36 replies to this topic

[Void Angel]

Please keep discussion productive; don't freak out - because freaking out makes irrational people defensive, and rational people may assume that you haven't thought things through. Strong wording is fine, but remember that while respect is about who you think other people are, courtesy is about who you choose to be. So, with that in mind:

---

This is my analysis of the proposed skill tree system for MWO, as per This Trailer. First, let me say that this is an awesome idea - a true MMO-style customization system that allows players to select perks according to their needs. Virtually every MMO I've ever played does this, from Everquest to League of Legends, and I'm pleased to see MWO following suit: it increases player options, streamlines balance feedback and telemetry, and moves the game closer to role-based warfare (depending on what skills are available.)

However.

While I love the system, the investment cost I cannot love. Bear with me here as I break down the problem I see with this system:

---

Assuming I want to master all the 'mechs: there are currently (unless I mis-keyed my calculator) 348 unique variants in the game - as per my Mech Trees skill list. This means that if Pirhana stopped making 'mechs right now, I'd need two hundred and sixty-one million experience to make that happen. I've been playing for a while, but a quick guesstimation of my current excess experience says I don't havemore than a couple million lying around in excess 'mech xp.

If I played the game like a job instead of a hobby, 40 hours a week, let's see... Assume I average 2,500 experience a match, including Premium Time (which we shall assume I am always running) and first-win bonuses. Assume 7 matches an hour, including time in the queue. That's 8 minutes and 34 seconds a match, and all this works out to 17,500 experience an hour; ~700,000 a month; and 8,400,000 a year. It would take me thirty-one. fracking. years. to skill out just the current 'mechs - not including those currently in the pipeline.

Now, I own just 67 unique variants - so if that's all I ever bought, given the above assumptions I'd be set up for six years of grinding as a new player - which is longer than I've been playing the game. That's longer than anyone has been playing the game.

That's insane. That's more hours than I'd have to spend to level up all the character classses on an MMORPG like WoW, and remember - we're assuming that I'm buying Premium Time and treating the game like a job. If I even treat it like a part-time job and do twenty hours a week, you double those time estimates. If I only play a couple of hours and skip a couple of days like a normal person, you double it again.

---

This system will dramatically slow my investment in the game - both of time and of money. This will happen because personally I want to play on a level field - and under this system I will not be able to do that against players in their favorite 'mechs for months after I buy a new chassis. In addition, balance feedback for me, the player, will be reduced, because I can't tell if an enemy pilot's build is better, or if he's just burned a bunch of GXP to kit out his 'mech with 12.5% cooldown reduction, 25% extra structure, and 15% extra armor (and that's just 15 points!) Playing new 'mechs becomes less attractive, since I know I'll be at a disadvantage for a much longer time, and buying a new 'mech pack or variant means that I have to start all over again: it's literally like being a max-level character in WoW, but having to re-level your talent trees every time you changed your spec - by fighting other players who do have all their talents.

The staggering investment requirements for this skill system runs the real risk of creating a haves/have-nots divide. It's the Stock Trials problem writ large: new players start out with 'mechs that are totally under par, and must compete against more experienced players in superior, customized 'mechs. We solved that problem in the current game buy designing newbie-friendly, fully upgraded Trial 'mechs, and by giving new players a Cadet Bonus - plus a few million for going through the Academy. There are no Trial Skill Trees, however - no Cadet Bonus for 'mech skills exists, and players must repeat the experience of being significantly out-classed every. time. they pick up a new chassis.

I love the new system. It looks great, but the crippling investment cost looks game-breaking.

Edited by Void Angel, 05 December 2016 - 10:07 PM.

[BLOOD WOLF]

While I do love constructive conversation Void, and its something I always try to do. There is just not enough information present to do anything but speculate. I know people are exited, but you might find more opinion that fact. Don't let that stop the discussion, Carry on

[Void Angel]

Well, they put out a trailer, and everything I put up is based on that hard data. And the math don't lie - so while it's possible that some unannounced change to experience gain will shift the time investment required toward sanity, until PGI weighs in with something different, there's no speculation in this post.

They're expecting it to be done "early 2017," so there's still enough time to change the system, but short enough time left for them to be pretty confident of how it's going to look. That makes this the perfect time to have this conversation.

Edited by Void Angel, 05 December 2016 - 10:11 PM.

[Jingseng]

The root problem is that there is no offline method to earning exp (well, that and the sheer amounts needed).

