now they just need to read it and learn from it, perhaps a course in designing a modern UI would help too.
Edited by Carl Vickers, 09 February 2017 - 06:15 PM.
#7
Posted 09 February 2017 - 06:14 PM
I'd like for each tree to have more nodes so I could do something crazy like invest all of my points into pure defense or whatever. Right now, role specialization is impossible because we're forced to pick four to five full trees. Right now, every mech gets to pick almost everything.
Edited by FupDup, 09 February 2017 - 06:15 PM.
#8
Posted 09 February 2017 - 06:15 PM
Deathlike, on 09 February 2017 - 06:11 PM, said:
I skipped the upper chassis section because there's no way I'd bother getting arm quirks (torso twisting is important, but mechs with torso mounts don't give a damn about arm quirks).
I know Paul acknowledged the tree section with the consumables... but how did that even get there in the first place? Forgetting what MC consumables do compared to their C-bill equivalents is akin to saying "the left hand doesn't know what the right hand is doing". I don't even touch MC consumables, and that was forgotten so quickly?
There is some level of integration with the modules noone uses.. but that says a lot about those modules (360 target retention, hill climb, speed retention) that they aren't really balanced amongst themselves (they are simply not "equally good options" compared to the rest).
Oh well, it'll fall on deaf ears.
Oh yeah the upper torso skills are basically useless compared to the lower torso skills.
I'd rather get turn rate than torso yaw rate. Also I've only ever been bothered by torso pitch angle, not yaw.
Nobody in their right mind would get torso twist skills over accel, deccel and speed tweak.
#9
Posted 09 February 2017 - 06:18 PM
Ultimax, on 09 February 2017 - 06:12 PM, said:
Here are some numbers if you want to reference costs for things (I didn't get to every weapon tree, but they are all designed the same so it's easy to eyeball the costs for a particular subtree).
https://docs.google....yzpQ/edit#gid=0
If anyone wants to add those trees I skipped (Missiles, LB ACs, Auxiliary) - PM me your e-mail and I'll give you access.
I do have experience with trees like this (Path of Exile has an awesome system - and it still has its own faults/errors/BS) and min-maxing that stuff in MWO feels kinda like child's play comparatively speaking.
FupDup, on 09 February 2017 - 06:14 PM, said:
I'd like for each tree to have more nodes so I could do something crazy like invest all of my points into pure defense or whatever. Right now, role specialization is impossible because we're forced to pick four to five full trees. Right now, every mech gets to pick almost everything.
I think you can reasonably pick about 70 or so points for any mech, and then specialize in just "one weapon" in a similar vein to what we already have in weapon modules (some have it @ 1 weapon system max).
I think some things can be tweaked/optimized for two, but I've already gotten a cheap-lazy ACH tree with pulse laser quirks, Full Radar Derp, Full Seismic (200m worth, which happens to be a nerf), and a host of other stuff... and it wasn't like it was a "choice" than it was "mandatory".
Edited by Deathlike, 09 February 2017 - 06:18 PM.
#10
Posted 09 February 2017 - 06:25 PM
FupDup, on 09 February 2017 - 06:14 PM, said:
I'd like for each tree to have more nodes so I could do something crazy like invest all of my points into pure defense or whatever. Right now, role specialization is impossible because we're forced to pick four to five full trees. Right now, every mech gets to pick almost everything.
considering that as presented all half way decent players will specialize in the exact same nodes, and ignore the other 2/3 of the options entirely.
-Max Defense, Minus AMS Nodes
-Max Lower Movement Stuff
-Pick and choose some upper movement MAYBE
-Max out Operational mostly
-Grab just enough info tech to access Radar Derp and Seismic
-Pick ONE Weapon Sub Tree, Maximize.
I don't see things ending up that generalist once the dust settles. Because just like the Modules now... not all the Nodes are created equal. And if they were to actually diverisfy the Trees to truly reflect Roles, we'd have a whole hell of a lot more options, and thus, be spread even thinner.
As it is, like I said... I applaud that we are not able to maximize everything. But what WILL be maximized, if this is the Skill Tree going forward, is pretty cut and dried, and will be a monolithic choice.
You'll likely see a less than 10% Skill Node divergence across the board with the higher tier players, as the game doesn't reward scouting enough to convince our Lights to do things like actually maximize Infowar, etc. If I felt like spending the rest of the evening on it, I could probably doctor up the official MetaMechs Comp SkillTree within a 95% accuracy. Because of how many of the options are really a no brainer, and which are a no reason to take, or simply nowhere near as comparable.
Deathlike, on 09 February 2017 - 06:18 PM, said:
I do have experience with trees like this (Path of Exile has an awesome system - and it still has its own faults/errors/BS) and min-maxing that stuff in MWO feels kinda like child's play comparatively speaking.
I think you can reasonably pick about 70 or so points for any mech, and then specialize in just "one weapon" in a similar vein to what we already have in weapon modules (some have it @ 1 weapon system max).
