152 replies to this topic

[Horseman]

Nimnul, on 11 February 2017 - 04:27 AM, said:

I like it: it's a step in the right direction.

Piranha do not listen to all the whiners. Do this as needed. no matter what.

It's a step in the right direction, but the implementation as we currently see it is two steps backwards - it encourages boating more than the quirk/module system it replaces, and provides less room to specialize.

[PyckenZot]

Just did my first steps with the new Skill Tree on the test server today.
The major improvement I would like to see so far in order to give a better view on what one is actually changing to a mech is a stats page that shows current quirks + what's added from modules + what the skills add.

I expect A LOT of maxed out boating in the future

[C E Dwyer]

this is a link to Fantastic Tuesday's you tube review



I think it pretty much encompasses everything that is good, bad and needs rework, and I agree with pretty much all his comments

I think everyone should look at it, and I hope the Dev's do also and take note

[ATR0P0S]

[AUXILIARY] needs some stuff for airstrike an artillery;
Fall damage nodes transfer to [JUMP JETS];
Gyros nodes transfer to [UPPER CHASSIS];
ECM nodes needs some improves like extra range of cover area or reducing counter-ECM effect from PPC;
Different value of bonuses for different class of mech's, something like that:

[Lights mech] torso speed bonus is 0,25% every node.
...
[Assault mech] torso speed bonus is 1% every node.
And another ideas like "lights heat containment" > "assaults heat containment" BUT "lights cool run" < "assaults cool run".

[MeIsYou]

Dear PGI developer,

I really like the skill tree idea, because it will give you flexibility when costumizing a specific mech. I therefore won't discuss every detail of the skill tree and how it may affect the game. Instead I would like to highlight one issue I just couldn't believe when I saw it:

After spending 9.1M CBills to unlock 91 skillpoints, the player must spend CBills again to reassign skillpoints? This approach takes the flexibility away, which I loved since I started to play the game and which the skill tree could take to the next level.

My suggestions is: leave the skill point costs of 1500 XP and 100k CBills but remove the costs for skill point reassignment (i.e. once unlocked, a skillpoint can be used anywhere in the skilltree without additional costs). This will allow experiments and a flexible mech design.

[Aleski]

As some as always says in the topic, one of the big issue of the new skill tree is the cost of the respec.

One of the big pleasure and feature of MWO (and other battletech games) is to be able to play with different loadouts, try new stuff and etc... Making the respec not free is a huge step in the wrong direction. It kills the fun in the game and create a massive C-bills Black Hole !

Some peolpe argue that it's not that overpriced to put all the nodes on a mech, cause the modules are currently too expensive (6M for Radar Dep for example). It's completely wrong : you can swap the modules in game ! You can't swap the new skill nodes point with this new system... Need to rethink that part of the new system.

One good thing should be to made the respec free and reduce the amount of price and XP to buy one node point.

For some ideas, they should made more nodes for bad variants and redesign some trees. It is stupid on some trees like upper movement to have the obligation to take some nodes on yaw arm movement on some chassis that don't need it for example. It's the same when you want to pick seismic, you HAVE to take some useless nodes.

[ArmageddonKnight]

Greetings.

WOULD APPRECIATE FEEDBACK

This is some of my opinions bassed on my experience on the PTS in regards to the new skill system.

To stop yee wall o text im using spoilers

Overall

Spoiler

First and foremost, the layout of each individual bonus within a skill tree is bad.

The current system on PTS forces users to put points into possibly useless bonuses that they dont want or may not even need.
For example, Radar Deprevation is somthing everyone wants, but you have to unlock almost the entire tree to get it, and UAC jam chance reduction, any1 that uses alot of UAC's or uses the smaller variety which have fast CD's will NEED jam chance reduction, but again you have to unlock the usless CD reduction and velocity bonuses to get it.

Same goes for almost eveything in the current skill tree system on PTS, u ether unlock the entire tree, or dont touch it.

This is NOT going to give much variety on any given mech. Everyone will need or want speed tweek, all the bonuses to torso and arm turn speed/amount, heat managment bonuses, and the armor bonuses etc etc. But eveyrone will have to unlock the entire tree for said bonuses, leaving very few , if any, points available for variation.

