Jump to content

Feedback On Potential Armor Structure Quirks And Rolling Boosts Into Base Stats

Armor Structure

8 replies to this topic

#1 Nothing Whatsoever

    Member

  • PipPipPipPipPipPipPipPipPip
  • Bad Company
  • Bad Company
  • 3,655 posts
  • LocationNowhere

Posted 12 September 2015 - 12:13 PM

I figured I should share some of my thoughts relating to what could be applied to mechs' armor and structure values.

The AS7-D's structure quirks, for example, allow that variant to be much tougher, but I think a new baseline should be established for all mechs.

A little background, over the years I've looked at different balance elements to help myself understand this game better, and one thing I've been trying recently has been to standardize how I display the information in relation to original values and for MWO's values.

What I have noted is that when Armor and Structure was doubled to deal with the increased lethality of weapons that themselves have been more than doubled (when trying to translate their values of one turn to 10 seconds of real-time) that there is room to raise Armor and Structure a bit more.

So in other words Armor and Structure was raised 200%; where Heat Capacity and most weapons have been raised from a range of 211% to 1390%. Then factoring Fast Fire, quirks and modules that increase rate of fire, and those weapons are made more lethal.

Therefore, what I suggest is that there is room to increase base armor and structure either through quirks or adjusting base values. It seems that a 50% boost from current MWO values could be sufficient, then quirks could further accommodate hitboxes as necessary on a case by case basis.

So, that would take the current 32 units per ton of Standard and raise that to 40 units per tons.
IS Ferro increases from 35.84 to 44.8.
Clan Ferro increases from 38.4 to 48.

Here is a chart that increases the values of Armor and Structure and has the current values for comparison:

Posted Image

The values in green, are the increase in armor with the 50% boost. The other values in light blue are the calculated Tonnage using the new values of 40, 44.8 and 48.

So to restate, the idea is to raise the base stats of Armor and Structure, and then apply quirks to accommodate hitboxes from there as necessary.

Edited by Praetor Knight, 12 September 2015 - 12:23 PM.


#2 Nothing Whatsoever

    Member

  • PipPipPipPipPipPipPipPipPip
  • Bad Company
  • Bad Company
  • 3,655 posts
  • LocationNowhere

Posted 14 September 2015 - 11:40 AM

Following up with this idea, here is an example with the Atlas variants that could be considered in distributing the quirks:

Posted Image

This Table shows the totals for the Current Atlas Baseline for Armor and Structure, 2.5x Boost, Boar's Head PTS Quirks, AS7-D PTS Quirks, and how I would reconsider distributing quirks to boost up the Atlas variants for Armor and Structure.

Weapon and Equipment Loadouts would then be where there can be the unique differentiation of the variants should take place, and the variation could be in relation to Sensor differences with mounted Equipment and stock weapons, for example where the Atlas K could have the best long range sensors of the group having a stock long range loadout, with maybe the S having the best short range sensors of the Atlas.

And I would strongly consider sustaining similar mobility / agility profiles, where what needs to be turned off, until it can be addressed, are the Mech Tree Efficiencies in the PTS.

#3 Nothing Whatsoever

    Member

  • PipPipPipPipPipPipPipPipPip
  • Bad Company
  • Bad Company
  • 3,655 posts
  • LocationNowhere

Posted 15 September 2015 - 09:28 AM

Here is an example of redistributing the TDK structure quirks to both Armor and Structure.

Posted Image

#4 Nothing Whatsoever

    Member

  • PipPipPipPipPipPipPipPipPip
  • Bad Company
  • Bad Company
  • 3,655 posts
  • LocationNowhere

Posted 16 September 2015 - 12:23 PM

using the TDR-9SE quirks, here is a comparison for 65 Tonners:

Posted Image

#5 Thorqemada

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • 6,368 posts

Posted 16 September 2015 - 11:51 PM

OP - you unbalance the inter-weightclass Balance when you simply multiply values.