I think some things can be tweaked/optimized for two, but I've already gotten a cheap-lazy ACH tree with pulse laser quirks, Full Radar Derp, Full Seismic (200m worth, which happens to be a nerf), and a host of other stuff... and it wasn't like it was a "choice" than it was "mandatory".
I'm sure there is a Reddit somewhere with it already spelled out.
#11
Posted 09 February 2017 - 06:30 PM
Bishop Steiner, on 09 February 2017 - 06:24 PM, said:
-Max Defense, Minus AMS Nodes
-Max Lower Movement Stuff
-Pick and choose some upper movement MAYBE
-Max out Operational mostly
-Grab just enough info tech to access Radar Derp and Seismic
-Pick ONE Weapon Sub Tree, Maximize.
I don't see things ending up that generalist once the dust settles. Because just like the Modules now... not all the Nodes are created equal. And if they were to actually diverisfy the Trees to truly reflect Roles, we'd have a whole hell of a lot more options, and thus, be spread even thinner.
As it is, like I said... I applaud that we are not able to maximize everything. But what WILL be maximized, if this is the Skill Tree going forward, is pretty cut and dried, and will be a monolithic choice.
You'll likely see a less than 10% Skill Node divergence across the board with the higher tier players, as the game doesn't reward scouting enough to convince our Lights to do things like actually maximize Infowar, etc. If I felt like spending the rest of the evening on it, I could probably doctor up the official MetaMechs Comp SkillTree within a 95% accuracy. Because of how many of the options are really a no brainer, and which are a no reason to take, or simply nowhere near as comparable.
What I'll pick for most mechs:
-Full Lower Chassis tree
-Most of Operations for Cool Run and Heat Containment. Maybe Magazine Capacity if using a ballistic mech
-Most of one specific weapon tree based on build (ignoring cooldown because 5% is too low to be worth it)
-About half of InfoTech to get Seismic Level 2 (I personally skip Radar Derp)
-Finish off with Survivability
So, to recap. My mech has enhanced firepower, enhanced mobility, enhanced sensors, enhanced durability, and enhanced "general operations" all at the same time. If I do this, what it my role? What is my specialty?
A specialized skill build would be something like maximum tanky, super mobile ninja, glass cannon, super duper electronics scouting platform, etc. Right now the trees don't have enough nodes to let us do this.
Edited by FupDup, 09 February 2017 - 06:34 PM.
#12
Posted 09 February 2017 - 06:34 PM
FupDup, on 09 February 2017 - 06:30 PM, said:
-Full Lower Chassis tree
-Most of Operations for Cool Run and Heat Containment. Maybe Magazine Capacity if using a ballistic mech
-Most of one specific weapon tree based on build (ignoring cooldown because 5% is too low to be worth it)
-About half of InfoTech to get Seismic Level 2 (I personally skip Radar Derp)
-Finish off with Survivability
So, to recap. My mech has enhanced firepower, enhanced mobility, enhanced sensors, enhanced durability, and enhanced "general operations" all at the same time. If I do this, what it my role? What is my specialty?
A specialized skill build would be something like maximum tanky, super mobile ninja, glass cannon, extreme scout, etc. Right now the trees don't have enough nodes to let us do this.

Edited by Deathlike, 09 February 2017 - 06:34 PM.
#13
Posted 09 February 2017 - 06:38 PM
FupDup, on 09 February 2017 - 06:30 PM, said:
-Full Lower Chassis tree
-Most of Operations for Cool Run and Heat Containment. Maybe Magazine Capacity if using a ballistic mech
-Most of one specific weapon tree based on build (ignoring cooldown because 5% is too low to be worth it)
-About half of InfoTech to get Seismic Level 2 (I personally skip Radar Derp)
-Finish off with Survivability
So, to recap. My mech has enhanced firepower, enhanced mobility, enhanced sensors, enhanced durability, and enhanced "general operations" all at the same time. If I do this, what it my role? What is my specialty?
A specialized skill build would be something like maximum tanky, super mobile ninja, glass cannon, super duper electronics scouting platform, etc. Right now the trees don't have enough nodes to let us do this.
as you noted, one single weapon tree, and the same basic gimmes on everything else. Doesn't sound that specialist. And actually we do have enough nodes to do those things... it's just you have to give up a lot somewhere else to do it. And most of those other things, again, are nowhere near as important as the very things you just listed. You can call it generalist if it makes you feel better... but when everyone will be running the exact same thing except possibly which weapon tree they specialize in.... it's generalist in name only. Because everyone is still specializing in one thing: Gunbag.
#14
Posted 09 February 2017 - 06:41 PM
Bishop Steiner, on 09 February 2017 - 06:38 PM, said:
YOU TAKE THAT BACK!
Our balance overlord designed this precious tree and he won't go down w/o a fight with that word you use.