Specifics

Spoiler

Where is the replacment bonus for weapon Convergence ?

AC Magazine capacity bonus seems to be 15% to 20% bonus to ammo.
BUT
AC2 and UAC2 only get 13.4% increase. And this is a weapon that has a low CD so it churns through ammo.
It looks even worse if u look at the bonus damage your getting, with U/AC2 only getting 10 rounds resulting in 20 damage while U/AC5 gets 50 damage, U/AC10 gets 100 damage, U/AC20 get 80 damage ..and so on. The U/AC2 is getting shafted here.
Ether choose a flat % increase across the board like 20%, or a flat damage bonus across the board like 40 (so AC2 would get 20 rounds, AC5 8 rounds, AC10 4 rounds, AC20 2 rounds.)
Ether way at the very least AC2 as a fast firing low damage weapon should have the highest bonus to ammo capacity. .

Armor Bonus. it is not correctly applying 10% to the mech. If you have 130 Front CT armor and 5 Rear you should get 13 front bonus and 0.5 rear armor, im getting only 11 front and 0 rear. Getting the same odd numbers regardless of front / rear balance or mech size.
IMO the bonus to armor should be given bassed on the max amount of armor for a given part. So if CT is 135 split aross front and rear, then you should get 13.5 bonus armor for the PLAYER to decide where to put it.

UAC2 jam duration change. NO... just... NO. 4 seconds is waaaaaay to long for a UAC2 to be jammed for regardless of the jam chance. On live its 3 seconds, and even that is to long, if anything it needed lowering not increasing.
I honestly dont understand your thought process on this one.
The faster a AC reloads the faster it should unjam, Thus a UAC2 should unjam much faster than a UAC20, since the UAC2 is going to jam more anyway due to its faster ROF.
If you must increase UAC2 Jam duration to 4sec as a base, then replace the crappy 4% C-UAC2 cooldown bonus to -50% jam duration (4x -12.5% Jam duration).

The AC2 skill tree, its usless. I boat 6 Ac2's on a King Crab, i wont be touching the skill tree for them as they wont do sh*t all to their performance. Especialy when more important bonuses for armor, agility, speed, and utility use up most if not all the 91 possible skill points.
What the AC tree does need is seperat bonuses for the AC sizes, for the AC2 in particular it needs a bigger CD bonus and/or a reduction in heat. 0.6 heat down to 0.4 heat (33%) would make a noticable difference, as would 0.72 reload down to 0.5 (30%).
Changing the AC2 cooldown bonuses from 1% to 6.5% would do, and adding in 5 bonuses of -6% to heat would work.

Heat & Alpha meta

Spoiler

:So im thinking PGI need to address the high alpha meta, and this new skill system is the perfect opportunity.

With the smaller CD bonuses people r going to rely even more so on alpha striking, as such, a way to reduce alpha damage without stopping people from chucking on loads of weapons would be the following.

Remove heat containment from the bonus list.
This will lower heat capacity, from a possible 46 back down to the standard 40.
In its place....
Ether:
Buff 'Cool Run' by 50%
Or
In place of 'Heat Containment' have a bonus that adds the same as 'Cool Run' BUT with a negative effect of -3% heat capacity per skill.

The latter idea would result in Double Cool Run bonus with - 15% heat capacity lowering the standard 40 down to 34.
The effect:
A reduction in the amount of damage that can be fired in 1 shot (alpha) but an increase in stagered fire for those who choose the additional cooling efficency skill.
This would promote more balanced builds and allow players to stop being insta killed by multiple high damage alpha strikes.

I dont particularly care about what costs you put on the skill tree are so long as its worth it and gives you choice, the current system does not.

[Kael Posavatz]

Would I like each mech variant to have a completely customized skill-tree, with some nodes coming pre-unlocked in lieu of quirks?

Heck yes.

But frankly I don't think that's a practical solution. So the question then becomes what is necessary to take what PGI has done so far and make it workable?

