22 replies to this topic

[Doctor Dinosaur]

Hello,

I would like to discuss the most important and apparently overlooked matter with the new system.
We have too many skill points.
91 points is just enough to skill everything relavant for each mech in his role...and that's a problem. You have to take specific skills for each role and there's hardly any chance to pass up on them in favour of other skills (e.g. a brawler will probably never take auxiliary skills and drop armour for it).

This leads us to "mandatory" skills and trash skills (for each role) which every player will try to achieve. This in turn leads to a massive buff across all mechs (everyone who needs armour gets 10% more armour) which in turn prevents people to take distinct skills (e.g. a max sensor cyclops).
I feel that this is quite the opposite of what is the goal of this skill system overhaul.

Of course, values can be nerfed or buffed, but I will still take 0.0001% more armor over an +300% AMS skill on a Kodiak.

Solution?
Either
- Give us less skill points, perhaps around 60.
OR
- Perhaps there should be incentives for taking skills from different trees (the ammo capacity on the operations tree instead of the (U)AC tree is a wonderful example).
Add one +5% armor node on the auxiliary tree, so scouts are not totally screwed and heavies are tempted to go into that tree, too. Things like that.

[Fox2232]

Why not to limit people to 40 SP? Can you then predict what nodes will people take? I think I would be able to do that with 90% success rate.
If we go down to 20SP, I am sure I'll get close to 95% prediction on which nodes will people take.
For trees of current size 91SP is just fine.

Because if we excuse those few people who will completely ignore META nodes (And they should have choice to ignore them.), it takes like 40 nodes spent into important stuff. And then we have another 51SP to pick what we LIKE.
And customization is not about what we need, but about what we like.

People should ask themselves: "Where mandatory stuff I need ends, and where fun begins?"
Is it fun to have customized tree and play your way? Or was it more fun to have same bonuses as everyone around?

Less SP = Closer to old system = illusion of choice while your choice is very limited.

[Doctor Dinosaur]

Yes, I want a limited choice. The oppsite would be to just buy every single skill...we kind of have that right now.
With 91 skill points I STILL can predict close to 95% which nodes people will take. Because there are plenty of nodes, but you can pretty clearly divide them between must have and trash (results may vary between brawlers and scouts).

We have like 4 skill trees which are relevant (Weapon(*2), Armor, Movement). I want having to focus on 2 out of them...without being forced to take exactly THOSE two.
I want to make a choice between burst or DPS. Between speed and armor. I want the skills to be significant to the gameplay and to be relevant to my playstile.
If everyone "needs" the same skills, we don't need no skill system AT ALL. That's are huge problem. If everyone has speed tweak, noone has.

[Cato Phoenix]

So everyone currently gets a blanket buff - manueverability, speed, fire rate, turn, etc.

You have to have enough to get back to that baseline, and then the hope is that the remaining pool can be used for specialized tweaking.

[Doctor Dinosaur]

Cato Phoenix, on 13 February 2017 - 05:24 AM, said:

So everyone currently gets a blanket buff - manueverability, speed, fire rate, turn, etc.

You have to have enough to get back to that baseline, and then the hope is that the remaining pool can be used for specialized tweaking.

Exactly. My point is, that the blanket buff already "needs" (most of) the 91 points, giving almost no room at all for that specialized tweaking. It should be the other way round: No blanket buff, only specialized tweaking. Then it would be meaningful.

[Errinovar]

You know less skill points is a great idea if we are talking about meta mechs.. but when you start talking about mechs like Centurion chassis they absolutely need to maintain the current baseline, and even then they won't be remotely competitive simply due to the the structure of the mech itself. And I don't think the cent is even the worst possible mech, though I think the release of all the new mechs have pretty much made it obsolete.

So what is your suggestion for all the low performing mechs? Are you suggesting that they should share a baseline with the Kodiac or Timberwolf?

[Doctor Dinosaur]

That's another problem:
A Kodiak has 91 points to improve.
A Centurion will need X points just to stay where he is RIGHT NOW and has less than 91 points to improve. But no skill tree system will remove this issue, so no use of discussing it, I guess.

[Mechteric]

91 is too many to try to keep track of, and part of the problem is the trees can be too granular. Kanjashi's video about it makes a good point about streamlining the skills in a better way that could then lowes the total SP points to something like 50-70.

Edited by CapperDeluxe, 14 February 2017 - 05:39 AM.

[Gas Guzzler]

I didn't feel like 91 was too many at all, I mean, considering that its supposed to make up for the old skill tree system plus modules. Most of my builds were forgoing the upper chassis agility and infotech trees already, in favor of some extra weapon buffs, specifically when I had 2 different weapon trees to go through, which compared to original system means I gave up my mech module slots to specialize in offense.

I do however think some of the skills were unbalanced, like being able to get 50% accel/decel compared to the current 15% is a little high.

[RoadblockXL]

They need to have fewer points along with fewer, but more significant nodes.

You should be choosing between things like -10% energy cooldown vs -10% energy heat, rather than choosing between two different piles of quirks that each contain a bunch of quirks you don't care about mixed in with the ones that you do.

