64 replies to this topic

[Widowmaker1981]

Xzip, on 06 December 2016 - 08:34 AM, said:

It doesn't solve problem of undergunned mechs like Spider 5K/5V or Shadowcat or Commando or many others, sadly.

This system is absolutely going to help the Shadow Cat - it currently has pathetic quirks, and basically none for the good loadouts (2xERPPC / 2xLPL.. i guess 3xERLL if you can stand using them).. if the new system lets it have even a little bit of energy heat gen or ERPPC velocity it will help a geat deal.

Cicada-3C on the other hand.. that mech is boned. But its boned now anyway, even with godquirks.

[Tremendous Upside]

FireStoat, on 05 December 2016 - 05:41 PM, said:

As others have previously posted elsewhere, the time investment on mastering a mech is the key issue to this refund business. It's absolutely possible that the new skill trees will require more time to 'max out' a given specific mech in comparison to what is required with the current system.

It's absolutely UNACCEPTABLE if the refund conversion doesn't match the time taken by the player, hour for hour. It's also unacceptable for those players who spent MC on GExp conversion to not have that MC refunded or evenly addressed in some fashion.

If PGI drops the ball on either of those two points, this forum board will explode with older players with a great deal to lose taking up torches and pitchforks, laying waste to the company's reputation on far more than just this posting board.

This is my entire problem with revamping the skill tree. I couldn't care less what it does to balance. Everyone will have to play through it the same way. I "do" have a problem with the idea that I've gone for years having to level mechs up, throwing GXP conversions and real money into doing so - to have mechs that are essentially "done" and running at their max potential. Some of those mechs? I don't own anymore - but the money, time, etc put into them still had to count for me to master the ones I still have. If the conversion structure doesn't leave me with enough "credit" to max out all the mechs I currently have mastered in the end, that's an issue. And I suspect that'll be an issue for a lot of players and not just me. I hardly play anymore as it is. Dunk the time and money I've put into the game into a system where I'd have to grind more or throw more money into the game to get to where I've always been in the past and you can guarantee I'll just walk away entirely Not worth the bother...

[Jman5]

You laid out a lot of the issues very well. I hope PGI takes a look at this and considers how to address some of these.

Quote

2. Its Logistics Don't Work With the Current UX
Several players have pointed out that skill purchases are totally at odds with a smooth Mechlab experience.

I think the only option here is to make it possible for "respec" to be localized to individual trees instead of resetting everything every time you make a weapon change. That way at least you're only redoing a handful of skill points instead of all of them. I would put a little reset button over each section.

Quote

3. Every Solution to Balance Weak 'Mechs with Strong 'Mechs is Awkward
Take the two 100-ton 'Mechs bookending Mech_Con's stage backdrop.

This is going to be tough. I think the only realistic way to go about this initially is to at least partially cook the current quirks into the skill tree. So let's assume 2.5% weapon cooldown per level is what PGI goes with as default. At level 5 you're left with approximately what players were currently getting out of a level 5 cooldown module.

Now imagine you're trying to balance the Atlas D-DC for the new skill tree system using existing Quirks as a model. Ok the Atlas currently gets 10% energy and 5% ballistic/missile cooldown quirks. So you divide 5 or 10 by 5 and bake it into the 5 cooldown skill levels.

Now PGI has two choices here. They can go strictly by quirks so energy cooldown skills go up in increments of 4.5% (2 + 2.5) and ballistics go up by 3.5% (1 + 2.5). Alternatively they can just come up with one value, say 10% and so every cooldown skill goes up by 4.5%. The alternative approach would homogenize things more, but you wouldn't be favoring one weapon type over another.

Edited by Jman5, 06 December 2016 - 10:08 AM.

[C E Dwyer]

Have to say most of the concerns here are ones that mirror my own.

Even down to the same mechs I used as an example the Atlas and the Kodiak

Another example might be the Arctic Cheetah v Spider v Mist Lynx

Arctic Cheetah will be good from the word go.

Who is going to even bother grinding XP for a week on the Spider or Mist Lynx, only to find out that with all the skill hexagons lit up, they're still vastly inferior to the basic Arctic Cheetah.

The skills trees are a good, sound idea, you'll get no argument on that from me, It's a system that is pretty flexible and complete, and it will if introduced sensibly, be a big plus for the game.

Removing Quirks and using the skill tree to replace them, as promoted in the video, has a hint of coming disaster written all over it.

It just follows the standard pattern P.G.I follow time and time again.

