12 replies to this topic

[Gagis]

Expectations
I am cautiously optimistic about the general approach to giving MWO a skill system that has at least some relevance. The current system adds practically no value to the game after all. I did not participate in the first test but it seems that combining the weapons into a single tree was successful and the right thing to do.

I have a background in playing games such as Path of Exile with extremely complex skill trees, and I was expecting to see something similar but much more simple here. That expectation was met.

UI
The UI is otherwise fine, but since players will practically be building their trees by finding shortest possible paths to the skill nodes they really want, it would be very helpful if nodes that add new functionality or otherwise noteworthy nodes would be highlighted with a different border or shading. This would make navigating the skill trees easier, and make it easier to wrap your head around the idea of having some uninteresting nodes you travel trough to to reach the good stuff.

Oh, and the Portal is great. I will use it instead of the Launcher from now on.

Number of skill points and nodes
Choises you make when choosing skills feel relevant, so I think the order of magnitude is right at least. There might be some room for working out optimal distances to reach the most important skill nodes, but no part felt completely wrong at least.

The concept of traveling trough suboptimal nodes to reach the ones you want
I like it, and it is absolutely necessary for balancing skill point costs and build choises. Do not listen to the forum whiners who want to have their cake and eat it too. Making routing between desirable nodes easier will be detrimental to the system.

Travelling trough a lot of less interesting nodes should be a significant cost we need to weigh against alternatives.

My comment from elsewhere on why bad nodes matter

Queen of England, on 03 March 2017 - 06:04 AM, said:

A far simpler way to force people to make hard choices is to just cut down the huge number of skill points you currently get to something manageable. Letting people pick between 100 good nodes with 20 skill points creates more interesting choices than picking between 100 good nodes gated behind 400 bad nodes with 91 points.

Not quite. The system that includes bad nodes gets much more interesting when those bad nodes are more useful to some mechs than others, which becomes a subtle factor in balancing and loadout design.

The bad mech with varied hardpoints all over its body will benefit from the otherwise bad nodes and gain a little bit more value per skill point from experience. The good mech with uniform hardpoints in good high locations will have much less use for more than one weapon type, arm or torso stats or other additional stuff, but will have to spend just as much experience in order to get 100% weapon cooldown or speed or whatever. Both mechs wanted and took the same thing, maximum weapon cooldown in my example, but one got more bang for their buck.

The pilot of the good mech will either have to suck it up and travel trough 2 bad nodes for the final cooldown node, or to be satisfied with 90% of available cooldown bonuses and spend those 3 points on some other tree instead.


The difference is subtle, but beautiful.

Edited by Gagis, 03 March 2017 - 06:44 AM.

[Gagis]

After a bit more testing I quite like the general design of the skill system. I had to make relevant tradeoffs when skilling different mechs.

[cazidin]

Gagis, on 02 March 2017 - 05:09 AM, said:

The concept of traveling trough suboptimal nodes to reach the ones you want
I like it, and it is absolutely necessary for balancing skill point costs and build choises. Do not listen to the forum whiners who want to have their cake and eat it too. Making routing between desirable nodes easier will be detrimental to the system.

Travelling trough a lot of less interesting nodes should be a significant cost we need to weigh against alternatives.

Alright. My first argument here is - why do less interesting nodes exists? Sure, in RPGs you'll always have less appealing choices, but none of them should be sub-optimal by design. Why AMS Overload be part of the Armor Structure tree when I either have no interest in AMS, which should be my choice, or on mechs that simply have no AMS hard points? That is a waste of skill points that I could, in most other systems like this one, avoid and rightly so.

My second argument is that you think this is the ONLY way to make this skill tree interesting? You can have meaningful alternatives and specialize your mech if this were made more competently. Players will gravitate towards the optimal path regardless of how many holes in the road there are, it just makes the journey rougher and the players less happy about it overall.

Instead, you can have 3-4 viable ways to build your mech and truly specialize in a specific field that suits your playstyle. In most RPGs this would be called picking your class. An Archer would rarely waste skill points on Axes when he could much more wisely spend them on levelling his Archery or acquiring a feat that synergizes well with his kit of abilities.

[Gagis]

Copy of my comments on linear trees vs. the skill web

Linear trees with diminishing results are boring design. In the current system the cost of reaching a desired node is different based on how much your current mech can benefit from the nodes you travel trough, which encourages you to seek synergistic loadouts and to experiment with optimal balances of stats. Those who benefit a bit more from the nodes you travel trough gain a bit more for their skill points, which is a way to balance mechs.