Complete and utter heresy!
#15
Posted 09 February 2017 - 06:48 PM
Bishop Steiner, on 09 February 2017 - 06:38 PM, said:
What I mean by specialist in this case is spending my Skill Point pool on one or maybe two trees. Right now the system makes us pick four to five nearly full trees.
Let's pretend that I got to be the Balance Overlord and have my way with the new Skill Tree. I'd make each tree have a bare minimum of 50 nodes (likely higher), compared to each tree currently having only around 20 nodes.
Let's go full potato and pretend that it each tree allowed the full 91 SP in it. The value per node would be different in some cases, but overall the maximum possible values would be higher if you invested into the full tree.
For example, yes a lot of people would probably spend all 91 of their SP on one weapon system. In this case, they are now specialized into one role. They get no agility buffs, no durability buffs, no sensors, etc. Just pure firepower. They do one thing and one thing only. Glass Cannon in this case.
As another extreme, let's say that you decided to build your Atlas to have all 91 of your SP in the survival tree and nothing else. Now you're going to be a literal wall of armor, but you're going to be clumsy and not have all that much firepower. Once again, you do just one thing and one thing only.
As a more balanced example, let's say that a light mech user splits his skills to have like 46 weapon and 45 mobility to try to be a ninja.
Ideally, the values should be balanced such that some people would even choose to go 50+ skill points into the InfoTech tree. If somebody goes all the way with 91 SP in IT, I honestly think they should get the old MW game radar that goes through terrain in all directions with very good range. Suddenly, a dedicated scout that has no firepower at all and extremely low armor can be incredibly powerful and decisive in the match. Of course, if he gets caught in a firefight then it's ggclose for him.
Etc. That is what specialization and roles look like, not "pick any five trees of your choice." You would still have the option to spread across a lot of trees of course, but your build won't have a clear strength or role if you do that.
Edited by FupDup, 09 February 2017 - 06:50 PM.
#16
Posted 09 February 2017 - 06:52 PM
FupDup, on 09 February 2017 - 06:48 PM, said:
I don't think it's even intended to specialize in this way.
This is their way of saying "If you want to specialize in being a scout, then you take the sensors tree at the cost of something else like weapons).
I'm not saying that makes sense in MWO, but I also don't think specializing in some "super defense tree" and like one other tree make sense either.
I also think it would cause a riot, because as it is we are already being nerfed from what the current tree gives us. I need to check the numbers but I don't think we actually achieve the full values we have now.
#17
Posted 09 February 2017 - 07:03 PM
So instead of playing the PTS, I'm spamming x2s for everything because it's time to grind.
I see a lot of other problems with it, but they're theoretical on my part (Though I see a lot of other people complaining about them on my behalf).
Oh, and it seems like the quirks that are staying have been shotgunned randomly, some mechs keeping an avalanche of them, some mechs losing everything, which has the predictable effect of making them inferior to Clan mechs by even greater margins.
Edited by Bombast, 09 February 2017 - 07:04 PM.
#18
Posted 09 February 2017 - 07:41 PM
However... the numbers are the numbers; no playtesting is needed to see that the costs are stratospheric compared to what is currently required to max out a 'Mech. 9.1 million cbills on top of the XP cost, which itself is multiple times what the current tree costs? That's obscene. I experiment with my 'Mechs a lot and I don't get to actually play very often (1-2 nights per week, if that), so I'm constantly cbill starved. Currently, that's not a problem because I can still grind out XP at a reasonable pace and just swap around hardware as needed. With the new tree? Now there's a hard cbill cost for each and every variant that cannot be defrayed, on top of the equipment and upgrades it already requires to perform at an acceptable level. Even with a full refund for modules, I'm looking at being able to re-master about half a dozen 'Mechs... and let's not even talk about all the partially mastered variants. Half a dozen, out of ~100.
The quirks granted by the new tree also massively increase the performance gap between dealer-fresh new 'Mechs and mastered ones, which is an utterly horrible idea any way you cut it. For some of the new bonuses being 50% or greater... you might as well ask new players to drop in SHS laserboats, for the capabilities of their equipment vs. vets in fully mastered minmaxed meta cheese. That's insane. Totally bugnuts.
If the skill tree gets pushed to live in anything resembling its current form, I'm done. It won't be worth the expenditure in time to get caught back up to where I already was, let alone ever buying anything new.
Edited by WrathOfDeadguy, 09 February 2017 - 07:42 PM.
#19
Posted 09 February 2017 - 07:43 PM
Where did all the months go? This is the old quirks system with a generic skills layer plastered on top. With the added bonus of killing optimised build experimentation in the mechlab because everything inexplicably costs c-bills to change.
What a wasted opportunity for something fun that differentiated and specialised all the mechs.
#20
Posted 09 February 2017 - 07:57 PM
1 user(s) are reading this topic
0 members, 1 guests, 0 anonymous users