This skill tree system encourages boating only one weapon type, and then not switching away from that weapon type ever on the chassis that you've upgraded. A KDK3 player, unless he buys three kodiak 3s and then upgrades them separately (which is prohibitively expensive in both cbills and EXP,) will not be able to freely switch between LBXs, UACs, and Gauss loadouts without facing a steep opportunity cost in not receiving any bonuses for the switched weapons.
Chassis with multiple weapon types that were already fairly bad are now even worse. Mechs like the Vindicator that depend on bracket builds, which are terrible, have now had their baseline quirks drastically reduced. They would have to give up one of the other incredibly important skill trees to buy a second weapon skill tree to compensate for what baseline offense quirks and modules would normally have given them. Mechs like this used to get extra module slots so that they could take multiple cooldown/range modules for different weapon types, but they are now playing under the same node point restriction of optimal mechs.
Some mechs end up having to buy skills that are functionally useless: mechs with no arm weapons needing arm speed, mechs with no arms getting arm movement, etc., while other skills are useless regardless of what mech they're on: speed retention, 360 degree target retention, reduced falling damage, hill climbing, etc.
As a quick prediction, I think that the community would reach a consensus One True Build with very slight modifications-- maybe three to five points-- that will be common across all chassis and variants with exception to very specific gimmick builds:
1.One weapon skill tree, for the one weapon type that the mech is going to be primarily boating.
2.The lower mobility skill tree, mostly for speed tweak but also turning speed, breaking, acceleration, etc.-- All bonuses that are too important to pass up.
3. The defensive skill tree, with all skills except for AMS overload being taken, because extra armor and extra structure are too important to pass up.
4. Sensors, with people taking the most efficient path to try to max out radar deprivation and seismic as the highest priorities. Once again, they're too important to pass up.
5. Operations, because of cool run, heat capacity, and startup speed as priorities, because the first two directly impact damage output and the last gives more forgiveness for overheating your mech and not overriding.
The thruster skill tree is virtually useless, because any mech that wants a miniscule thruster bonus like that will only get a benefit if it's already covered in thrusters. Any mech that is already covered in thrusters won't find anything higher than it can jump onto with 5% more initial boost and 5% more burn time. What does vectoring even do? Does it increase the horizontal speed that you are going while you jump? Make this mechanically clear, because otherwise it seems like an utterly useless skill to have. If vectoring increased your land speed by a significant amount or let you turn in mid air, it would be a really cool and useful skill.
The upper body skill tree has the most important skills, torso twist degrees and twist speed, hidden through a pile of useless gobbledygook. It is not worth taking those when the other 5 important skill trees exist.
lol the consumables skill tree.
If there's another skill tree I forgot, it isn't important because the importance of the first five have made it so that I don't even recognize it as something that exists. Even if you had extra node points, you would likely end up spending them on a second weapon tree at this point.
------
I suggest making it so that you just plain buy X/91 nodes for a chassis variant, and these skill tree selections will stick. You can then buy another skill tree template that has the same number of skill nodes unlocked, maybe 3 million cbills or a handful of MC, and you will be able to select a different combination of skills. Let people buy as many different skill trees per variant as they want that they can swap between after games. Make these skill trees available across all mechs of the same variant. This will allow people to experiment without compromising builds that they know will work, and will allow omnimechs to actually change out equipment without losing stuff. If you program this, maybe you could also program in a feature to save mechlab builds to go alongside it.
Reduce the size of the skill tree by merging nodes, or make it so that players can climb up/laterally on the nodes instead of only down. This will help bad mechs with multiple weapon systems, and will be good for the upper body tree, which is vastly outshone by the others.
I'd also suggest just plain reducing the amount of nodes that unnamed supermeta mechs get to reduce the time to kill baseline (this IS what you guys were going for, right?), but I know that will just turn this thread into the 100th nerf X mech thread.
Edited by Commoners, 08 February 2017 - 07:19 PM.