23 replies to this topic

[Pinkie Pie]

It was a while after reading, experimenting, and digesting the new skill tree before my opinion on what was bugging me about it finally had a word to succinctly describe it.

Granularity.

Or to put it in another way, how fine, small, or precise something is. Something with a high granularity would be a 0-100 slider in increments of 1, while that same slider with low granularity would only increment in units of 25. Put in the context of MWO’s Skill Tree, we are confronted with something with an inordinate amount of granularity.

There are 229 individual nodes, by my last count, and of them we are expected to finish with 91 choices. There are only 55 distinct skills as well.

=Firepower= (73)
(10)Range
(10)Laser Duration
(5)Velocity
(11)Cooldown
(8)Heat Gen
(5)High Explosive
(5)Missile Spread
(2)Missile Rack
(5)LBX Spread
(5)UAC Jam
(2)Magazine Capacity
(5)Gauss Charge

=Survival= (22)
(5)Reinforced Casing
(5)Skeletal Density
(5)Armor Hardening
(5)Shock Absorbance
(2)AMS Overload

=Mobility= (40)
(5)Kinetic Burst
(5)Hard Brake
(5)Torso Pitch
(5)Torso Yaw
(5)Torso Speed
(5)Arm Pitch
(5)Anchor Turn
(5)Speed Tweak

=Jump Jets= (20)
(5)Heat Shielding
(5)Vent Calibration
(5)Vectoring
(5)Lift Speed

=Operations= (25)
(5)Quick Ignition
(3)Speed Retention
(3)Hill Climb
(4)Improved Gyro
(5)Heat Containment
(5)Cool Run

=Sensors= (27)
(5)Target Info Gathering
(5)Target Decay
(5)Sensor Range
(2)Target Retention
(5)Radar Deprevation
(2)Seismic Sensor
(2)Enhancd ECM
(1)Advanced Zoom

=Auxiliary= (22)
(4)Consumable Slot
(4)Capture Assist
(2)UAV Range
(1)UAV Duration
(1)Extra UAV
(2)Enhanced NARC
(2)Enhanced Coolshot
(1)Coolshot Cooldown
(1)Coolant Reserves
(2)Extended Bombardment
(1)Expanded Reserves
(1)Enhanced Spotting

As an example, laser duration increments at a rate of -1.5% for each node. On the laser with the longest burn time, that’s only 0.0225 SECONDS of reduction. For cooldown, it’s 0.8% rate increase per node, shortening a 4s cooldown to 3.968s. UAC Jam chance reduces it by a paltry 1%, that reduces the chance of a uac5 jam from 15% to 14.85%. These choices are far too granular.

No choice, if it is to be a meaningful one, should have an effect that is too small to be noticed by the player in the course of a game. It cheapens the experience choice is supposed to represent, as well as making it many times more tedious to actually use.

The solution is both simple in practice as well as in design. Reduce the number of nodes throughout the tree and the number of selections to the player by a proportionate amount. As a corollary to this, certain nodes would serve to be combined into the same node to bring their value up, and in the case of the firepower tree, making it less prone to boating singular weapon types.

As an example, I’ve provided suggested number of nodes for each skill.

=Firepower= (15)
(2)Range
(2)Laser Duration & Velocity
(3)Cooldown
(2)Heat Gen
(1)High Explosive
(2)Missile & LBX Spread & UAC Jam
(2)Missile Rack & Magazine Capacity
(1)Gauss Charge

=Survival= (8)
(2)Reinforced Casing
(2)Skeletal Density
(2)Armor Hardening
(1)Shock Absorbance
(1)AMS Overload

=Mobility= (10)
(2)Kinetic Burst & Hard Brake
(1)Torso Pitch & Yaw
(2)Torso Speed
(1)Arm Pitch
(2)Anchor Turn
(2)Speed Tweak

=Jump Jets= (4)
(2)Heat Shielding & Vent Calibration
(2)Vectoring & Lift Speed

=Operations= (8)
(1)Quick Ignition
(1)Speed Retention
(1)Hill Climb
(1)Improved Gyro
(2)Heat Containment
(2)Cool Run