1) Skill nodes need to follow independent progression paths, whether it be 'armor hardening' or 'X Weapon cooldown'. The paths should be homogenous, allowing acquisition of sequential nodes without necessitating the acquisition of 'other' nodes. This is particularly important for the UrbanMech since the torso Yaw increase nodes fill two rather important cross-over points but will have absolutely no effect on this mech. Also, a complete weapon system (NARC) is buried behind two particularly obnoxious walls (one a risk of a pay-to-win mechanic [see #2] and the other seldom useful).

2) Consumable skill trees need to be made independent of mech skill tree. If part of broad 'pilot' skill tree that tree needs to have no restrictions placed on it. Otherwise, instead of players paying MC for early access to max-performance consumables, they are instead paying for the privilege to use skill nodes to further buff their mech (or potentially pilot). This is the definition of a pay-to-win game mechanic.

3) The c-bill cost to respec a mech needs to be drastically reduced. 125-k is decent payout for a QP game without premium time. That is too high to encourage tinkering, especially for any OmniMechs in the game where tinkering might include removal or replacement of a hardpoint type (ballistic for energy, for example) rather than swapping around specific weapons without a subtype (AC10 for AC20)


Other things that would be nice
a) reduce cost of skill nodes. Yes, the cost is less than buying a full set of modules. But modules could be swapped between mechs, and the skill tree previously came free. Playing without equipping any modules wasn't terrible, but I don't see that being the case with skill nodes

b] 'Generic' quirks need to be renamed since there are no 'generic' weapon nodes. As it stands there are two terminology systems using the same words in different ways. For example, 'Laser Range+10%' quirk would appear to affect only the Standard/ER lasers since that is what the 'Laser' skill nodes affect, rather than a buff to all laser weapons.

c) Magazine buff should be placed with either relevant weapon trees, or if it remains approximately where is, it needs to be made clear that it is a one-time buff, or based on ammunition tonnage. MG, missile, and AMS ammunition should be affected as well.

Edited by Kael Posavatz, 12 February 2017 - 03:05 PM.

[BigScwerl]

Chiming in with the rest of the crowd: I don't like it.

It imposes significant Nerfs on weapon cooldowns, What is the point of a 4% cooldown?

It requires that I spend points on stuff like fall damage.

It will enforce singularly minded builds and prevent any adjustments or customization because it is too expensive.

It should not cost money to unlock a node, Only XP.

Make the categories linear, maybe it would be better...

I like some of the new features (mag capacity, JJ thrust vectoring, etc)

When you impose this skill tree on something like the summoner and remove its quirks, it will be useless.

What about mechs that people money or hard earned c-bills for? now they are unquirked and nerfed?

I present the following example..

I sell you a car for $500 with nice soft seats. Its a crappy car, but the seats make it comfortable at least.....6 weeks later I remove the nice seats and replace them with a couple pieces of plywood. How would you feel about the $500 car? were the seats part of the deal when you bought the car? do you feel like you still own a $500 car? would you want a refund from me for the seats I took?

[Otherium]

I really don't like to pay for things I don't want in order to get things I already have in the current system

[Uncle Totty]

I would also like to see some give-n-take from the nodes.

Such as energy range nodes also adding heat.

Laser duration reduction nodes also reducing range.

Faster running adds movement heat.

ACs, more damage for less range.

SRMs, shorter reload for a bigger spread and so on.

[50 50]

Odanan, on 09 February 2017 - 03:35 AM, said:

Here is my (succinct) feedback:

1- The current system favors weapon boating and less experimentation with builds. Solution: consolidate weapon trees into categories that affect all three weapon types at the same time (skill trees like: Rate of Fire, Range, Heat, Precision).

2- The current system doesn't encourage owning several different variants of the same mech (heck, it doesn't encourage owning several chassis at all) because of the cost in mastering them. Solution: severely reduce the C-Bills cost to buy skill nodes. Let people spend money in mechs, not in mastering. Less mechs = less MCs spent (with mechbays, camos, heroes...).

3- The current system boosts equally mechs that perform very differently. A Kodiak 3 has the same number of skill points than a Victor, which is wrong. Solution: make the number of skill points variable for each mech/variant. This could even (almost) eliminate from the game the need of basic quirks.