[n00biwan]

They have to give you lots of points since they expect you to buy 20 things to get to the ones you want.

It's a set of trees that are about preventing things, disabling choices, not about enabling them. Scared of min-maxing meta players? Get over it, I'm not one, but they're out there, and whatever your system is, the "best fit" will be found, so make it easier for the other 80%, please?

If it becomes that EVERYONE has to have seismic etc and will go down those trees to get them, why not just let ppl just choose a skill and level it, like, whats that thing.... a character sheet?

Odds are ppl will drop a few points down it, and will or won't spend the additional points based on how worth it it is... maybe level based bonuses with diminishing returns? I won't be buying lvl 3-5 of seismic on a mech where I won't feel the extra range as relevant as more speed/turn, beefing up my srm/laser combo etc instead.

Why have people buying turn rate to get to speed tweak? How about "Spend 4 points on mobility to unlock lvl 1 speed tweek, 6 for lvl 2" type ties to prevent people from just taking the "top" skills?

Plenty of RPG style character development methods out there to choose from, why like this? All that will happen is that you'll have 10k mechs boating 1 weapon type with all the same "other" skills that they've been forced to buy to get to the few appropriate to their role.

I have no doubt that my dual UAC5 HBR will be tougher, deadlier and 20% cooler than before, but so will everyone else's mechs be just as enhanced, in probably the same ways. Players might be "individuals" but we act like algorithms when it comes to games, we'll find the obvious/best solutions, which will be the same ones for everyone.

All that said, I'm happy in general with the fact it's changing, kinda like the direction, and the balls it's showing to take on a move like this with a historically angry and easily disappointed player base. (I mean there is some substance to that, so fair play really).

edit> I didn't really mention the cost of the whole thing (clearly too incensed by the the complete non-sensemakingness of the skill trees as are), but it all costs TOO DAMN MUCH! You expect someone skilling a locust to be able to earn all that money and even save for another locust? Good player maybe, average player? not so much...

Edited by n00biwan, 14 February 2017 - 04:20 PM.

[Pjwned]

Cato Phoenix, on 13 February 2017 - 05:24 AM, said:

So everyone currently gets a blanket buff - manueverability, speed, fire rate, turn, etc.

You have to have enough to get back to that baseline, and then the hope is that the remaining pool can be used for specialized tweaking.

No, you really don't need to have enough to get back to that baseline actually; just because there's a baseline that we have now doesn't mean it's good.

It would be better for the game if the baseline was zero, but a compromise works too by making it closer to that zero baseline and also making choices in the skill tree actually matter so that at least it's not a boring piece of crap like the old (current) tree.

Fox2232, on 13 February 2017 - 04:58 AM, said:

Why not to limit people to 40 SP? Can you then predict what nodes will people take? I think I would be able to do that with 90% success rate.
If we go down to 20SP, I am sure I'll get close to 95% prediction on which nodes will people take.
For trees of current size 91SP is just fine.

Because if we excuse those few people who will completely ignore META nodes (And they should have choice to ignore them.), it takes like 40 nodes spent into important stuff. And then we have another 51SP to pick what we LIKE.
And customization is not about what we need, but about what we like.

People should ask themselves: "Where mandatory stuff I need ends, and where fun begins?"
Is it fun to have customized tree and play your way? Or was it more fun to have same bonuses as everyone around?

Less SP = Closer to old system = illusion of choice while your choice is very limited.

Or there could be some actual effort put into it so that we don't have skills that are always picked first.

[50 50]

Fox2232, on 13 February 2017 - 04:58 AM, said:

Why not to limit people to 40 SP? Can you then predict what nodes will people take? I think I would be able to do that with 90% success rate.
If we go down to 20SP, I am sure I'll get close to 95% prediction on which nodes will people take.
For trees of current size 91SP is just fine.

Because if we excuse those few people who will completely ignore META nodes (And they should have choice to ignore them.), it takes like 40 nodes spent into important stuff. And then we have another 51SP to pick what we LIKE.
And customization is not about what we need, but about what we like.

People should ask themselves: "Where mandatory stuff I need ends, and where fun begins?"
Is it fun to have customized tree and play your way? Or was it more fun to have same bonuses as everyone around?

Less SP = Closer to old system = illusion of choice while your choice is very limited.

Going to completely.... disagree.

With less skill points what would the meta be?
As soon as it boils down to having to sacrifice certain enhancements, we can't say that taking tree A over tree B is the best option as you can't take skills from tree B.... or C, D or E etc.

So if one player says: I've got a good ballistic boat here, I'm going to max out my Autocannon tree. That's 20 points gone. Leaves me with (lets go with the 60 point max) 40. I'll put 20 into lower mobility because I want my speed tweak. Left with 20 and suddenly you've got the hard decision of whether to take upper mobility, defense, sensors or one of the others.
You max defense.
Well, your sensors are rubbish. Can't detect a mech 300m in front of you and you've got no radar dep, seismic etc.
Plus, now you can't torso twist to spread your damage and the mech doesn't have some of the other options that are equally as cool.

Right at the moment on the live servers we get all the bonus. Our mechs get faster, run cooler, become more mobile.... and then we add in the modules to make them shoot faster, have more range, get better sensors.

