88 replies to this topic

[Sjorpha]

MischiefSC, on 06 March 2017 - 02:27 AM, said:

I encourage anyone who wants to to also test just what a brutal nerf to IS XLs the mobility hit is. I don't want to just go off anecdotal experience but I found it absurdly easy to focus a ST. Torso twist is now slow enough that even someone trying to twist can not avoid a full laser burn to a single ST. I was consistently able to get 2 CLPLs and 3 CERMLs almost completely on a side torso of pretty much any/every heavy or assault mech. At least 80%. The inability to 'jink' in light mechs made ST focusing them out terribly easy as well. I found it easier to kill a Locust via ST than trying to leg it.

Which is to say that even with full 'survival' tree perks the loss of mobility has left IS XLs in an untenable situation, at least in what I've seen.

The thing is that it might be a good thing that the extreme imbalance between IS and Clan XL is more exposed and harder to deny in this system, this might force the issue a bit and we'll get to actually balancing the engines equally quicker.

Point being that the XL imbalance is a huge problem in any system, and a system that obfuscates this fact will only serve as an excuse to not fix it.

Let's go with this and make the problem plain to see, and then push harder for PGI to finally fix the faction tech imbalance.

[Greyhart]

Reno Blade, on 05 March 2017 - 11:15 AM, said:

I've added a radial skill tree to my original post here:
https://mwomercs.com...ranch-like-this

with the center starting point, you can use more than 3 roots.
I've sorted the nodes, so you can always get all nodes of a certain base-type and weapon-types are reachable by passing 1 or 2 nodes of 2-3 different base-types.

e.g. Energy can be reached after 1 node of cooldown/heat or 3xrange (please ignore the missing link lines) and Missile can be reached by 1 cooldown or 2 heat/velocity nodes.
only ballistics are left with 1 cooldown/range node, without direct connection to velocity.

Option 4
Radial starting point with skill types grouped next to each other (eg. heat between missiles and energy)

I think if we are going along the lines of hexagons then the start should be in the centre and that you are then running in 6 different directions. you could then use the "filler" skills to traverse between branches that are close too each other.

So may be not exactly as above but it makes better use of the nodes

[Horseman]

Alistair Winter, on 05 March 2017 - 04:08 AM, said:

Players want a more simple system.

Yes, even if individual SPs are made more expensive.

Alistair Winter, on 05 March 2017 - 04:08 AM, said:

Players don't like being forced to take random, unrelated skills.

Agreed. They should be "opportunity upgrades" secondary to the more desirable skills.

Alistair Winter, on 05 March 2017 - 04:08 AM, said:

A linear skill tree would be better.

I would not go fully linear, but rather apply additional restrictions to the current branching design.

Back during PTS1, I've thrown this together as a proof of concept: http://imgur.com/a/jLa3P
Now, take it with two assumptions:
1. You can only choose only one child node of the last node you bought
2. The node values are buffed to be more substantial.

[rook]

I'm torn between the number of nodes/skill points we have. There are a few ideas that sound good to me:

• A path choice system, were you just pick a major path, and get skills based on that path (ie - sniper path vs. brawler path). The "nodes" are limited to one path or another (ie - the brawler would have torso speed and armor hardening while the sniper path would have AC/PPC velocity and sensor range)
  
• Removal of the skills that are overall bonuses to any type of build, such that you don't really have a choice in picking them and/or are not really customizing your build
  
• Having a large number of skills and points (the current PTS) because a major aspect of the game is the mechlab and lots of effort should go into your build with lots of nobs and tiny tweaks

If you have lots of skill points, it seems like a stragegy to force lesser useful nodes to make you "use up points". I'm ok with this, if lots of points/lots of nodes is the system of choice.


My Cyclops-S seemed very agile (a bit slower) with the skills I picked out, didn't try it without skills. I heard lights and meds are less responsive (without having tried them myself). Overall, I would like to see a slight nerf of agility across the board for mechs. I want to have more of a feel of a giant death machine, instead of a agile humanoid (as compared to other FPS). Though, whatever level of agility we have, lights and mediums need to have enough agility that they're sacerfices in firepower and armor is approperite.


While the complicated hex system is pretty in some ways, I want more of an industrial feel to it. This is a derivative of battletech: makeshift parts, half-arse solutions, mcguyvered components...

[MechaBattler]

Are people really ready to make meaningful choices? I doubt they'll increase the percentages if they do make the trees smaller. They'll probably just consolidate what we already have. So you'll see a fuller firepower bonus. But you're agility will suffer more or your heat performance. Are people ready for that?