4- The UI is complicated. Too many currencies, clumsy XP conversion (why do you need to convert historical XP at all?)... Solution: keep it simple! Intuitive interface, less clicks to do the job, more visible buttons, less currencies, etc.

5- Some Skill Trees are mostly useless, like the "Jump Jets tree" (I mean, really useless). While some are absolutely mandatory. Solution: to merge some trees, balance the nodes, decrease the number of too specialized nodes (but greatly increase their bonuses).

Conclusion: the new skill tree system is a great concept but far from ready to release.

1 - Might work.

2 - Except to get different hard points or have mechs setup for different purposes.

3 - Can't do that. If one mech has more skill points than another it's the same problem as having one mech with better quirks. People simply gravitate towards that mech and the other one never sees the light of day. Plus this is then a difficult item to monitor as it becomes a balancing point on it's own. What happens if it is decided that Mech A has too many points and it needs to be reduced? Try explaining why one mech has more skills than another as well as why it costs more to level up to a new player. I would think that a better option would be to provide a unique tree to the different mechs to give players that choice of taking advantage of that design.

4 - UI is good. At the moment while we convert the historical XP across it has that extra 'currency' but seeing as we can no longer earn historical XP that part of it will disappear in time. It's like this because of the shift from one tree and XP pool for all duplicates we might have of a single variant so we can now split that XP to specific individual mechs of that variant.

5 - Going to disagree again. Now having skills for equipment that we previously didn't have is good. We were missing options for jump jets entirely, now we have them. If you don't want to spend points on them, that's fine, but don't get rid of them. The problem here is that people view certain skills as 'mandatory' regardless of what mech they have or how they might want to experience piloting it. This is the same view that has been imposed on everyone with which mechs we should use and which weapons we should have on those mechs and which modules we should equip etc etc.
I say no.
If we want to break up the single meta, it means more skills trees and less skill points to spend on them. It is currently far too easy to build the mechs with what we had before, with our 4 favourite modules and then add in another 4 modules worth on top of that.
We have too many points.
If we want to make our mech a rockin' weapon platform, it has to be absolute rubbish in other areas. We have to make the hard decision to either mix and match more, OR specialize and therefore miss out on other skills.

[Xozia]

I like this system, I just hope that you (PGI) don't leave it alone post-implementation as there is so much you can do with this. For example, people were upset about the change from predator heat vision to the current, but it was changed to limit view as with night vision. We could bring back the old night vision through skills. Similarly, if players have trouble getting knocked down (when it gets re-implemented), it would be nice to have improved gyros in the skill tree affect that as well as screen shake. I would love to see those types of things added into the skill tree to add depth for non firepower/armor related bonuses.

[Delas Ting Usee]

Cathy, on 11 February 2017 - 07:22 AM, said:

this is a link to Fantastic Tuesday's you tube review



I think it pretty much encompasses everything that is good, bad and needs rework, and I agree with pretty much all his comments

I think everyone should look at it, and I hope the Dev's do also and take note

Thanks for bringing up this video. It raises ALOT of question I don't think PGI can answer now.
Please Please watch and LISTEN to what this fella has to say.
Personally, those of us with over a hundred mechs are screwed. Might be time for ANOTHER Hiatus.
Thank God Battletech is coming in March!

[Odanan]

50 50, on 12 February 2017 - 06:29 PM, said:

3 - Can't do that. If one mech has more skill points than another it's the same problem as having one mech with better quirks. People simply gravitate towards that mech and the other one never sees the light of day. Plus this is then a difficult item to monitor as it becomes a balancing point on it's own. What happens if it is decided that Mech A has too many points and it needs to be reduced? Try explaining why one mech has more skills than another as well as why it costs more to level up to a new player. I would think that a better option would be to provide a unique tree to the different mechs to give players that choice of taking advantage of that design.

But mechs are not equal. Number of hardpoints, location of hardpoints, hitboxes... all this makes some mechs much better than others. That's why we have some mechs with big boosting quirks and some with negative quirks.