While in one sense, everyone will suffer the same problem (insufficient exp to distribute among all their mechs, possibly even just all their favorite mechs), as VA pointed out there is a bigger problem created in the uneven playing field. MWO is online pvp only - so that exp you need to level up your mechs can ONLY come from playing against other players.

This creates a potential imbalance that is not easily cured in the match maker system. As imperfect as PSR is, it will be even worse trying to put together 24 players of roughly equal PSR, tonnage, AND exp spent. It already defaults to skipping PSR from time to time - but the imbalance created from invested exp bonuses will be much worse in impact.

Which means if not changed, more players (as also noted, most particularly new players and vets on new mechs) will have unsatisfactory gaming experiences. Even giving exp bundles with mech packs or exp boost for X time after purchase wont even the field really (that is probably the solution they are eyeing... akin to bundled premium time and modules).

I wonder what they have in mind to combat this. In other MMOs, you'd just grind mobs... or else everyone starts each match at the ground floor (or else the opponent is always cpu). It could open the door to buying exp bundles for MC (given that gxp is awardable from the supply cache, and bundled exp with new mechs would be the simplest solution...).

Edit: always assuming they don't change required costs to skill (which wouldn't address the imbalance, just make the painful whipping period shorter)... If they detached skill trees from individual mechs and applied it to individual variants, that would be better... applied to the entire chassis group (all phoenix hawks, for example) would be unlikely...too restrictive on build variety

Edited by Jingseng, 05 December 2016 - 10:34 PM.

[Duke Nedo]

Somewhere between 1 and 2 k per Sp would be reasonable.

[NoiseCrypt]

Void Angel, on 05 December 2016 - 10:08 PM, said:

Well, they put out a trailer, and everything I put up is based on that hard data. And the math don't lie - so while it's possible that some unannounced change to experience gain will shift the time investment required toward sanity, until PGI weighs in with something different, there's no speculation in this post.

They're expecting it to be done "early 2017," so there's still enough time to change the system, but short enough time left for them to be pretty confident of how it's going to look. That makes this the perfect time to have this conversation.

I really like the new skill-tree concept.
And the demoed values has to be either a cruel joke to mess with the negative players or a just normal "pricing strategy".
The real values will look and feel so much better because everyone has had to wrap their heads around 750k xp to master a single mech.

[NoiseCrypt]

Duke Nedo, on 05 December 2016 - 10:44 PM, said:

Somewhere between 1 and 2 k per Sp would be reasonable.

I'm more in the 2-3k range. that would be one month of "grinding" for me to master a mech.
But it also really depends on how many skill points a specific mech will actually need to be "viable".

[Gaden Phoenix]

I really hope that they have a town meeting and this xp issue is brought up. For now we just have to wait. I think many many people have brought this up in the forums, including me. So, sadly it is now a waiting game.

[invernomuto]

Apart from excessive grind, what worries me is the possibility to take wrong decisions in skill unlocking (like lrm quirks on an Atlas). As far as I understood, your choices are limited so you have to spend xp wisely. I really hope that you could "undo" some skills unlocking if you discover that they not fit the mech...

[o0m9]

invernomuto, on 05 December 2016 - 11:36 PM, said:

Apart from excessive grind, what worries me is the possibility to take wrong decisions in skill unlocking (like lrm quirks on an Atlas). As far as I understood, your choices are limited so you have to spend xp wisely. I really hope that you could "undo" some skills unlocking if you discover that they not fit the mech...

It says right in the trailer that all mechs have full respeccability. Ostensibly you can just click a button and reset them but there might be a C-bill cost or something to try and force people to be careful with how they spend their points.

Edited by o0m9, 05 December 2016 - 11:44 PM.

[Duke Nedo]

NoiseCrypt, on 05 December 2016 - 11:00 PM, said:

Firstly: I am very very positive to the concept of a skill tree to replace quirks if done correctly.

That said:

I'm more in the 2-3k range. that would be one month of "grinding" for me to master a mech.
But it also really depends on how many skill points a specific mech will actually need to be "viable".

3k would be OK I guess, but still a bit harsh imo. I mean, I understand that they want progression to continue, give us something to grind for, but (this is a big butt) this is after all a PVP game, and as such there need to be a more or less level playing field where player skill matters more than how long you have been grinding your mechs.

I am not a comp player, but I would anyway not even consider playing handicapped (less than max skilled) mechs in a game I'd like to win. Therefore I'd sure as hell ace out all my best mechs first no matter the cost and what ever mechs are not top tier mechs and not covered by the Legacy XP would be left to rot in the mech bays or sold. Any drive to buy new mechs would be gone too since I'd with the original values would have 170 unskilled mechs waiting to be skilled up if I ever found the motivation to start that. I guess I am not alone reasoning like this...

