

Scaled Skill Bonuses To Encourage Mixed Builds
#1
Posted 11 February 2017 - 04:24 PM
For example, the current skill tree applys LRM velocity bonuses in equal increments, with 5 levels of 4% each for a total increase of 20%. A scaled system could apply the velocity bonuses thusly:
Level 1: 6%
Level 2: 5%
Level 3: 4%
Level 4: 3%
Level 5: 2%
The total velocity increase is still 20%, but the bonuses are front loaded to provide the most benefit in the early levels. A pure LRM boat would likely fill the entire LRM tree anyway as it would receive the same total benefit as the current PTS system. But a mixed build would be more likely to benefit from spreading skill bonuses around, maybe taking only Level 2 of the velocity boost (for an increase of 11%) and use the remaining skill points to delve into the energy or ballistics bonuses, which would be similarly scaled to provide a significant improvement in the early levels.
The aggressiveness of the scale could be adjusted for each bonus type as needed. Generally, more aggressive scaling should promote mixed builds (diversity) and less aggressive scaling promotes specialization.
#2
Posted 11 February 2017 - 04:40 PM
It's the fact that all of the buffs are must-haves; there's no give to taking except an arguably value-less opportunity cost to take something else.
#3
Posted 11 February 2017 - 08:24 PM
Pilots who boat weapons are usually min/maxing mechs. This occurs because it is about efficent and effective builds. In Beta days you would actually see min/max builds that had mixed weapons. When Quirks were introduced min/maxers naturally jumped on whatever weapon any mech was quirked for which made "boating" more effective and efficent than finding weapons that work well together or "synergy".
An example of that old min/maxing was the gauss+ppc combo (before the charge time was added to firing gauss). It was very effective because gauss and ppc were both long range weapons. The charge time for gauss put an end to min/maxers use of those builds.
Currently I run an Cyclops 10q with a mix of SRMs and LRMs. It can do a lot of damage, but a LRM boat will outperform it at long range and an SRM boat will tear it apart close in. Diversity can mean weakness or strength depending on your opponent.
Like I said your idea has merit, if min/maxers find that spreading skill points into multiple weapons creates deadilier mechs we will see less "boats". But since quirks are staying for most mechs min/maxers will probably tend to build mechs to boat whatever weapon it is quirked for despite the option to diversify with skillpoints.
Edited by SilentScreamer, 11 February 2017 - 09:19 PM.
#4
Posted 11 February 2017 - 08:38 PM
But, really, what I'm driving at is that when I go into the skill tree, there are all these nodes which are obvious must-haves, because to leave them would be suicide. These include cool-run, heat containment, speed-tweak, structure density, and armor hardening to name a few. I spend points on all of these and they give me all these nice perks, but in game-play terms I didn't really pay anything at all. I'm not actually specializing my 'Mech to do anything, I'm just making my 'Mech strictly superior to the same one driven by a guy who is jumping into his first matches. The weapon buffs are not that big of a deal, and PGI is actually abusing the hell out of them to try to make crappy guns perform more closely to better guns. It's an awful, awful tax.
So, even if you front-load the progression, that doesn't change what I have. I just don't have to dive as deeply into each tree and the other guy can maybe attain closer performance in less time. But what I should be seeing are skill trees that compel me to truly give something up in order to gain a buff in a particular area. You can either do it in the skill values (i.e. specifying for longer range lasers results in increased heat and cool-down), or you can do it by placing a limited subset of skills into different trees, with the best bits deeper in, and not enough points to fully spec more than one.
You cannot let people just pick all of the most potent bonuses a la carte.
Edited by Yeonne Greene, 11 February 2017 - 08:39 PM.
#5
Posted 11 February 2017 - 08:43 PM
However, actively punishing mixed builds should also be avoided.
Yotta's solution is a partial fix- It reduces disparity between maxed boat and new player or mixed loadout mechs- but there is still some incentive to max the tree for a boater. It has great merit.
Maybe a small boost in max unlockable nodes for mechs with a low number of mixed hardpoints (or other underperformers) would assist to limit this disparity also? (similar to what was done with module slots- only more tuneable).
What would address yeonne's issue is balance among choices- a hard proposition that will never be fully met, but one can at least try (and the current system suffers that issue also- more so possibly).
Edited by Wibbledtodeath, 11 February 2017 - 08:45 PM.