The only problem I see with different number of skill points for different mechs is this: balance tweaking. If PGI decides to decrease the number skill points for some mech, she will have to refund what the players spent on it.

50 50, on 12 February 2017 - 06:29 PM, said:

4 - UI is good. At the moment while we convert the historical XP across it has that extra 'currency' but seeing as we can no longer earn historical XP that part of it will disappear in time. It's like this because of the shift from one tree and XP pool for all duplicates we might have of a single variant so we can now split that XP to specific individual mechs of that variant.

UI is good? I better stay quiet about that...

50 50, on 12 February 2017 - 06:29 PM, said:

5 - Going to disagree again. Now having skills for equipment that we previously didn't have is good. We were missing options for jump jets entirely, now we have them. If you don't want to spend points on them, that's fine, but don't get rid of them. The problem here is that people view certain skills as 'mandatory' regardless of what mech they have or how they might want to experience piloting it. This is the same view that has been imposed on everyone with which mechs we should use and which weapons we should have on those mechs and which modules we should equip etc etc.

But some skill tress are simply not worth the points spent on them. Even if you take all the jump jets skill tree, you will end up with marginal benefit. No one with a minimum of know-how will even spend points on them. The only thing some trees will do is to serve as newbie traps.

[Tethyss]

- I tried the PTS system the first day but after that I cannot find a match. It seems it's popularity has waned so more feedback may not be forthcoming.

- I like the fact I can customize each mech without having to 'build out' 2 more chassis

- The cost seems pretty high, c-bills especially, XP less so.

- The respec cost seems punishing

- (Sometimes) the pre-req skills seem like a waste in order to get to the skills I want. Perhaps flatten out the tree, or just give each weight class/chassis a skill point max to balance. Maybe like an MMO

- I appreciate the change overall as the old system was OK but the module purchase and swapping was not desirable. FYI I spent a lot of XP on the old system but made very few purchases of modules.

- P.S. sorry, forgot the boating problem. The new skill tree system absolutely encourages taking one weapon type, which overall is bad.

Edited by Tethyss, 12 February 2017 - 09:27 PM.

[EgoSlayer]

My 2 cents, probably some repeats:

This respec is way too costly on many fronts.

First - I have nearly 200 mechs in various states of Basic/Elite/Master. Even on mastered Mechs I am not refunded enough XP (or C-Bills) to bring my inventory back to it's current level. WTF? Let me buy skill points directly with Historic XP at 3x conversion rate to allow mechs to be returned to current level and Historic XP bought skill points should have zero C-Bill costs. It's not like I can spend this HXP on anything but the mech it came from. And this conversion interface is painful anyway, why not a slider or +/- instead of the dynamically changing as I type interface that isnt intuitive at all.

I get that the costs are to replace the module costs, fine, only put the costs on the items that are former modules. Which brings it to the next point:

Gating of skills behind undesirable or useless skills. Why do we have to take arm skills to unlock the upper torso, when many mechs don't have arm weapons or arm actuators that these skills require. Really? Have we learned nothing from 'Pinpoint' and 'Arm Reflex' in the current system? Fall damage gating the defensive skills? Hill climb and speed retention gating Cool run? And so on.

Don't gate the skills behind stuff we don't want. Reduce the total number of skill points if necessary, but if goal is supposed to 'Empower players with the ability to customize their 'Mechs performance characteristics according to their own desires and goals.' then this is an utter failure. My desire is not to take unnecessary or useless skills.


UI - This is a hot mess. I can buy any skills I want with XP or GXP. Even if I don't have either. And after going and spending 91 skill points, I have no way of telling which was bought with which to go fix any nodes that are a problem. Either don't let me buy them, or color code them with how they were purchased before committing so any errors can be quickly identified.


Lastly - Time. It's going to take many, many, dedicated hours to respec all the mechs I use. There should be quick build/template options for each tree so that with one or two clicks I can have the common most useful things selected and then fine tune if necessary. The level of apathy that amount of work just to get back into the game is going to generate is a serious disincentive to playing the game at all.

[LordLeto]

Respec costs are punitive to new players who will make mistakes and players who like to tinker and change their builds. This is the most important change in my book. A cost for respec is fine if reasonable, rebuying nodes is out of the question. Especially with new weapons coming this summer!