With 2k I would be able to ace out 85 of my 200 mechs, I guess that would be OK. I could live with only skilling up fun-mechs to some 80% and leave the rest for grinding and spread the love a bit and maybe end up with only having some 50-80 corpses in my hangars. None of this will change the way I play when I play to win, but it would remove the possibility to fool around in fun-mechs, hence variety and fun-factor will go down a lot! It would have to go as low as 1k per SP to work out the best imo, Forcing grind-to-win in a PVP game based on mech sales is an incredibly bad business decision.... it will hurt the players, the game and PGI sales.

The only way the current pricing would work out would be if there were only 7 useful skills per mech and the other 68 skills were fluff skills that doesn't improve performance. Then you could "basic" up all your mechs and grind the fluff for your favorites. Not likely...

So, I have major concerns. The way this is presented I expect clan-is-variant-omni balance to be completely broken, and I expect to have more than half of my mechs obsoleted by a grind-wall.

Edited by Duke Nedo, 05 December 2016 - 11:51 PM.

[Jingseng]

1k-3k?

I mean, you get like 2k per victory in qp (maybe 500-800 when your team gets rolled)... That would be very fast to skill up. FP, double exp weekend/events (not to mention daily), etc would all make that run even faster. Somehow I doubt that is what they are looking to do - I'd be surprised at less than 3.5k (anchor turn, avg'd mastery level excl speed twerk), and more than 5k.

But really, at the 1k range, why even bother with experience points and such. Just let everyone unlock X skills from the go and let that be that =p

[NoiseCrypt]

Duke Nedo, on 05 December 2016 - 11:50 PM, said:

3k would be OK I guess, but still a bit harsh imo. I mean, I understand that they want progression to continue, give us something to grind for, but (this is a big butt) this is after all a PVP game, and as such there need to be a more or less level playing field where player skill matters more than how long you have been grinding your mechs.

I am not a comp player, but I would anyway not even consider playing handicapped (less than max skilled) mechs in a game I'd like to win. Therefore I'd sure as hell ace out all my best mechs first no matter the cost and what ever mechs are not top tier mechs and not covered by the Legacy XP would be left to rot in the mech bays or sold. Any drive to buy new mechs would be gone too since I'd with the original values would have 170 unskilled mechs waiting to be skilled up if I ever found the motivation to start that. I guess I am not alone reasoning like this...

With 2k I would be able to ace out 85 of my 200 mechs, I guess that would be OK. I could live with only skilling up fun-mechs to some 80% and leave the rest for grinding and spread the love a bit and maybe end up with only having some 50-80 corpses in my hangars. None of this will change the way I play when I play to win, but it would remove the possibility to fool around in fun-mechs, hence variety and fun-factor will go down a lot! It would have to go as low as 1k per SP to work out the best imo, Forcing grind-to-win in a PVP game based on mech sales is an incredibly bad business decision.... it will hurt the players, the game and PGI sales.

The only way the current pricing would work out would be if there were only 7 useful skills per mech and the other 68 skills were fluff skills that doesn't improve performance. Then you could "basic" up all your mechs and grind the fluff for your favorites. Not likely...

So, I have major concerns. The way this is presented I expect clan-is-variant-omni balance to be completely broken, and I expect to have more than half of my mechs obsoleted by a grind-wall.

I'm am also very concerned...
If done well this will be really really cool and just what the game needs.
If done bad it will probably kill the game for most players.

[Besh]

Imho, looking at the new SkillTree with thoughts of "mastering a 'Mech", or unlocking ALL the skills/quirks for a given Chassis/Variant, is looking at it from the wrong angle(s) . You are looking at it from the past .

What if "mastering" a 'Mech has NO meaning at all in the future of MW:O (i.e. does NOT exist ) ? . What if it is MEANT to be nigh on impossible to "finalize" even ONE Varaint in terms of Unlocks....let alone several, and what if it is intentional that in the future, every single Unlock is a meaningfull choice, and one you should not take lightly ?

What if SkillTree 1.0 is a Step in the direction of: there is no EndGame and final Goal, not even for a single 'Mech Variant . Keep playing. Keep making your 'Mechs better . For a pretty long time you can, on every single one .

Would that be so bad really ?

Edited by Besh, 06 December 2016 - 12:19 AM.

[NoiseCrypt]

Jingseng, on 06 December 2016 - 12:06 AM, said:

1k-3k?

