51 replies to this topic

[BearFlag]

I congratulate PGI on the Skill Tree and for opening it up to review. I think it's the right direction and something much needed. I do, however, agree with many of the analyses offered and agree some tweaking is called for. This thread, though, will not be a reiteration of concerns already thoroughly discussed. Nor will I talk about the economics of the Skill Tree.

Instead, I'm going to offer a conceptual approach to skill trees that differs from the current. At present, the skill trees (old and new) and modules are all upside. They're all costless improvements (in performance terms). Consequently, stock mechs are laughed at, especially in FW, because unmastered and unmoduled they really are inferior. The idea here is that the skill tree should be more about specialization and less about total improvement.

To that end, dichotomies are used to present tough choices and definitive specialization. Most "improvements" in the game have a reasonable nemesis. So improving in one could cost in another. This may not be a popular idea, but it does tone down the free lunch aspect and makes stock mechs more competitive with specialized mechs. Of course not all improvements need have a downside, but most probably should.

A couple examples are here, one laser weapons and one a non-weapon. Ignore the provided numbers - they're spitballed.

The first is lasers and is based on the already in-game dichotomies present in the "Clan Paradigm". Increased laser damage and range at the cost of increased heat and duration.



Now the choices are real. It's not just gimme, gimme. IS lasers can specialize to be more Clan-like at a cost of heat and duration. Clan lasers can specialize to be cooler and quicker - at a cost. Ideally, this would be calculated to be a 50% reach. That is, an IS ML could be tweaked to the half-way point of a C-ERML. Likewise the C-ERML could reach half-way to being an IS ML. Both so tweaked would be very similar in performance at that half-way point. Leaving lasers stock is viable option now if the pilot sees the specialization costs as too high. The IS also has the option in this example of taking their lasers down a notch. A Locust 1E might want to bring in six "light" versions of the ML with reduced damage and heat. The choices are tough, but the possibilites are pretty interesting.

The next example is Speed vs. Maneuverability. Seems a likely trade-off. A Fokker Dr.1 at 90 mph can turn on its wing while an SR-71 at Mach 3 and 80,000 feet needs nearly 20 miles to turn around.



Here again, a trade off choice is presented. Either branch could be worthwhile depending on the performance wanted. And here again, gains in one area are offset in another. The stock mech is compensated, so to speak, for being unspecialized.

The examples present the general concept. It is easily extended to other trees. Specialization over raw power gains is the idea. This approach may or may not encourage build-out diversity. Seems it would definitely promote specialization variety, though. It's also possible players won't want to part with their gimmes.

Edited by BearFlag, 12 February 2017 - 03:18 PM.

[Pyed]

I very much agree with the concept. This is similar to my post today:
https://mwomercs.com...al-tradeoffstm/

[BearFlag]

Pyed, on 12 February 2017 - 04:14 PM, said:

I very much agree with the concept. This is similar to my post today:
https://mwomercs.com...al-tradeoffstm/

Ah yes. Very similar. Nice.

[Johnny Z]

I have only been on test quickly and checked out some of the trees entirely only quickly so far, but I really like the direction also. Some tough choices and a 100000% improvement over the current skill tree.

[Quicksilver Aberration]

So basically like the original weapon modules that everyone avoided for a reason?

[Pyed]

Quicksilver Kalasa, on 12 February 2017 - 06:29 PM, said:

So basically like the original weapon modules that everyone avoided for a reason?

I'm not familiar. What did they do?

[BearFlag]

Quicksilver Kalasa, on 12 February 2017 - 06:29 PM, said:

So basically like the original weapon modules that everyone avoided for a reason?

I wasn't here then. But after looking back, it appears those weapon modules were pathetically weak and offered near zero specialization. They were replaced with the million CBill power creep modules. Am I wrong?

Edited by BearFlag, 12 February 2017 - 07:16 PM.

[Pyed]

Oh, this:

Quote

- Increase the maximum range of a Medium Laser by [5] meters, increase heat output by [0.2]
- Cost: [8.500] GXP

?

The concept is fine. Those values are terrible though.

[BearFlag]

Pyed, on 12 February 2017 - 08:19 PM, said:

Oh, this:


?

The concept is fine. Those values are terrible though.

That's what I found too. Crazy. Point 2 heat for 5 meters? Big ouch, little gain. Really, it's not even the correct trade off given the IS/Clan spectrum. Range/Duration seems more appropriate.

[Quicksilver Aberration]

Pyed, on 12 February 2017 - 08:19 PM, said:

The concept is fine. Those values are terrible though.

Not really, especially like the trade-off in the example. No one is going to make the trade-off for adding range at the expense of duration because extra duration makes that extra range kinda moot. Probably the only point that is a reasonable trade-off is for IS pulse weapons because they already have a stupidly short duration and are too short of range.