They have a really great idea, then ruin it with the mechanics of the system.

Edited by Cathy, 06 December 2016 - 09:39 AM.

[DAYLEET]

Animus41, on 05 December 2016 - 05:13 PM, said:

"Give me a 1 to 1 conversion or give me my money back!"

We dont need a 1 to 1 conversion because its not like we will "master" all out hundreds of mech next time we log. We will do those we play that night and lets say we can only "master" 100 out of those 150 at patch release... then we have a long time to play before the xp runs. But it shouldnt run out because we gain it as i play and we have no other use for XP.

However, paying cbills for mastering is just wrong, we could have taken this opportunity to removed a cbill sink from the game. Theres no way i play 3 mech in a single night and make up for the millions in cbills required for the 3 mech. We only get the cbills back from the modules, we arent going to get the same legup as the XP conversion. You could say that its the same gimmick as XP but it isnt. Cbills is needed to buy mech, weapons, Engine, consumable and change armor and heatsink. XP is only needed for one purpose. Suddenly, we could be refunded 2 xp for every 1 xp we have and we still would be losing because the cbills required to master will never catch up as we play.

I personally think that the XP tree PER mech variant is the best idea PGI had since 2012. Being able to have 2 of the same variant but different "quirks" is awesome. Cant wait to see if the associated cbill grind nullify all my purchase and make me play another game for good.

Edited by DAYLEET, 06 December 2016 - 10:35 AM.

[AnimosityMonk]

@DAYLEET It needs to be a 1 to 1 conversion because there is no way for PGI to justify anyone having to regrind mechs that have been mastered or mastered then sold for whatever reason. PGI can not arbitrarily say all the work we players have put into this game for years is now only enough to unlock 10% of the skills in the new system. Hence,

"Give me a 1 to 1 conversion or give me my money back!"

[Xzip]

Animus41, on 06 December 2016 - 11:09 AM, said:

@DAYLEET It needs to be a 1 to 1 conversion because there is no way for PGI to justify anyone having to regrind mechs that have been mastered or mastered then sold for whatever reason. PGI can not arbitrarily say all the work we players have put into this game for years is now only enough to unlock 10% of the skills in the new system. Hence,

"Give me a 1 to 1 conversion or give me my money back!"

Can you STOP posting this bit about 1 to 1 conversion in EACH topic about skill tree? We discuss here what caveats new system may potentialy have instead of what grind it will be....

Widowmaker1981, on 06 December 2016 - 09:16 AM, said:

This system is absolutely going to help the Shadow Cat - it currently has pathetic quirks, and basically none for the good loadouts (2xERPPC / 2xLPL.. i guess 3xERLL if you can stand using them).. if the new system lets it have even a little bit of energy heat gen or ERPPC velocity it will help a geat deal.

Cicada-3C on the other hand.. that mech is boned. But its boned now anyway, even with godquirks.

True - but here I see bit problem still. Shad Cat will have same max potential quirks like Stormcrow and Nova. I know that Shad Cat power is mobilty, agiltiy, ECM, hit and run tactics... however it feels silly that all three of them have max potential in quirks even if boating potential is completly different.

Personally I love to run 2ML + ERPPC + ECM Shad Cat. Funny little mech~

[Nickredace]

East Indy, on 05 December 2016 - 04:23 PM, said:

Let me preface by thanking PGI for being open about game design plans — especially works in progress. It's not easy to subject an idea to criticism. It's extremely helpful, all the same. I'd like to calmly and rationally explain why embedding quirks in skill trees is not a positive change for this game, and is so fundamentally flawed that no amount of testing can fix it. So, again: hey, it's an idea, interesting concept. But it's not going to work. 1. It's a Solution in Search of a Problem Why? Because MWO is not a game with a shortage of potential variety. If a player wants a certain style with certain weapons in a certain weight class moving at a certain speed, there are dozens of 'Mechs available, each with a half-dozen variants. And then each one can be pulled into the 'Mechlab. Open-ended customization would make sense if MWO were a limited-roster title like Wipeout or F-Zero. But it's not. We have scores of choices to suit a playstyle, and as the Roughneck's unveiling suggests, if there's a burning desire for a loadout regardless of lore, PGI can oblige.

Your Concern is... there already a good amount of variety so why have more? Open ended customization leads to a varied play-styles with varied mechs. For example faster Direwolves, More speed and heat management with Artic cheetas, or maybe lrm, and ac cool down and better twist speeds with the King Crab.