The new costs are prohibitive to our pokemech players. This is alienating to some of your most loyal and consistent customers. Those with mastered mechs need to be grandfathered into having mastered mechs in the new system, no cbill cost. I know the module refund was sposta address this but you failed to account for your cheap player base that shared a handful of modules over dozens of mechs.

In general the new costs are fine to me, exp and cbill, given the above two points are addressed. Doing away with the rule of three and costs of modules for 9.1 mil per mech is a bargain, especially for a new player.

The tree structure and some design choices should be reconsidered. Eliminate wasteful choices(arm reflex) that are mandatory, add front loaded bonuses and deminishing returns. Having 10%-15% CD f.ex. easily attained will help diversify load outs as they will make them more competitive with the boats who can specialize. As it is now it's all or nothing. This is no better then the current module system that also promotes boating.

[FunkyT]

So after looking a bit into the PTS I would like to share some of my thoughts about the system. Sorry in advance if I'm simply repeating what others said (too lazy to read through 6 pages and several threads).


The Skillsystem

Idea: Why not make skillpoints a one-time buy?

So skilling through your mech the first time would be just as it is now: you select the nodes you like and pay the price for each node / skillpoint. Maybe even keep a small cost for respeccing, guess 25% or 20% of the initial skillpoint price doesn't really hurt. But after respeccing, don't give them back the EXP they spent. Why would you even? They'll just spend just as much EXP on their next build, so there's no difference. Just give back the skillpoints.
In this case, buying the skillpoints would cost you just once. Reallocating them after a respecc would come with no additional cost. With the small respecc-fee you would still discourage people to just jump all over the skilltrees and would encourage them to think about their skill-choice. But you also wouldn't discourage people from trying new builds on their mechs. Changing their skills would be almost free then and wouldn't punish you for a change of mind.


Would someone think about the Childre- I mean the Trial Mechs?

Idea: Preallocate some skillpoints for the trial mechs.

Just imagine: A new player gets fresh out of the tutorial, in high spirit picks a trial mech and eagerly starts into his first match...Only to be greeted by enemie mechs with fully skilled defensive quirks. So the new player just tries his best, but he doesn't seem to deal any signifikant damage to the enemie and gets himself picked apart very quickly. And he has no idea why. Since he'll be stuck with trial mechs for maybe 25 matches or more, he'll likely lose interest. He'll just think "this game is totally pay2win" and leave.

To lessen this burden, why not preallocate some skillpoints to the trial mechs, maybe depending on the pilots tier.
Say a Tier 5 pilot picks a trial mech. Give this mech maybe 1/3 of the skillpoints, so roughly 30 points in this case. Have 10 of them put into the defensive quirks (structure and armor), distribute 15 over Upper- and Lower-Chassis and Heat Management (probably 5 for each) and put the last 5 points into weapon quirks.
That way the trial mech is still weaker than a fully skilled mech, but the power-gap is way smaller. Fighting wouldn't feel as unfair anymore.
For a Tier 4 pilot, maybe allocate 2/3 of the skillpoints, so roughly 60. Again, most of them probably into defensive quirks, because those will likely be picked first for every mech.
And from Tier 3 and up, just allocate all 91 points for the trial mechs.
Or maybe just share the points 50/50 for Tier 5 and Tier 4, don't know.

If the trial mechs stay unskilled, they will be completely useless from Tier 3 upwards, maybe even earlier. So instead of basically "punishing" new players for having to use trial mechs and instead of making trial mechs completely useless later in the game, try and go the extra mile and put together some basic builds for them.

Something simple, while on the subject:
Inform the player, that skilled (or "mastered") mechs are more powerful than unskilled ones. Make that a hint for the loading screens or something. Should help to not alienate new players, by informing them why there's a power-gap between them and other players.

[Flitzomat]

It was all perfectly preparred:
https://mwomercs.com...nd-skill-trees/

Seems they only listened to the XP cost parts though, which I totally not agree upon.

Edited by Flitzomat, 13 February 2017 - 03:01 AM.