It takes around 50 skill points to get our mechs back to where they are under the current system before modules.
But we then have another 41 points to spend which is like adding 8 modules onto the mechs.
It's way to much.

So at the moment, dropping the skill point maximum to 60 (seems like a good number) does force us to diversify. It will either mean missing out on some skill trees entirely and creating specialized mechs, or, taking a piecemeal approach and having a more general build.

[MechaBattler]

I honestly believe they made the skill trees the way they did to increase TTK. In particular they only give us 5% cooldown. Instead they give us duration, jam chance, or velocity. Then everyone is obviously going to take survival. Heat skills aren't as high as on live. There's also more acceleration and deceleration than we have on live.

But I think if they were to introduce more things to the trees. We could have customization on the level of not being able to take everything you want. Especially if they add more unique skills. That aren't just more damage.

Edited by MechaBattler, 14 February 2017 - 08:20 PM.

[Reno Blade]

Yea that topic came up few times. Kanajashi suggested some options and I also had these listed in my post (see signature).

Mainly to split the skill-classes in 3 groups similar to what we have in current module system (weapon, mech, infotech).
Then limiting the max SP per group, so you can NOT put 75 SP into the Mech class and then maybe use Weapon/Infotech branches ( all mechs spending the same 50-75 SP on survival, mech-ops and movement ...)

e.g. Weapon / Mech / InfoTech maximum points per skill class (total 91 SP available)
light = 40 / 50 / 40
med = 50 / 60 / 30
heavy = 60 / 50 / 30
assault = 65 / 45 / 20

I didn't see any better option yet, but just reducing the max SP will only most likely lead to people still using the most important survival points and only very few others (as they will just skip the least important weapon/info skills).

Edited by Reno Blade, 14 February 2017 - 10:25 PM.

[Andi Nagasia]

the thing is we could make do with 50-60 points, then make all the Trees useful,
we could also do with keeping the number of points we have but increasing the number of nodes,

imagine an defense tree, that is split into
=Structure=
lets say 15 points of 3% each 45% Buff,
More Structure but more likely your things will get Crit'ed out,
=Armor=
lets say 15 points of 2% each 30% Buff,
More Armor less than Structure but less likely your things will get Crit'ed out,

now you have to dedicate 30 points to these inorder to get both,
in a 60 point max thats half your points, but its a choice people can make to make their mech tanks,

[50 50]

Andi Nagasia, on 14 February 2017 - 10:38 PM, said:

the thing is we could make do with 50-60 points, then make all the Trees useful,
we could also do with keeping the number of points we have but increasing the number of nodes

now you have to dedicate 30 points to these inorder to get both,
in a 60 point max thats half your points, but its a choice people can make to make their mech tanks,

Dropping the number of skills we could buy would be a very easy change at this stage and has the double benefit of suddenly making the cost of getting all of them a fair bit cheaper as well.
A trade off to a reduction in points could be more streamlined paths, though I think there is an element of needing to pick up some additional skills to get the maximum levels so it cannot be completely optimized. Be nice to make those paths have an appropriate flow to them. Eg: If you want to get Seismic, you must also get some points in sensor increase along the way OR the benefit of some skills are improved with others. ie. You can go straight to level 5 in seismic, but it's range is very limited unless you also boost the base sensors.

Completely agree that there is room for more skill nodes, I think the defense and sensor trees could be expanded on. Bit like what you are suggesting there Andi, but it could be that there would be a tree for structural defensive bonuses and another related to defensive equipment.
For example:
Structural advantages could include:
Armour
Internal structure
Better item HP
Reduced critical chances
Reduced fall Damage.

Where as Defensive equipment options include things such as:
AMS
CASE(Not sure what bonuses would work here but it's an example)
.. and probably a few others I can't think of right now.

The same could be done for sensors with an offensive and defensive tree.

Lets also consider that with a tech change we are potentially up for an increase in different trees, particularly weapons.

Either way, there is more that could be added to make our choices more diverse but ultimately reducing the points is what will make the biggest difference.

Edited by 50 50, 15 February 2017 - 02:13 AM.

[Vybz]

I might have a problem. Being a bit unaware of the new skill point system (came back after a longer break) I assigned more then 91 skillpoints to a mech.

Is there a way to get the points assigned that exceed 91 back to be used in other mechs?

[Reno Blade]

Vybz, on 01 May 2018 - 04:13 AM, said:

I might have a problem. Being a bit unaware of the new skill point system (came back after a longer break) I assigned more then 91 skillpoints to a mech.

Is there a way to get the points assigned that exceed 91 back to be used in other mechs?

No.
The only use you will now have is to be able to unlock more skills for any "experimentation" you want to do (e.g. switching weapons, or armor for agility...).

Only other option could be to contact support and ask very nicely if they could help, but I think skill points are out of their reach/area even.

[Thrudvangar]

Hmm looks like what you guys want is everything good, nothing "bad".... a super hero skilltree...

Skilltree is good as it is right now. You have to sacrifice something to make another thing good...

Either sacrifice on firepower tree to get like agility/armor or the other way. What you all want is all buffs at once.

Hopefully this will not happen...