East Indy, on 05 December 2016 - 04:23 PM, said:

2. Its Logistics Don't Work With the Current UX Several players have pointed out that skill purchases are totally at odds with a smooth Mechlab experience. Let's say you've made a build with four LB-X autocannons. You'll show everyone! You pop your range and cooldown modules in. You're about to push "Quickplay" when you come to your senses, rip out the LB-Xs, and choose a sensible loadout. You hit tonnage, swap modules, hit solo queue and immediately earn Ace of Spades. Now, let's try that with quirked skill trees. You rip out the LB-Xs and choose the sensible loadout. But you'd already invested in LB-X quirks. So, you go to Skills. You deactivate all the LB-X points. (You can, can't you?) Then you need to reactivate all the points for your new loadout. Well, actually — had you purchased them for this specific 'Mech? If you didn't, time for some button clicks, and maybe a microtransaction. Then you've got to work your way down trees, hopscotching to get the right combination of quirks. And you may need to do it three or four times for each system. What if it's not solo queue you're playing. What if four unit-mates are waiting? Heaven forbid you change your mind again and decide to go True Blue on the LBs.

Again your Concern is, its deeper customization. Also taking more time to stat a mech isn't really a bad thing now that their is more deeper customization per mech. As for the deactivate all points, (yes its a thing it even showed it in the trailer) I completely doubt re-specing is going to cost a micro transaction. For some reason your complaining about having more options, simply because it takes more time to get the mech you want. Its not a downside.

East Indy, on 05 December 2016 - 04:23 PM, said:

3. Every Solution to Balance Weak 'Mechs with Strong 'Mechs is Awkward Take the two 100-ton 'Mechs bookending Mech_Con's stage backdrop. The Atlas needs its brains quirked out to be half-effective. The Kodiak is so powerful it's been nerfed once, and indirectly nerfed twice. In order for MWO not to fold in on itself the day quirked skill trees are released, the Atlas needs some way of at least keeping pace with the Kodiak. Oh, there are options. But all of them stink. a. Give the Weaker 'Mech Points to Start With. This is by far the most straightforward solution, and I imagine the first PGI will try in an effort to save the design. Whatever 'Mechs receive in skill points as a baseline, the Atlas receives more. But how many more does it need? The Atlas isn't slightly more quirked than the Kodiak. It's exponentially more quirked than the Kodiak. Does this mean it has four times the skill points? Five? Ten? Is that even enough to match today's (insufficient) quirks? Is a new player ready to invest scores of points? Awkward. b. Increase the Bonus of Weaker 'Mechs' Skill. Instead of having a Scrooge moneybin of skill points, a skill point would purchase more than 2.5%, or whatever the bonus, so that the same number of skill points could allow the Atlas to reach its current level of quirks. One problem: many 'Mechs are better off than the Atlas but not as good as the Kodiak. A few 'Mechs are worse. This mean every single 'Mech will have a different bonus factor. Hey, our balance solution created a new balance problem! Awkward. c. Give Weak 'Mechs Deeper Trees. The least baked of the solutions, it nevertheless needs to be identified. If an Atlas needs more quirk bonuses, let the player keep buying them! Except this saddles the player with a kind of 30-year mortgage where they pay their way out of performance debt and finally reach parity long, long after the Kodiak has been getting 1K games like it's breathing air. And, like increased bonuses, every 'Mech will have its own depth. What happens when a tree gets truncated — refund? Awkward. -- Universalizing quirks is not going to work. Save your dev time and PTS electricity. My own humble suggestion is to simply make quirks more uniform — less eclectic, looking at you AMS RoF — so that QUIRKS = BETTER all the time, which in turn makes them easier to understand. It doesn't matter if quirks are simple and similar, because MWO provides so many more dials for players to use for customization. Just trying to help!

I will admit this one makes sense Mech balancing is a problem. But with this system you'd be able to fix alot of mech problems that exist (for your own personal taste).

[Bradigus]

The issue of mech packs providing essentially worthless content could easily be solved if the new skill tree system is outfitted for it in such a way: every mech of that chassis you own and play with can use it's experience points to unlock the skill tree steps of that chassis. After unlocking that skill, you could then choose whether or not to allocate a skill point into the slots you previously unlocked.

For example; I play the newly released Marauder IIC, but only one variant catches my eye. If I want to unlock the skill tree quicker, I can play for the first win of the day bonus on multiple variants and use the extra experience to fully unlock the skill trees for it quicker.