=Sensors= (13)
(2)Target Info Gathering
(2)Target Decay
(2)Sensor Range
(1)Target Retention
(2)Radar Deprevation
(2)Seismic Sensor
(1)Enhancd ECM
(1)Advanced Zoom

=Auxiliary= (15)
(4)Consumable Slot
(1)Capture Assist
(1)UAV Range
(1)UAV Duration
(1)Extra UAV
(1)Enhanced NARC
(1)Enhanced Coolshot
(1)Coolshot Cooldown
(1)Coolant Reserves
(1)Extended Bombardment
(1)Expanded Reserves
(1)Enhanced Spotting

Again, this is just an example spread of reduced node count, and this is assuming the total bonus provided from maxing nodes is roughly the same. This reduces the total number of nodes available to around 73, and the number of choices to be made on this tree somewhere between 25-30 Skill Points. A huge deal friendlier to the player and also makes it easier to do TRUE balancing on the skill tree. Because at the end of the day, for it to be a choice, picking one node needs to be just as worth taking as any other node for it to be a real choice, and not the illusion of one.

[BadBird]

This! Sooooooooooo much this! Too many nodes too little impact.

[Little Details]

but...but...then they'd have to make each node cost 3-4x the current amount! They are trying to be fair with such *cough* low prices as they are now in this 2nd PTS attempt.

There is a lot of utility (read: fun) in picking 91 boxes * XXX mechs.

Edited by LT Satisfactory, 04 March 2017 - 04:33 PM.

[AngrySpartan]

"Less is more", that's what should hang on the wall of PGI's employees who developed current skill tree.

Definetely like your thinking Pinkie, skill tree will definetely win by collapsing half of the nodes.

Yes, it will decrease overall cost of levelling a mech, which isn't something in line with PGI's agenda I assume. Yet, gameplay should go before C-bill sink they trying to implement, that's the only long term solution players would tolerate.

[Alistair Winter]

I agree 100%. That's one of the many things that PGI could learn from popular RPGs and their skill trees, or even combat sims like Armored Warfare and Star Conflict. In those games, when you unlock a new skill, you usually feel it. It's exciting, it's fun, you suddenly feel more powerful and you want to keep playing.

What the hell does a 10% improved laser duration feel like? Unless you're a humming bird, you can't really tell the difference at all. It feels like nothing. There's no joy, except the tiny satisfaction that you're 1/91 closer to mastering your mech.

Fewer skills would feel a lot more rewarding, it would be easier to balance and it would be a lot more fun to do decide your skills for each mechs if they only had 20 skill points to use. With 91 skill points per mech, I'm probably looking at more than 4 hours of assigning skills to my 150 mechs. That's not a good time, it's literally just clicking the skill tree 13,000 times.

[AngrySpartan]

Alistair Winter, on 04 March 2017 - 04:40 PM, said:

Fewer skills would feel a lot more rewarding, it would be easier to balance and it would be a lot more fun to do decide your skills for each mechs if they only had 20 skill points to use. With 91 skill points per mech, I'm probably looking at more than 4 hours of assigning skills to my 150 mechs. That's not a good time, it's literally just clicking the skill tree 13,000 times.

Good point I can easily imagine how to spend all the time I'll waste levelling my mechs on more important things.

[Tier5 Kerensky]

Many of the skills are just some you have to take, to get what you really want. For example those laser duration skills you'd likely get in a energy boat just to get heat reductions. That is why we have 91 of them for total, as most are fillers.

[Malrock]

too many choices is literally making people exhausted and unhappy with the game.

http://www.nytimes.c...7shortcuts.html

and

https://www.fastcomp...ing-you-unhappy

Edited by Malrock, 04 March 2017 - 07:41 PM.