[Greyhart]

MechaBattler, on 06 March 2017 - 09:08 AM, said:

Are people really ready to make meaningful choices? I doubt they'll increase the percentages if they do make the trees smaller. They'll probably just consolidate what we already have. So you'll see a fuller firepower bonus. But you're agility will suffer more or your heat performance. Are people ready for that?

I am, makes things interesting.

But the whining will be epic. People only ever want power increase not a cross the board power reduction.

reduce damage and everyone will complain. increased armour and structure and they will cheer. end result is the same of course just playing with the emotions of loss and gain.

[AngrySpartan]

Sjorpha, on 06 March 2017 - 05:48 AM, said:

The thing is that it might be a good thing that the extreme imbalance between IS and Clan XL is more exposed and harder to deny in this system, this might force the issue a bit and we'll get to actually balancing the engines equally quicker.

Point being that the XL imbalance is a huge problem in any system, and a system that obfuscates this fact will only serve as an excuse to not fix it.

Let's go with this and make the problem plain to see, and then push harder for PGI to finally fix the faction tech imbalance.

It's 99,9% probability there will be light engines (LFEs) when the new tech eventually arrive. That's the only fix that is required unless PGI will start counting engine crits properly.

Besides, skill tree is not supposed to fix that (as of now), though there might be some interesting options if some skill nodes will be tied to technology. E.g. "cXLFE/IS LFE durability" - x% less heat/speed penalty if losing a side torso, or "Standart structure toughness" - x% less damage transfer if using Standard structure, etc.

[Reno Blade]

Thinking about radial and overlapping attributes in different builds or roles...

if we would have some kind of radial distribution of points, we could have some kind of "role preset" which automatically assigns skills and similifies the whole skill points amount to 6 attributes.
Then if you enable "advanced" you can change the skills manually before saving.

similar to Never winter nights 1/2 or other RPGs provide presets for classes.

this would provide a few benefits:
- similification (cause you can select roles and get the full set of benefits needed for that role)
- customizability for the customizers (the ones who want to spend more time in the mechlab than the game to fine tune their mechs)
- role identification (could be used for ingame icons depending on point distribution)

maybe the hexagon tree will then look like this if using the "simple" view:

[Shu Horus]

Proposal for streamlining the Skill trees:

I can understand to use more Skill Points instead of fewer, as it means smaller steps to the next advancement. So you can get some reward without having to wait for the next big XP cap. I'm OK with that.

What I do not like are these big spreadsheets for skill-selection without any sense. Why do we Need x-time the same skill skattered over the whole sheet ?

How do Players actually select skills ?
They aim for a bonus they like and will find out the "cheapest" route getting most of it.
So what options do we have: Either harden this attempt by placing useless skills in between, or just give them the freedom to actually chose the options they like.

Going road one you will just annoy most people because there is absolutely no reason fo doing so. All people have the same access to the same skill trees, so there is no disadvantage for anyone if you make it plain and simple. So the only People who will not be annoyed by this are the ones who do not face unwanted Skills for their chosen path.

If you go road two and just make the bonuses available straight Forward, there is no longer a need for such wide spread trees. There is only one hex needed for i.e. Laser Duration but you do have the option to sink multiple Points into it until you reach the maximum bonus allowed.
You can still "block" out some more valuable Skills behind others. i.e. you can only Access the extra UAV after you already taken at least one point either in UAV-Range or UAV-Duration. (if that's neccessary for balancing).
Easy thing and the skill tree would be easily understandable for everyone.


This Option would also open another opportunity to actually make this system more diversive without making it more complex:

The implementation of negative Skills.

For some skills it might be an option to actually make reducements available, to free some extra skill Points (definately not for all skills, as it should not be possible to take negative quirks that actually do not have any impact on your mech).
This would have to be limited of course (i.e.: you can only pick up to 8 negative quirks to free 8 extra points) and the negative quirks must be of impact to actually matter (i.e. it will block the corresponding positve quirk completely and all the skills that might be locked behind it and the negative effect you are taking will be far more than what you would get if you take the positive variant, like factor 3 or more).
Maybe there also should be a limitation that gained Points may only be available for the same skill tree the negative quirks came from.

Quick example:
Jump Capabilities
The heat shielding quirk gives you a 6% Heat reduction per Point spent.
The player would like to free some extra Skill points to spend in this tree and decides that he is willing to take more heat, if he could get another 3 Points to max. out the Lift Speed of the mech.
So he takes 3 negative quirks for the Heat shielding (but as it is a negative quirk the factor is x3 for the effect) which will bring a negative effect of 54% extra heat when using the jump jets, but enables him to buy the missing 3 Points for Lift Speed, he wanted so bad.