This way, you still have incentive to own multiple mechs of a chassis, you do not disrupt the current mech pack selling system, and you make it slightly easier for collectors to master the chassis.

TL;DR:
Skill tree unlocking is chassis wide
X mechs of same chassis = more experience gain with first win of the day bonus for faster mastery
Each individual can still pick and choose the final composition of their skill tree separately
Grind isn't as severe

[WrathOfDeadguy]

I'm not overly concerned with my badmechs getting worse. Most of them, I bought already knowing they were bad because screw it, they're fun... or because I get a sick sense of satisfaction out of the challenge of being an already flatly average pilot driving a badmech. Whatever. I will still drop in my PXHs, VNDs, and (insert every light that isn't the current meta darling). IF.

And this is important: IF I don't have to grind out 'Mechs I've already bought skills for. Not just the mastered ones- the basic-ed and elited ones too. And the ones that I only got a few skills for but never completed one tier or another. All of them.

Thing is, I'm very limited on playtime these days. Down to a handful of hours per week, if that. I do not have the time to sink into re-grinding. I won't say that I'll never play the game again, because I know myself. I get nostalgic. I'll miss playing stompy death robbits with a particular chassis and jump back in for giggles somewhere down the line. I won't even ask for a refund, because I accept that I spent the money and it's gone. But I won't spend another penny on it, and I will give other games the overwhelming balance of my available gaming time- which probably means I'd just uninstall, because by the time I played next I'd be patching the full installation anyway.

New tree with no loss of time invested? Fine by me. There are more 'Mechs I could see myself putting money on; the same limited play time that makes me retch at the thought of re-grinding XP is also what makes it worthwhile to buy virtual robbits for real bucks to shortcut the other grind. Money is not the issue. Time is the issue. That is what makes a more expensive skill tree unacceptable. If the time I've already spent grinding doesn't translate to exactly as much progress along the new skill tree as it did with the old skill tree, the game suddenly goes from being a fun little diversion- which I throw money at every now and then- to something that is not worth my time, because time spent in the game would have less value.

I'll reserve judgment until the changes go live. Until then, I'm wary.

[Weeny Machine]

Viktor Drake, on 05 December 2016 - 04:37 PM, said:

You have hit most of the issues on the money but I will add one.

The Grind.

Right now it is set at 10,000 XP per skill point or 750,000 XP to Master a mech. Now take your two examples, the Atlas and the Kodiak. As you point out, the Atlas needs a truly massive amount of quirks to make it competitive while the Kodiak really needs none. Even if they do some sort of asymmetrical bonus system where the Altas gains more power from skill, quicker, right out of the gate the Atlas is going to be at a massive disadvantage. The question then is how long before that Atlas will gain enough SP's to offset its initial disadvantage? 10 SP,? 30 SP? 50 SP? 75 SP? Even 10 SP will require about 66 matches if the Altas can earn 1500 XP a match. 50 SP will require over 300. That is a hell of alot of time playing a gimped mech.

...or light mechs which will have most likely access to more skill points but in turn needs an eternity to get those.

Edited by Bush Hopper, 06 December 2016 - 12:20 PM.

[Kael Posavatz]

1) Solution in search of a problem

Existing Skill tree is an artifact of Closed Beta and has been in need of replacement for a while. Modules were a generic form of customization. And Quirks have been a point of contention for a while now

2) logistics

Since it looks like the weapon quirks are going to be generic-ballistic/energy/missile rather than specific-weapon oriented this is less of a problem. Where it could be a problem is in regards to OmniMechs where choosing a particular weapon-type node will incentivize players to only make customizations within the same set of pods.

3) Balance issues

We only had a look at a single mech's tree. In all likelihood underperforming mechs will have nodes with stronger buffs, and overperformers will have weaker nodes. In the same way IS mechs will probably, at least initially, have much stronger nodes than Clan equivalents, with nodes becoming more balanced as both mechs accrue experience.



Frankly I'm not as concerned with some regarding XP costs. Russ was careful to stay away from the matter in his speech, but some of the devs discussed it briefly at the Q&A. Basically, all the numbers in that video are placeholders. I fully expect mastering one mech will take more than the slightly less than 60-k it currently takes, but I rather doubt it will take twelve times as much experience which is what the video had.

Where I am concerned is there has been no word if respecialization will be monetized or not. In games with similar mechanics it is. The issue with MWO is that you have one type of mech, OmniMechs, where versatility is supposed to be a key feature. Monetizing respec then becomes monetizing the feature and that is very nearly the definition of 'play to win'.