Sorry, but it is going to end up stupidly hard to balance because of how important those values are and how they interact with each other differently depending on the laser type. That's why these kind of "choice" trees end up being false choices, because there is an optimal path that just depends on your weapons whether you even want to go down it. The worst part about these sort of "trade-off" trees is that in some cases you won't even go down the tree; if I'm better off avoiding a whole thing, then is that even a good feature?

[Y E O N N E]

BearFlag, on 12 February 2017 - 08:35 PM, said:

Really, it's not even the correct trade off given the IS/Clan spectrum. Range/Duration seems more appropriate.

Mmmm, I don't think that's right, either. Mechanically, an increase in duration most directly offsets an increase in damage. That PGI decided the longer range ER LL should take longer to deal damage than the standard LL was because they couldn't figure out any other way to differentiate the two weapons (TBQH, they could have and should have used heat or cool-down).

[BearFlag]

Quicksilver Kalasa, on 12 February 2017 - 08:43 PM, said:

The worst part about these sort of "trade-off" trees is that in some cases you won't even go down the tree; if I'm better off avoiding a whole thing, then is that even a good feature?

Rather makes the point about pilot decision, doesn't it? Stock may be YOUR best choice for a given weapon class. Use your SP's in other areas.

Quicksilver Kalasa, on 12 February 2017 - 08:43 PM, said:

"...stupidly hard to balance..."

I disagree. It's a computational problem with a computational solution informed by experience. It's only a serious problem for dart-chuckers.

Quicksilver Kalasa, on 12 February 2017 - 08:43 PM, said:

"No one is going to make the trade-off for adding range at the expense of duration because extra duration makes that extra range kinda moot."

Tell that to pilots who gravitate to Clan tech. The paradigm is already in play.

I strongly doubt that the potential of hyper-specialization would be deemed by all pilots to be "not worth it." On the contrary, and for both Clan and IS, build specialization opens up vast flexibility. I could go down the list of "what ifs." Would an assault brawler trade 4 kph for +50% turn rate? Would an ER sniper trade +25% duration for +25% range?

When it comes to "trades", it's really a matter of reasonable scales, isn't it?

Edited by BearFlag, 12 February 2017 - 10:15 PM.

[Pyed]

Quicksilver Kalasa, on 12 February 2017 - 08:43 PM, said:

The worst part about these sort of "trade-off" trees is that in some cases you won't even go down the tree; if I'm better off avoiding a whole thing, then is that even a good feature?

That's because people tend to design these things with even or near-even tradeoffs, which I would agree is stupid. The gains should clearly outweigh the (still meaningful) negatives and materially affect the way the mech/system handles and what role it promotes.
--Which is still a problem with values, not concept. BTW I don't agree with BearFlag's value choices, as they're too incremental and the penalties are way too relatively high.

I think it can be done, however:
https://docs.google....dit?usp=sharing

[Cato Phoenix]

Yea, I'm not a fan of this.

There's no point in leveling a mech if the end result is just somewhat different but not improved once the +/-s add up. Something like speed at the cost of mobility or laser range at the cost of heat is not going to be generally beneficial.

Edited by Cato Phoenix, 13 February 2017 - 06:49 AM.

[Fox2232]

Nice try. It may even look good to someone who is mathematically impaired.

While I agree with trading one stat for another and would easily propose many meaningful ones. One should fist understand what is benefit and what is disability.
So, lets take all of your propositions and put them into practical battle perspective:
-4% Heat & -4% Damage => This node does not alter Heat/DPS ratio and therefore does nothing. Its effect is same as removing Laser... With small difference, removal of Laser does not cost SP and has greater effect than 5SP spent in tree
+3% Damage & +5% Heat => This case even makes heat/DPS ratio worse. I'll rather take 6th laser which will increase both Bamage and heat by 20% than investing 5SP here as that gives... +15% damage & +25% heat.
-5% Duration & -5% Range => In this case -5% Duration translates to approximately 1.6% higher rate of fire (and apparently 1.6% more Heat generated per second as result) so full effect of 5 nodes is: 8% more DPS and 8% heat generated per second - 25% range
+5% Range & +5% Duration => This node is actually great, Not sure where negative part is tho. Because as previous node, it does not alter Heat/DPS and while damage is lower, I can just pop-in another laser to compensate for loss and gain free +25% range.
-5% Cooldown & +3% heat => Two nodes back, we had less dumb version of this: In total, This will increase DPS by around 17% and from this increased rate of fire we'll get proportionate 17% more heat generated per second. Then We will add bonus 15% of heat generated
Really? So +17% DPS and +32% more heat? Why not to take 6th laser and gain 20% more damage and 20% more heat per second?
- - - -
Then there are those "Maneuverability" & "Speed" nodes: They are attempt to somehow simulate basic real world physics.
Those things should be implemented implicitly without stating them on nodes:
Faster you go, harder it is to change direction: E=m*c^2, I am sure everyone heard of this. But this is not important.