Edited by Shu Horus, 06 March 2017 - 11:29 AM.

[Reno Blade]

Ok, now I've went ahead and created my own radial attribute based role-tree.

Nodes in tier 2, 3 and 4 (outer region) cost more, but also provide more bonus.
Unlocking a tier will provide certain extra bonus values.

Skirmisher preset uses
5 red (operation) nodes unlocking tier3 operation bonus
3 white (sensor) nodes unlocking tier2 sensor bonus
10 purple (firepower) nodes unlocking tier4 weapon bonus
12 blue (mobility) nodes unlocking tier4 mobility bonus




Command preset uses
12 red (operation) nodes unlocking tier4 operation bonus
1 white (sensor) nodes
4 purple (firepower) nodes unlocking tier3 weapon bonus
13 yellow (defense) nodes unlocking tier4 defense/survival bonus
10 green (command) nodes unlocking tier4 command bonus (incl. aux and other module improvements)

[MischiefSC]

Sjorpha, on 06 March 2017 - 05:48 AM, said:

The thing is that it might be a good thing that the extreme imbalance between IS and Clan XL is more exposed and harder to deny in this system, this might force the issue a bit and we'll get to actually balancing the engines equally quicker.

Point being that the XL imbalance is a huge problem in any system, and a system that obfuscates this fact will only serve as an excuse to not fix it.

Let's go with this and make the problem plain to see, and then push harder for PGI to finally fix the faction tech imbalance.

Except PGI left Clan/IS tech broken for years. Like pants on head, dancing in clown shoes stupid obvious broken. There is no such thing as 'obvious enough' for PGI. You could light their office on fire while they're in it, they would be in a room full of raging flames and one might say 'does it seem hot in here?' the response would be 'Maybe. Ask me again in November.'

Supporting PGI releasing a broken system isn't going to make them fix the other things they've been ignoring any faster. I'm sure they look at the analytics and say something stupid like 'Clans only win more because the higher skilled players play them'. Except the higher skilled players gravitate to Clans because the tech is better.

I work in analytics. I watch office manager types make bad conclusions out of otherwise clear data every single day. If PGI even wants to balance tech (maybe they don't - maybe they want balance broken for some even more stupid reason) they're likely not separating data to comparable skill performance in comparable fields. As in, IS brawler vs Clan brawler with comparably skilled players, that would give you useful, directly actionable and supportable data.

Anyway. I could rage all day on bad decisions in MWOs design process. I would not support encouraging PGI to make another terrible mistake in the hopes it would teach them something because history indicates that they absolutely would not learn from it.

[Quicksilver Aberration]

I voted for no skill tree, but since that was multiple choice I also voted for linear because no skill tree just isn't realistic. If I could have my dream skill tree it wouldn't have skills, but would have unlocks (like camos, colors, etc) and would be chassis based.

Edited by Quicksilver Kalasa, 06 March 2017 - 01:03 PM.

[FireStoat]

If I was allowed only a single suggestion to make to PGI's devs, it's that 'filler / useless' tree nodes feel extremely unrewarding to spend Cbills on to unlock. The filler nodes should either be adjusted in placement or simply formed into a big group and given to the player for free after spending certain thresholds of points into a mech, so that every point actually spent and paid with Cbills feels relevant and worth the price.

This game doesn't exist in a vacuum. Newer players or current players that play a diverse set of games are going to compare the unlocking process with other games and either have a good feeling or a bad feeling in putting time into this one. At the moment, it's not comparing favorably with other games and the process they use.

[JC Daxion]

while i agree with, having less nodes, just to make it a cleaner system.. I think the way it's done is ok.


as far as not taking nodes you don't need/want.. It's completely a balance issue, I'm sorry min/maxers should NOT get everything they want, it only throws the system more out of balance.


The only other changes i would make are make it so some more earlier nodes reflect how the basic skill tree works now on line. meaning early nodes, getting less heat gen, better movement in twisting/turning, accell/decelll.

Over all i think It's going in a good direction.. Clean up the bugs and get it out there for a Mass live test. And give it a good month, with perhaps some events, aka mech class, followed by some individual leader boards so people will be piloting all the mechs on live, to get better numbers.

Also give people time to adjust.. this is a game changer, people are not going to adapt and like it over night.