[oldradagast]

Malrock, on 04 March 2017 - 07:39 PM, said:

too many choices is literally making people exhausted and unhappy with the game.

http://www.nytimes.c...7shortcuts.html

and

https://www.fastcomp...ing-you-unhappy

Very good articles. I recall a similar one where they did a study that involved too many choices in toothpaste. Seriously, go to a grocery store and stare in amazement at the bewildering array of toothpaste choices, each of which claims to fight one or two things, but not everything. So, what do you pick? Needless exhaustion from too many choices. Oh, and in that case, the active ingredient is basically the same in all of them, so beyond a certain point, it doesn't matter.

With regard to the skill tree, I honestly hate it at this point. Not just because of the nerfing of weak mechs, the loss of capability, and overall slap-dash nature of the thing. No, there's also the upcoming utter waste of mental energy needed to go through this tangled web of lunacy and try to figure out what I actually need since most of the skill maze is nothing but false choices and pointless grind. It is simply a waste of my time and energy and about as much fun as doing my taxes. That is NOT why I play games.

In the end, out of sheer frustration, I'll probably find 1 or 2 generic skill maze builds (aside from weapons) that work "well enough" on all my mechs, and just slap that on there because I have better things to do with my time than stare at a tangled mess of 220 skills. Oh, and the best part - think of how much "fun" it will be if the skill maze goes live... and then PGI decides to start mucking with it every few months, just like they enjoy doing with quirks! Argh!

Edited by oldradagast, 04 March 2017 - 07:55 PM.

[Malrock]

oldradagast, on 04 March 2017 - 07:51 PM, said:

Oh, and the best part - think of how much "fun" it will be if the skill maze goes live... and then PGI decides to start mucking with it every few months, just like they enjoy doing with quirks! Argh!

Oh yes talent trees are never static, and all games that have them recognize as they get changed / re-balanced the points need to be refunded so people can freely spec again. This is a basic tennent of every game that has a talent tree that can and does undergo changes.

[oldradagast]

Malrock, on 04 March 2017 - 07:56 PM, said:

Oh yes talent trees are never static, and all games that have them recognize as they get changed / re-balanced the points need to be refunded so people can freely spec again. This is a basic tennent of every game that has a talent tree that can and does undergo changes.

Exactly. So how much mental energy and time am I going to be forced to waste every few months relearning the skill maze because, you know, forcing people to respec is the same as adding content?! And let's not forget the respect cost, which (as far as I know) still exists in the proposed skill maze. Because I'm sure folks will just buy lots of premium time to grind past that.

The skill maze itself is a horribly flawed and illogical mess, but the door it opens - to ever-changing skill mazes and constant regrinding of skills - paints a dismal future for this game if this is implemented.

Edited by oldradagast, 04 March 2017 - 08:00 PM.

[omnomtom]

I agree entirely. A tree with ~75 nodes and ~30 points available is far easier to navigate for the player, far easier to balance, and makes each choice of where to spend a point far more meaningful. It makes future changes to the tree much easier for players to understand and adapt to. It also cuts down on the "going through every 'mech and assigning 91 points" issue.

As to the C-Bill and XP cost... I'd like to be optimistic enough to say that players would understand that cutting the number of points in 3 and tripling the cost per point works out to a net zero. There might be backlash from players who don't understand rudimentary math that "NOW IT'S 180k CBILLS PER NODE? OMGWTFBBQ" but... as long as the actual cost from 0 to mastery (and approximate benefit of 0 to mastery) stays the same, everyone who can multiply will recognize that the economics didn't change.

The one thing I can see being a problem for PGI's current design of the tree is that with fewer nodes, it makes it harder to 'hide' nodes they think will be stronger/more 'meta' behind less critical ones. This is something a lot of players are frustrated with, but I can understand the reasoning behind it - putting radar dep at the bottom of the sensors tree means players that want it do have to invest in other sensor nodes. It turns sensors tree from "Well, take 5 points of derp and 2 points of seismic and never look at that tree again" to an actual choice - "Do I need all 5 derps? That costs me 19 points... or I could only take 3 and a single seismic for 10 points." It's a way to balance the opportunity cost of different node choices, so that those "meta" nodes effectively cost more, and I think that's reasonable. Drawing up a smaller scaled one that creates a similar relative cost might take some work, but I certainly think it's feasible.