Would not you rather see engine vs mech weight => speed & acceleration dynamic behavior?
- Like, You take 100t mech and configure it to 95t and it implicitly moves faster than it would while being configured to full 100t.
- Or someone shoots off your AC/20 arm. Now you are 17 tons lighter. Why not to move faster?

[Quicksilver Aberration]

BearFlag, on 12 February 2017 - 10:06 PM, said:

Rather makes the point about pilot decision, doesn't it? Stock may be YOUR best choice for a given weapon class. Use your SP's in other areas.

Is it really a choice if it is purely based on your build? If it is solely based on the build you are running, and running stock is your best choice, why do we even need the skill tree? Answer is we don't, it is unnecessary.

BearFlag, on 12 February 2017 - 10:06 PM, said:

I disagree. It's a computational problem with a computational solution informed by experience. It's only a serious problem for dart-chuckers.

Tha computation is never going to be perfect, and it adds to the complexity that quirks already add because certain skills have different impacts on weapons. For example increased duration has a MUCH larger impact on things like iERLL and cERLLs than it does IS pulse.

BearFlag, on 12 February 2017 - 10:06 PM, said:

Tell that to pilots who gravitate to Clan tech. The paradigm is already in play.

Except there is more to Clan tech than just longer range and longer duration, if anything the longer duration is to keep the increased damage in check, not the range.

BearFlag, on 12 February 2017 - 10:06 PM, said:

trade 4 kph for +50% turn rate?

Most definitely since the speed isn't as big of deal.

BearFlag, on 12 February 2017 - 10:06 PM, said:

Would an ER sniper trade +25% duration for +25% range?

Lol no, that would never be something ER snipers trade-off with.

Pyed, on 13 February 2017 - 04:53 AM, said:

That's because people tend to design these things with even or near-even tradeoffs, which I would agree is stupid. The gains should clearly outweigh the (still meaningful) negatives and materially affect the way the mech/system handles and what role it promotes.
--Which is still a problem with values, not concept. BTW I don't agree with BearFlag's value choices, as they're too incremental and the penalties are way too relatively high.

What exactly is the point in negatives if the gain outweighs them though? You are better off just going with smaller positives.

[Pyed]

Quote

Would an ER sniper trade +25% duration for +25% range?

Quicksilver Kalasa, on 13 February 2017 - 10:38 AM, said:

Lol no, that would never be something ER snipers trade-off with.

IS LLAS:
Range 450
Duration 1
DPH 1.29

IS ERLLAS (LLAS comparison):
Range 675 (1.5x / +50%)
Duration 1.25 (1.25x / +25%)
DPH 1.13 (-0.16)

*CERLLAS (ERLLAS comparison):
Range 740 (1.096x / +9.6%)
Duration 1.5 (1.2x / +20%)
DPH 1.1 (-0.03)

You sure about that?

[Pyed]

Quicksilver Kalasa, on 13 February 2017 - 10:38 AM, said:

What exactly is the point in negatives if the gain outweighs them though? You are better off just going with smaller positives.

What is the point in having different weapon types?
Compare pulse lasers vs. normal/er lasers. They have different advantages and drawbacks. They're for different roles!
All the weapon choices exist for this very reason!

The point is that you can tweak your mech to fill the role you want it to fill.
Also, it is a heck of a lot more interesting when you have to make meaningful decisions sacrificing one part of your mech's performance for a gain elsewhere.

And, you know, that's how literally everything works in reality. You want more top speed? Sacrifice weight or acceleration or fuel economy or something else!

[rook]

Why make the hard choice picking choices that have a negative effect, why not have the choice be a one or the other? Say you have 5 points, with 5 levels of range increase and 5 levels of heat reduction: pick your favorite combo. All range, all heat reduction, or a smaller combo of both.

[Pyed]

rook, on 13 February 2017 - 11:30 AM, said:

Why make the hard choice picking choices that have a negative effect, why not have the choice be a one or the other? Say you have 5 points, with 5 levels of range increase and 5 levels of heat reduction: pick your favorite combo. All range, all heat reduction, or a smaller combo of both.

For the same reason we don't have long laser, medium laser, and short laser with slightly adjusted stat bonuses from the same base weapon with uniform crit slots and weight.

It's an actually interesting choice whether you want to go with mediums or small pulse or medium pulse with fewer heat sinks or fewer medium pulse or whatever. Ya gotta pick, and they each have their relative drawbacks.