Expecting people to travel trough bad nodes to reach the good ones is perfectly fine and will alienate far fewer players than you think. It would just be easier to accept if the great nodes were clearly highlighted so they scream "I AM HERE, AND I AM GREAT, HOW DO YOU PLAN TO REACH ME, PUNK!?". The current UI doesn't really show much, so players get misconceptions about what it means to have to travel trough the skill tree for the best bits.

Quote

What I want to do is just go down and say, unlock all the range nodes and then not spend skill points on any other firepower skills.

This would be bad and encourage min-maxing. It is better that you can probably take about half of the range nodes with minimal waste of resources, but will have to consider if you really want range so much you want to travel to the other end of the skill tree for it, or if you'd be better off spending all those points on survival instead.

Edited by Gagis, 03 March 2017 - 06:49 AM.

[Gagis]

I edited the title of the post and added some musings on why I prefer a skill web with bad nodes and good nodes mixed together over linear design. This is now a PGI *** kissing thread where I tell them that their skill web is good and the people crying on the forums about having to buy useless nodes are bad.

[cazidin]

Gagis, on 03 March 2017 - 06:47 AM, said:

This is now a PGI *** kissing thread where I tell them that their skill web is good and the people crying on the forums about having to buy useless nodes are bad.

Thanks for your honesty, but you know that PGI doesn't often read these forums? If you want to kiss butt, try Twitter.

[1453 R]

I play Path as well (Scorching Ray CwC Firestorm for Legacy start, hoping to try some interesting Knife Nova builds later), and I can definitively state that PoE does not have outright worthless skill nodes, the way half of each of these trees do.

All nodes in Path of Exile's skill tree are at least minimally desirable for some manner of build - pathing and efficiency is based on spending the least points for the most impact, maximizing per-point efficiency in a maze-like tree with a hefty number of pathing options. Similar-ish nodes are ganged together in regions of the tree, and if you don't want a particular stat because it doesn't help you, it's easy enough to find a different route that gives you something you can use in order to get to something you really want to use.

Piranha's current skill trees do none of these things.

Pathing is extremely restrictive, and a significant number of nodes in almost every skill tree are flat-out worthless for a significant number of builds. Nobody needs gyro stabilization. Nobody needs hill climb. They are poorly designed, objectively not good nodes that do not accomplish the task of making a pilot's machine more dangerous or durable in a fight. These 'bonuses' are no such thing and should be replaced outright, not grandfathered in because they were bad modules nobody used before, and not used as efficiency-wrecking filler in front of nodes that actually accomplish things.

Even the weakest, most boring road node in PoE gives me ten points of something that at least marginally benefits my build. I don't have to deal with taking evasion nodes or life nodes or dodge nodes or whatever scattered throughout my energy shield nodes on a CI build because 'Defense' nodes are all mixed into a ridiculous slurry. Certain combinations of things are inherently less efficient in PoE because of the way the tree is laid out, but you never have to take something you just actively do not need at all (Ascendancies aside) to get to something you really like.

That doesn't feel good. That doesn't make for subtle balance, that makes for frustration and bad feels. If you're going to argue for road nodes to indirectly increase the cost of desirable skills more than undesirable skills, ask yourself why those undesirable skills should be there at all if their only objective is to act as a point sink for actual, desirable nodes.

[process]

I'm starting to think a tiered skill system would work best.

50 total skill points to spend.

Tier 1: basic skills, requires 30 points to unlock Tier 2.
Tier 2: advanced versions of basic skills, requires 15 points to unlock Tier 3.
Tier 3: end game skills, e.g. seismic, radar dep. 5 point.

There would be a lot of skills to create a variety of paths to tier up. No compulsory skills.

[FupDup]

Gagis, on 02 March 2017 - 05:09 AM, said:

The concept of traveling trough suboptimal nodes to reach the ones you want
I like it, and it is absolutely necessary for balancing skill point costs and build choises. Do not listen to the forum whiners who want to have their cake and eat it too. Making routing between desirable nodes easier will be detrimental to the system.

No, it's not necessary at all. There are two fairly easy things that can be done to balance out the "desirable" choices.

1. Split up the "good" nodes into a higher total number of nodes so that you have to invest more SP to max it out. "Bad" skills on the other hand should only take a very small number of SP to max out (like 3 at most).

2. There's also this strange, alien concept of making the bad nodes not be bad anymore. I know this sounds really heretical and stuff, but maybe, just maybe, we could like BUFF the values of "crappy" nodes to make them NOT be crappy anymore?