I mean, you get like 2k per victory in qp (maybe 500-800 when your team gets rolled)... That would be very fast to skill up. FP, double exp weekend/events (not to mention daily), etc would all make that run even faster. Somehow I doubt that is what they are looking to do - I'd be surprised at less than 3.5k (anchor turn, avg'd mastery level excl speed twerk), and more than 5k.

But really, at the 1k range, why even bother with experience points and such. Just let everyone unlock X skills from the go and let that be that =p

I avg 1500xp pr match. With premium time. So one skill point for the first match a day and every second match after that actually makes a lot of sense. And you would still need something like 140 matches to master a mech.
That is a lot of game time for me... i have managed 1k matches since i started playing right after steam launch.

Sure they can raise the xp needed to get all 75 skill points.... but if the grind to make a mech valid turns out to be more than a month of dedicated gaming, then i would rather play EVE or WOW again.

Edited by NoiseCrypt, 06 December 2016 - 12:17 AM.

[Duke Nedo]

Besh, on 06 December 2016 - 12:15 AM, said:

Imho, looking at the new SkillTree with thoughts of "mastering a 'Mech", or unlocking ALL the skills/quirks for a given Chassis/Variant, is looking at it from the wrong angle(s) . You are looking at it from the past .

What if "mastering" a 'Mech has NO meaning at all in the future of MW:O (i.e. does NOT exist ) ? . What if it is MEANT to be nigh on impossible to "finalize" even ONE Varaint in terms of unlocks....let alone several, and what if it is intentional that in the future, every single Unlock is a meaningfull choice, and one you should not take lightly ?

What if SkillTree 1.0 is a Step in the direction of: there is no EndGame and final Goal, not even for a single 'Mech Variant . Keep playing. Keep making your 'Mechs better . For a pretty long time you can, on every single one .

Would that be so bad really ?

Point is, the best mechs that you play the most will be the ones that will be maxed out. That's now the ruler by which every other mech will be measured by.

Edited by Duke Nedo, 06 December 2016 - 12:21 AM.

[NoiseCrypt]

Besh, on 06 December 2016 - 12:15 AM, said:

Imho, looking at the new SkillTree with thoughts of "mastering a 'Mech", or unlocking ALL the skills/quirks for a given Chassis/Variant, is looking at it from the wrong angle(s) . You are looking at it from the past .

What if "mastering" a 'Mech has NO meaning at all in the future of MW:O (i.e. does NOT exist ) ? . What if it is MEANT to be nigh on impossible to "finalize" even ONE Varaint in terms of Unlocks....let alone several, and what if it is intentional that in the future, every single Unlock is a meaningfull choice, and one you should not take lightly ?

What if SkillTree 1.0 is a Step in the direction of: there is no EndGame and final Goal, not even for a single 'Mech Variant . Keep playing. Keep making your 'Mechs better . For a pretty long time you can, on every single one .

Would that be so bad really ?

I agree.
The discussion should be centered around "time to valid" and not "time to mastered".
They might not be the same thing.
If spending skill points in a single tree would give diminishing returns, then high sp mechs would still have an edge, but low sp mechs wouldn't be completely screwed.
This is one of the things that work really well in EVE IMO.

Edited by NoiseCrypt, 06 December 2016 - 12:24 AM.

[ACWILD]

I wonder....

If i have specced my mech for dakka, and, just for fun, i want to try a laser vomit build.....will i be able to do that or i'll have to spend an absurd quantity of c-bills/xp/gxp to do that?
This approach to the mech builds won't discourage the experimentation of new builds on old chassis?

I'm pretty worried.

[Vervuel]

Did they mention if you get exp credit for having already mastered a mech, or do you have to master them all over again?

[627]

as no-one linked the video yet:


And just a thought; they stripped all quirks and with the new skill tree, you can add "quirks" to any mechs. And that'll cost you.

The quirks were introduced to help mechs that performed poorly, it was a bonus for subpar mechs. Then this whole idea went nuts in it's current implementation where top of the line clan omnis start with quirks or people riot.

But think about it... you got some quirks for poor mechs, for free to level them against tier 1 mechs. Now, you not only have to invest quite a bit to get these quirks "back" - the tier 1 mechs can do the same and use the very same (or better fitting) quirks!

And the plan to all this is that it'll be balanced because you get less skill points in a tier 1 mech. as if you can't minmax a chassis with 60 skills instead of 75. Yay.

Edited by 627, 06 December 2016 - 12:48 AM.