#6
Posted 11 February 2017 - 08:59 PM
Then I build my mech with UACs and PPCs. I have to spend 10 points to get the same level of weapon performance, 5 points that could have gone to my performance.
So how does scaled bonus help?
#7
Posted 11 February 2017 - 09:57 PM
#8
Posted 12 February 2017 - 03:35 AM
#9
Posted 12 February 2017 - 03:43 AM
Yeonne Greene, on 11 February 2017 - 08:38 PM, said:
But, really, what I'm driving at is that when I go into the skill tree, there are all these nodes which are obvious must-haves, because to leave them would be suicide. These include cool-run, heat containment, speed-tweak, structure density, and armor hardening to name a few. I spend points on all of these and they give me all these nice perks, but in game-play terms I didn't really pay anything at all. I'm not actually specializing my 'Mech to do anything, I'm just making my 'Mech strictly superior to the same one driven by a guy who is jumping into his first matches. The weapon buffs are not that big of a deal, and PGI is actually abusing the hell out of them to try to make crappy guns perform more closely to better guns. It's an awful, awful tax.
So, even if you front-load the progression, that doesn't change what I have. I just don't have to dive as deeply into each tree and the other guy can maybe attain closer performance in less time. But what I should be seeing are skill trees that compel me to truly give something up in order to gain a buff in a particular area. You can either do it in the skill values (i.e. specifying for longer range lasers results in increased heat and cool-down), or you can do it by placing a limited subset of skills into different trees, with the best bits deeper in, and not enough points to fully spec more than one.
You cannot let people just pick all of the most potent bonuses a la carte.
well technically Gauss+PPC is as dead as it has been for years since the charge mechanic killed it... PGI just introduced Clans so cGauss+cERPPC has become the go to metamech tryhard cancer player choice.
#10
Posted 12 February 2017 - 03:54 AM
Personally I think I would prefer if all weapon skills were general, like 10% cooldown applies to all weapon systems. That would even the field a bit, even tho boats would still perform better due to better weapon sync.
#11
Posted 12 February 2017 - 03:57 AM
YottaHurts, on 11 February 2017 - 04:24 PM, said:
Level 2: 5%
Level 3: 4%
Level 4: 3%
Level 5: 2%
I get what your hoping to achieve with this, but you will also open the door to boost mech's across the board with the least amount of effort by grabbing all the level 1 skill improvements that *aid* your mech for your preferred load-out and play style quickly.
Having a curve would be more suiting to achieve more or less the same thing, but *forcing* to invest a bit more effort to get to the good stuff.
Those who focus & boat will still work there way down the skill tree to min/max the numbers, regardless of what is used...
For example,
Level 1: 3%
Level 2: 4%
Level 3: 6%
Level 4: 4%
Level 5: 3%
#12
Posted 12 February 2017 - 04:54 AM
Wild Cat, on 12 February 2017 - 03:57 AM, said:
You are overthinking it. Using many different weapon types is a disadvantage in itself and there is not bad compensating for it.
@ topic:
Similar threads:
Andi’s
Mine
#13
Posted 12 February 2017 - 08:32 AM
I_AM_ZUUL, on 12 February 2017 - 03:43 AM, said:
well technically Gauss+PPC is as dead as it has been for years since the charge mechanic killed it... PGI just introduced Clans so cGauss+cERPPC has become the go to metamech tryhard cancer player choice.
No. There are several IS 'Mechs which can do it, too, and they are properly powerful with it; they just can't jump, and it's that jumping ability which is really necessary due to the KDK and which pushes the Clan version over the top. You don't see anybody complaining about cGauss + cER PPC combos on Ebon Jags or Warhawks, because there's nothing superlative about that. It's only those 'Mechs which can pop-tart or bring bring 50 points.
#14
Posted 12 February 2017 - 08:42 AM
However, if you have armor buffs of 5%, 4%, 3%, 2%, 1%, and then weapon buffs of 1.5%, 1.25%, 1%, 0.75%, and 0.5%, then you might do something different. There would be value in going 5-4-3 and 1.5-1.25. If you value survivability, you could go for 15% armor, and if you value damage you can get 5% damage, just like before.
But like others have said, weapon skills need to be combined to not encourage boating.
4 user(s) are reading this topic
0 members, 4 guests, 0 anonymous users