[Alistair Winter]

I would actually be very surprised if PGI didn't listen to the feedback and try to reduce the complexity of the skill tree while reducing the number of nodes. They may not go as far as some people want (e.g. 10-20 skills per mech), because that means each node will be more expensive and the cost of mastering mechs will be more blatant. Don't forget, mastering mechs is supposed to be a huge C-bill sink, in the same way that modules were. This is the MWO "endgame".

On the other hand, I would be very, very surprised if PGI fundamentally changed the structure of the skill system. I think they are 100 times more likely to just reject skills outright and go back to modules, rather than doing a radical change like a non-linear system, or taking inspiration from Witcher 3 or Mass Effect 3. I think what we'll get is going to be a variation of what they've already made, if we get anything at all. Because that's the way PGI has worked this far. They either make small adjustments or they rip it out of the game all together.

My biggest concern is that it will be watered down to have very little impact on how the game is played at all. Then it just becomes a tax to make your mech as powerful as enemy mechs, rather than actually letting you do new stuff, which is the most fun part of skill trees. Having new skills. If every mastered mech is just 10% faster, 10% tougher and 10% deadlier... what's the bloody point?

[JC Daxion]

I just think lineal tree is just a bad idea for exact reasons of your comparison. It's not an RPG, but even in RPG's we have skills we don't want/use ext. I can name tons of games that had a skill i needed to get, just to unlock the one i wanted. this is just not new and always about wasting points in the name of balance.


I do think the amount of skill nodes can be cleaned up a bit, but there is a fine balance to, making players want to use all there points, and not be, well if i just unlock X, Y and Z i don't need anything else because it won't make a diference.

My solution to this is make nodes in a couple of sizes. Your very early unlocks on the tree should give a noticeable difference. then the fluff, lesser, filler, must use to unlock nodes.. then finally your end game nodes.

I also wouldn't mind it if they priced um accordingly. early nodes, use little XP, and little C-bills.. Say 25k/1k XP. Maybe typically 20-30 nodes would fall under this catagory. (spend just under a mill on this level and 20k XP

then make the filler nodes cost 50k/2k these nodes would be a majority of your nodes. and cost around 2m, and 30-40k XP

Then the final nodes, or END GAME nodes, which have a good boost, and cost with them. be it a big heat or recharge, maybe speed tweak, some module type ones.. that sorta thing. Make these cost like 250k each and maybe 3k XP. say a typical mech would have maybe 8-10 of these

Cost is around 4m or so depnding on how many endgame, and early level nodes you take, and XP is around what it takes to master now.. I'd drop the total number of nodes, to maybe 60? a bit more powerful, and less choices.


Basically it would give people a nice bump early, so new players are not stuck in poorly performing mechs.. the middle area, is more slanted to going in a given direction towards your end game nodes. But it doesn't make all end game nodes end game. you could choose to say get into radar dep, or speed tweak or maybe a single end game weapon heat reduction node with maybe a mill and 20k XP (like unlocking basics now) Or you could spread them out, and get some random bigger boosts in all trees, and then wait for your end game unlocks.

It is very similar to the way the trees are now, with some minor reduction in nodes, and changing some placement, and values along the way.

They could also limit the end game nodes per tree as well.. So while you might have a general or two end game node, you can't have 3 or something like that.. so boats wouldn't benefit more.

Edited by JC Daxion, 06 March 2017 - 02:43 PM.

[pyrocomp]

Alistair Winter, on 06 March 2017 - 02:34 AM, said:

I just think people enjoy winning.

People enjoy having somethuing that they (individually) call enjoyment. That can be simple winning like in a slot machine. Or it can be some challenge accepted and afterwards beaten (a friend of mine dedicated loads of time to beat FF 7, IIRC, without a single level-up). Or just for some challenge disredarding win/loss.
I prefer some challenge and a feeling of completion or achievement. Thus a loss in a game against a well match opponent that outmanueverded or outsmarted me is better than the win in a game that was baby seal clubbing. But that's me.
Light currently can be one-shotted which is frustrating. Bigger mechs rarely get killed in a single blow. All the novelisations give you the impression that the mech is a strong sturdy machine and you want that feeling. Bigger mechs give it, Lights don't. What to do is... well, there is no way to do this now, it was decided somewhere at closed beta time (like speed dependence on actuall mech mass, e.g. factoring spent ammo, or malfunctions), I think. But this is not the main point of discussion now.