And I would point out that under the current timeframe, seeing an Inner Sphere Omni (Raptor, 3052) is possible. Russ has stated repeatedly that he would be 'very surprised' if we didn't see the Mad Kat Mk II in 2017. This would necessitate a time-skip to mid-3060s which would put the Black Hawk KU, Firestarter and Blackjack Omnis, the Owens, and something like six or seven other OmniMechs on the table for the Inner Sphere.

[WildeKarde]

I think one concern I have is if the skill tree's are all the same for every mech then it promotes the build focussed around less or a single weapon type. Meaning that Omnimechs are likely to be built to boat as they will get the best bonuses compared to using differing weapon types.

How do you promote any sort of balance when the skill tree promotes boating and added to that clan mechs with their option to build for boating potentially

[Khobai]

I think each mech should have "tag skills". And if you unlock a tag skill for a mech you get additional bonuses from that skill. That encourages you to build your skill trees towards specific skills which would differ from mech to mech.

For example, an atlas might get level 5 armor as a tag skill. And if you take level 5 armor youd get the normal skill bonus for taking level 5 armor plus an additional bonus on top of that because its one of the atlas' tag skills. That would allow the atlas to still have more armor than other assaults, which it rightfully should.

And obviously less powerful mechs would get more tag skills than more powerful mechs. Its a simple way to differentiate mechs and buff less powerful mechs that works completely within the new skill system.

Edited by Khobai, 06 December 2016 - 12:55 PM.

[Y E O N N E]

Khobai, on 06 December 2016 - 12:47 PM, said:

I think each mech should have "tag skills". And if you unlock a tag skill for a mech you get additional bonuses from that skill. That encourages you to build your skill trees towards specific skills which would differ from mech to mech.

For example, an atlas might get level 5 armor as a tag skill. And if you take level 5 armor youd get the normal skill bonus for taking level 5 armor plus an additional bonus on top of that because its one of the atlas' tag skills. That would allow the atlas to still have more armor than other assaults, which it rightfully should.

And obviously less powerful mechs would get more tag skills than more powerful mechs. Its a simple way to differentiate mechs and buff less powerful mechs that works completely within the new skill system.

I really like this idea. It builds the quirks that some 'Mechs need into the new system in a fairly organic way.

[Quicksilver Aberration]

Yeonne Greene, on 06 December 2016 - 12:58 PM, said:

I really like this idea. It builds the quirks that some 'Mechs need into the new system in a fairly organic way.

The only problem is it still means at the very beginning that mech is at a larger disadvantage until you start to pile on those tag skills unless we are just assuming this is meant more for flavor between mechs and not balance between mechs.

[Khobai]

Quote

The only problem is it still means at the very beginning that mech is at a larger disadvantage until you start to pile on those tag skills unless we are just assuming this is meant more for flavor between mechs and not balance between mechs.

the tag skills dont necessarily all need to be deep in the skill trees though. some of them can be shallow in the skill trees too. for example the atlas could get armor 1 as a tag skill and armor 5 as a tag skill.

ideally tag skills should be distributed in such a way that it only takes about half your skill points to get all the tag skills and the other half you can spend wherever you want. that still allows for plenty of customization options.

Edited by Khobai, 06 December 2016 - 01:26 PM.

[Quicksilver Aberration]

Khobai, on 06 December 2016 - 01:23 PM, said:

the tag skills dont necessarily all need to be deep in the skill trees though. some of them can be shallow in the skill trees too. for example the atlas could get armor 1 as a tag skill and armor 5 as a tag skill.

I was already assuming that, but there will still be a divide even then since not all the tag skills will be shallow in the tree.

Edited by Quicksilver Kalasa, 06 December 2016 - 01:28 PM.

[Khobai]

I dont see it as a big issue because regardless of the mech youre going to be at a disadvantage anyway if you havent spent all 75 skill points.

yeah some mechs might be at a bigger disadvantage than others while leveling up, but what matters most IMO is that the mechs are balanced at having 75 skill points.

[Y E O N N E]

Quicksilver Kalasa, on 06 December 2016 - 01:02 PM, said:

The only problem is it still means at the very beginning that mech is at a larger disadvantage until you start to pile on those tag skills unless we are just assuming this is meant more for flavor between mechs and not balance between mechs.

True. They could also just start the 'Mech tree off with the minimum already populated.