0
I'm writing this all from the perspective of the few hours I spent playing and studying the PTS. Probably gonna go through and revise it a little with some more experience under my belt.
To start: The PTR is buggy, but provides a decent enough feel for how it'll play once it's all hammered out. Some quirks on some mechs straight up don't work, the JJ tree doesn't work, SRM's are busted, etc. so on and so forth.
The JJ tree was kind of a downer for me because I wanted to test out the SDR-5V with ******* insane jump jets, but I'll deal with it.
The biggest thing I could suggest to PGI is to let the system sit in the oven a little longer. It's not ready yet.
The first thing that struck me was just how expensive the skill system was. I get that it's supposed to be a replacement for modules and mastering mechs, but it's gonna make something like a Raven go from 2.5m stock > 7m for the kit, to 2.5+7+9.1m.
It's a huge increase in cost, that hurts players who wanna play Pokemechs (Like me) and newer players who don't quite get what's going on yet.
IMO the cost of the skill tree should be changed. Not necessarily the EXP cost, though that wouldn't hurt too much, but bringing down the cost of a skillpoint by half or (optimistically) 75% would go a long way in that regard. That way we're not paying double the cost of a light mech and a third of the cost of an assault to make it workable.
*Upon reflection a blanket change probably isn't smart. The end nodes (Radar Derp, Speed tweak, etc.) should get more expensive as the trees go on. Speed Tweak 5 shouldn't cost as much as Speed Tweak 1. Also, PGI could think about frontloading the skills. Speed Tweak 1 should be more powerful than speed tweak 5, even if just by a bit, so that it'll reward people who really wanna go fast (By dumping more points into that tree), but if you just want the first couple points of speed tweak you can do that and still get a good benefit.
A flat bonus across the whole tree is not a good choice.
The second thing that struck me is the actual system, in regard to a mech by mech basis, could use some refinement.
A Kodiak shouldn't have the same number of skill points as the Pretty Baby, full stop. The PB will need the structure/armor (Billboard chest + XL Engine) in addition to some badass weapon skills (It's got **** hardpoints), but the KDK should only have access to either:
1) Limited structure/armor points,
2) Limited twisting points, or
3) Limited weapon points.
Pick two or even 1 and a half.
It doesn't need to change on a tonnage basis, either. The 91 points that you can dump into armor/structure for a light is HUGE in terms of fixing them from the Rescale. My lights can actually take a hit before folding, now, which is amazing.
Just keep it on a chassis by chassis basis and we're good.
The trees themselves are a problem. I'm probably gonna go make a writeup on that later, but the way the trees are laid out is incredibly linear and should be reworked.
They're less like trees, and more like branches without that core tree, that require a lot of filler to get to the stuff you need.
I shouldn't need Hill Climb modules on a spider to get to Speed Tweak.
Weapon skills and boating prevention
This requires a restructure.
Right now the system only rewards boating with how much the tree requires you to sink into a particular weapon type's branch.
It needs to be laid out like an actual tree.
Just because I like lasers I'm gonna give my example using them.
Right now you need to sink 19 skillpoints into the tree to make Lasers good, when on my Raven I'd be totally happy with just Range and Heat quirks.
IMO it should have a catchall skill, let's say hypothetically Laser Heat 1. That branches four ways, to the left for Cooldown, the right for Duration, Center-Left for Heat, Center-Right for Range. That way I could sink 5 points into Heat, and 5 points into Range, and we're good.
Again, frontloading is necessary here.
In terms of the functionality of the tree design, It's not bad right now. It could use some edits, like if you're unlocking a node with GXP vs. Mech XP, they should be different colors. That way if you can't apply the changes at the end, you don't have to respec the whole thing.
There's a lot of clicking you need to do, which is kind of a pain in the ***. Some people suggested click+drag functionality, which ain't bad. A tree restructure might fix this whole thing anyways.
TL;DR
Fantastic framework. But big problems that, if pushed to live, will hurt more than they will help.
Firstly: Way too expensive. We shouldn't have to pay the cost of a medium/heavy to make a light good.
Second: Tree needs a lot of work, more specifically a restructure and another look at node rewards.
Second Addendum 1) Points in the skill tree should be modified on a per-chassis basis. KDK-1 will probably need the full 91, but the KDK-3 does not.
Third: Some functionality tweaks would be nice.
PGI, I'm begging you not to throw this out. Just refine it a little more. It just needs a bit of work and it'll be fine to release.
Onward to the horrible, personal opinions PGI could take inspiration from.
These are all just numbers I'm throwing out there.
So in regard to what I was talking about earlier with the skill tree, I figure the best course of action is to have 3 seperate trees, Utility, Firepower, and Operations.
This not only would help clean up the UI a little bit (less boxes, more tabs possibly), but also help refine the actual trees and help sort out the various problems I personally have with the system.
Let's start with firepower, just for shits and giggles. More specifically the IS Laser tree.
So you've got your 3 brands of lasers, ER, Standard, and Pulse.
From here we can see a few things, that 1) it's consolidated all of the lasers, and 2) it doesn't completely shut a player away from one choice or the other.
You can take a triple ERLL double ML jester, spec into the ER tree, and not get punished for taking STD ML's. You still gain *something* from it, opening up the door to mixed builds. Alternatively, you could take an ML Black Knight out, spec'd into the STD lasers, and mix in a few ERLL's for ranged poke.
Alternatively, you could go for the more generalist STD tree and gain access to the duration quirks, and by extension be able to mix it up with LPL's or MPL's with an STD spec, a la the BLR-3G with 3 ML 3 LPL, or change it up and take ERLL's without worrying too much because of the Range quirk they share, making them less efficient but still an option.
This also comes with a built-in safeguard for keeping clantech in check, just by not giving them access to the shared quirks, so they're forced either into the Pulse branches or the ER branches. The tech is good enough on its own that it'll live.
The important part, if you're going to take anything away from this idea, is to share quirks between weapons systems, so you can flex between the two as necessary.
It's no fun being shoehorned into one weapons system or another, Mechwarrior, and by extension MWO, is all about pots and pans and sticking them to your mech to see how it works out. That said, some specialization is necessary in order to make sure that you can't just take a Kodiak with 4 LBX's, get sick of it, and slap in 2 UAC/5 2 UAC/10 or double ERPPC/double Gauss without drawbacks. You want a mech that's good at brawl? Spec for it. Just don't be surprised when that thing can't trade for **** because it's cooldowns are long or its UAC's keep jamming.
Alternatively, frontloading the quirks with the most powerful quirks first allows people more flexibility with their skillpoints.
If you wanna take a Cheetah with points in everything and swap back and forth between PPC, SL's, ML's, and SPL's, that's great. But chances are you'll be beaten by someone with a more specialized build and spec, who is playing the mech for a specific reason, which is how it should be. You're paying for that flexibility with effectiveness, and being a jack of all trades that's a master of all wouldn't make for a good design.