Anyway, to the OP. The skill tree should be renamed to the 'upgrades tree' (or 'tweak tree') and get logic in it. Thus no random skills placement. There was a good start with separation of upper and lower agilty. I'm for further aplication of this logic. The actual tree can be scrambled or linear I don't think that those options are really different.
I'm for less nodes and less points to take, BUT ONLY IF. If the nodes become complex in their returns and for higher nodes the trade-off are a must. E.g. getting further speed upgrade should also reduce your turn rate or getting hardened armor should reduce heat dissipation for example. The nodes are tweaks, not a strait-up OPgrades.

[Alistair Winter]

JC Daxion, on 06 March 2017 - 02:40 PM, said:

I just think lineal tree is just a bad idea for exact reasons of your comparison. It's not an RPG, but even in RPG's we have skills we don't want/use ext. I can name tons of games that had a skill i needed to get, just to unlock the one i wanted. this is just not new and always about wasting points in the name of balance.

Well, first of all, "It's not an RPG" is never really a convincing argument. Different game genres borrow ideas from each other all the time. RPGs borrow elements from FPS games and vice versa. There's no principle that dictates FPS games should not have RPG elements, except personal preference. And most of the skills in the current skill tree can be explained through technological upgrades to the mech in question, not just magical abilities.

Second, all game genres have bad designs. If buying the low level skills are just "wasting points in the name of balance", I dare say you're doing it wrong. Ideally, there's a more important philosophy behind the linear structure, such as (but not limited to) gradually introducing more complex abilities as players level up (making the gameplay increasingly complex and challenging) or getting players to make the above-mentioned choice of versatilty vs specialization. Versatilty is not a waste, it's a strategical choice.

[AngrySpartan]

Alistair Winter, on 06 March 2017 - 02:11 PM, said:

On the other hand, I would be very, very surprised if PGI fundamentally changed the structure of the skill system. I think they are 100 times more likely to just reject skills outright and go back to modules, rather than doing a radical change like a non-linear system, or taking inspiration from Witcher 3 or Mass Effect 3. I think what we'll get is going to be a variation of what they've already made, if we get anything at all. Because that's the way PGI has worked this far. They either make small adjustments or they rip it out of the game all together.

Well, Russ was very devoted to make it happen. "It will happen by March latest" or smth. Anyway, I don't think they can back off as they did with Energy draw, that's like admitting that they aren't capable to do it right.

JC Daxion, on 06 March 2017 - 01:44 PM, said:

while i agree with, having less nodes, just to make it a cleaner system.. I think the way it's done is ok.
as far as not taking nodes you don't need/want.. It's completely a balance issue, I'm sorry min/maxers should NOT get everything they want, it only throws the system more out of balance.

That is why WItcher 3 style system works so well. Every node is independent, yet you have to unlock certain amount of nodes to get the most tasty. That's everything that players were byzzing about in the feedback topic: less skills, no need to take every junk to unlock the best, limited min-maxing, cleaner UI. And on top of that it's flexible!

What surprises me is that players still prefer linear trees according to the poll. Can't see any pros in the linear trees, compared to the Witcher's system.

Edited by AngrySpartan, 06 March 2017 - 03:09 PM.

[Prof RJ Gumby]

1. Size is OKish I think. Didn't go deep into that, I kinda feel the node cap is a bit too high, but maybe I'm wrong

2. The complaints about having to buy "useless" nodes to get "best" are kinda strange to me. Yeah, it's not a perfect system, but the point is in deciding whether it is worth it to grind through the tree all the way to that last node to have all of them. Yet somehow people start their reasoning from the axiom that "YOU MUST" have every speed tweak/range/cooldown node there is. Barely any math and tests were done, but people KNOW that they HAVE TO get all that "BEST" nodes to have your mech optimal, no matter the price. Well, maybe you don't? Maybe having that last +1,5% is just not worth all those "useless" nodes? Maybe trading that last of the "best" nodes that will cost you triple for 3 a bit less useful, but easily accessible nodes is also a viable choice?

People wanted a proper meaningful skill tree with actual choices, but now they're complaining they're forced to make hard choices and actually sacrifice something for something instead of just going: "firepower - check, mobility - check, durability - check". You want a proper meaningful skill tree with actual choices that let you easily max out most important skills? What?

3. Hard to me to comment on mobility, but judging from votes it seems OKish. I bet lights and meds could use some more love, as always.

4. Don't care much about the looks of the skill tree, as long as it serves its purpose.
EDIT: I would prefer it to be called upgrade tree or something. These are mech upgrades, not skills.

Edited by Prof RJ Gumby, 06 March 2017 - 03:12 PM.