[Elrik Stormbringer]

I only have two main complaints with the new pts system and the phenomena of undesired skills and they are both cbill related. They are 1) That we may have to pay for upgrades that are not useable on the mech in question. If it were just XP I would be annoyed, but ok with this, but the fact that it is also costing Cbills is not acceptable 2) that there is no mechanism for testing the validity of skill changes before having to buy, and part and parcel with this lack, is that reskilling has a monetary cost.

Otherwise I am kind of digging the major changes to gameplay that this may be representing, it will be almost like a whole new game and play less twitchy with a longer ttk.

[Haji1096]

Gagis, on 02 March 2017 - 05:09 AM, said:

Expectations
I am cautiously optimistic about the general approach to giving MWO a skill system that has at least some relevance. The current system adds practically no value to the game after all. I did not participate in the first test but it seems that combining the weapons into a single tree was successful and the right thing to do.

I have a background in playing games such as Path of Exile with extremely complex skill trees, and I was expecting to see something similar but much more simple here. That expectation was met.

UI
The UI is otherwise fine, but since players will practically be building their trees by finding shortest possible paths to the skill nodes they really want, it would be very helpful if nodes that add new functionality or otherwise noteworthy nodes would be highlighted with a different border or shading. This would make navigating the skill trees easier, and make it easier to wrap your head around the idea of having some uninteresting nodes you travel trough to to reach the good stuff.

Oh, and the Portal is great. I will use it instead of the Launcher from now on.

Number of skill points and nodes
Choises you make when choosing skills feel relevant, so I think the order of magnitude is right at least. There might be some room for working out optimal distances to reach the most important skill nodes, but no part felt completely wrong at least.

The concept of traveling trough suboptimal nodes to reach the ones you want
I like it, and it is absolutely necessary for balancing skill point costs and build choises. Do not listen to the forum whiners who want to have their cake and eat it too. Making routing between desirable nodes easier will be detrimental to the system.

Travelling trough a lot of less interesting nodes should be a significant cost we need to weigh against alternatives.

My comment from elsewhere on why bad nodes matter

Not quite. The system that includes bad nodes gets much more interesting when those bad nodes are more useful to some mechs than others, which becomes a subtle factor in balancing and loadout design.

The bad mech with varied hardpoints all over its body will benefit from the otherwise bad nodes and gain a little bit more value per skill point from experience. The good mech with uniform hardpoints in good high locations will have much less use for more than one weapon type, arm or torso stats or other additional stuff, but will have to spend just as much experience in order to get 100% weapon cooldown or speed or whatever. Both mechs wanted and took the same thing, maximum weapon cooldown in my example, but one got more bang for their buck.

The pilot of the good mech will either have to suck it up and travel trough 2 bad nodes for the final cooldown node, or to be satisfied with 90% of available cooldown bonuses and spend those 3 points on some other tree instead.


The difference is subtle, but beautiful.

I disagree with your assertion completely.

Is arm pitch relevant to a mech with no arm mounted weapons ? No.

Is shock absorbance relevant to a mech with no jump jets ? No.

Is Reinforced Casing relevant at all ? No. critical hits don't matter. It simply easier just to kill whatever you are shooting at instead of hoping for a critical.

Is Target Decay relevant for a mech with direct fire weapons ? No.

Its a waste of skill points, that I have already earned. A waste of C-bills that I won't have enough of to master over 130 mechs.

If PGI wants to lower min/maxing they should have better gameplay mechanics, instead of punishing players with ham-fisted changes that are poorly implemented.

I don't enjoy driving well-rounded piece of crap mechs.

Lights and mediums agility is completely borked on the PTS. I will be canceling my Javelin pre-order if this PTS is impelmented.

[oldradagast]

There is no defense for bad skill nodes. They are flawed game design, full stop.

It's bad enough when presented as false choices - in short, you can choose from one of several options, but only one of those choices is actually worth taking. But it is even worse when the false choices are mixed in as XP paywalls to reach the good skills since now players first have to figure out which skills are actually worth taking amid all the trash, and then they still have to grind through trash to get to those skills. The end result is even more chances for new or inexperienced players to be hosed AND everyone has to waste their time pointlessly getting there.

More mindless grind through junk we don't need is not "content," and doubling the number of skills by filling the skill tree with trash is not "increasing skill diversity" since most builds are going to use the same narrow number of viable skills anyway.

Edited by oldradagast, 03 March 2017 - 03:44 PM.