Definite +1 to scaling down the number of nodes and increasing the effect per node. I think that would be a big improvement for player comprehension and satisfaction with the tree.

Edited by omnomtom, 04 March 2017 - 09:25 PM.

[Sauron]

I agree mostly with the OP. I do also feel though that in addition to that, we cannot accept our "new and improved system" to nerf the hell out of all of our builds. Why should I be happy that I'm not going to get the same top speed? mobility, etc, etc. It's not just tradeoffs it's flat out impossible to get a mech back to the base it was at. I want my mechs to be better as a result of the new system, else why support it? After playing our mechs for a long time, grinding to get them mastered, etc. I really don't think it's going to be well-received to not only have to rebuild all and tweak all your mechs assuming you can afford to do so for all the mechs you currently have mastered, but also learn to accept your mech will feel inferior....IMO these are the kinds of things that drive big chunks of players away to other games. I am not a fan so far of this new system, I liked the idea of it and was excited about it, but so far, it's falling far short of any expectations I had for it.

[Warchild Corsair]

So much this! It was my first thought with the new skill tree and has stayed ever since.

PLEASE give us meaningful (!) choices, PGI!

[oldradagast]

omnomtom, on 04 March 2017 - 09:25 PM, said:

The one thing I can see being a problem for PGI's current design of the tree is that with fewer nodes, it makes it harder to 'hide' nodes they think will be stronger/more 'meta' behind less critical ones. T

The problem is hiding good choices behind trash is poor game design. WoW got away from it years ago, tabletop D&D started moving away from it in 2003, when 3.5 edition came out, and so on. Meanwhile, in the land of PGI where time forgot, we're going to move TOWARDS that failed gaming model with the skill maze.

GRIND is not CHOICE, and it is NOT content. The skill maze is simply a huge illusion of a deep and complicated system to level your mech and give it a role. In reality, it's a frustrating exercise in lunacy as you muddle you way through the tangled mess to get the handful of skills you need on basically all your mechs... after giving them worthless skills, like missile perks on energy boats, jumpjet boosts on mechs without jumpjets, etc.There are no roles - just which weapon you shoot better with, and which junk skills you take to get the same good ones.

PGI doesn't get it. The way to balance a powerful skill is NOT to bury it behind a long, boring grind of junk. That in no way changes the fact that you need that skill - it just takes more time to get there, which just ticks people off since their mech is less effective for a longer amount of time. And it's even worse for new players who will lack the resources to quickly get to the needed skills and who probably won't know what the needed ones are anyway and will instead get lost in the skill maze.

The skill system needs FEWER real choices that are reasonably balanced and a lot LESS GRIND.

Edited by oldradagast, 05 March 2017 - 07:16 AM.

[Lucky Rookie]

1000 times this!

[Malrock]

Yes the sheer size of the talent trees is overwhelming, and i suspect a new player wouldn't even know where to begin there are so many options. Plus because the buffs are so minor, you would hardly be able to tell that you have achieved anything. No one can reasonably expect to feel a 1% or smaller buff.

[Arkroma]

Yes I criticized that too many node caused tons of problems in my feedback as well.

Couldn't have said better myself, it just destroys a player's interest to even look at it.

I know I will just copy-paste something that I can find from the Internet if this is the final version.

[Bud Crue]

How did I miss this thread?


Spot on Pinkie Pie, spot on.

[Malrock]

Arkroma, on 05 March 2017 - 04:51 PM, said:

Yes I criticized that too many node caused tons of problems in my feedback as well.

Couldn't have said better myself, it just destroys a player's interest to even look at it.

I know I will just copy-paste something that I can find from the Internet if this is the final version.

Agreed.

And the solution to send people out side of the game is not a good one. People shouldn't be required to go outside the game to find solutions for poor implementation. that is how you get people to stop playing. But you are exactly right that is what people will likely do. It is what happened in WoW. everyone went to elitist jerks and found the mathematically best possible build and ran it, and never thought about the tree again.