A Light suddenly needs a multitude of additional damage to kill an Assault while an Assault only needs a fraction of that additional firepower to kill the Light.

You have shown it very well in the "Difference" row of your table.

Edited by Thorqemada, 17 September 2015 - 04:24 AM.


#6 Nothing Whatsoever

    Member

  • PipPipPipPipPipPipPipPipPip
  • Bad Company
  • Bad Company
  • 3,655 posts
  • LocationNowhere

Posted 17 September 2015 - 04:51 AM

View PostThorqemada, on 16 September 2015 - 11:51 PM, said:

PO - you unbalance the inter-weightclass Balance when you simply multiply values.

A Light suddenly needs a multitude of additional damage to kill an Assault while an Assault only needs a fraction of that additional firepower to kill the Light.

You have shown it very well in the "Difference" row of your table.


No need to be childish and dismissive. Remember that this is not being looked at in a vacuum. This topic is focusing on armor and structure, yes. But other factors are at play here that I haven't mentioned or haven't been clear on.

For example, when has that situation/relationship not existed in these Mechwarrior games (Between the average Light and the average Assault)?

Take the TDK with 4xMLs or MPLs against a brawler Atlas now. It would not take much if the Atlas pilot has good aim and the Light makes a mistake.

Even with 1.0x armor, there is a wide difference:

Posted Image

Also if you take the values and compare 132 to 459 versus 393 to 1261, the ratio is rather tight with these values actually.

But what matters is how the values play out live. In the PTS, the AS7-D quirks really helped make it feel tough to fight. The TDK was the variant with the largest boost, so that is why I used those sets of values to compare to 2.5x values.

And remember that weapons have seen significant boosts as well, with armor trailing behind. 4xMLs on a TDK with quirks and modules can pump out 20 damage, every 3.44 seconds. So ~60 potential damage in less then 12 seconds. So when Hit Reg works, and a light player is using cover, and focusing fire with allies ~240 Damage to core that Atlas is not impossible to see in a single engagement.



So regarding Inter-weight Class Balance, a better criticism would be that only certain mechs or variants should receive such boosts to armor and structure. And it can be as a means to provide different mechs different Armor profiles.

That way an Awesome could have more armor than a Zeus or Victor (as it was in lore and then have quirks boost up needy hitboxes,) here's and old look at one idea for Armor Profiles using weight (so Ferro could then see better trade-offs against Endo too):

Posted Image


So 19 tons of Standard Armor at 40 units per ton is 760 units total. 5.5 tons of Standard at 40 is 220 Units.


Another can be to change the Heat System, or directly address weapon values. I'd have no problem resetting everything to 1.0x and going from there.

Lastly, another alternative is connecting weapon convergence to the new InfoTech variations. Where the little light not only has better speed, agility, and hitboxes than an Atlas, but also has better sensors so it can utilize it's smaller number of weapons more efficiently than the Assault. There's a big thread on that topic.

#7 Thorqemada

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • 6,368 posts

Posted 17 September 2015 - 05:05 AM

What is childish about saying that giving the Atlas 2,5 times more additional Armor than the Jenner will unbalance the weightclasses?

It is about that additional shots, the additional exposure time you need to kill a Mech of a given weightclass and multiplaying values will favour the big weightclasses that have big numbers bcs their absolute gain outclasses the absolute gain of the light wiéigthclasses.
You yourself have posted it in your table in the difference row where a 35t light has 60 Points of additional Armor and the 100t Assault has 154 Points of Armor (roguhly 2.5 times the gain of the Light).

Strucutre works the same way - the Light has a gain of of 31 and the Assault a gain of 78.

Now the gain of Armor+Structure for the Light is 91 and the gain of the Assault is 234 Points.

That means that you change the balance between the weigthclasses in favour of the heavier Mechs that have higher base values.

That is all i do say... :)

PS: Quirks are quirks and should not be taken for granted - better have a good base balance that is independant from to much PGI Quirkening.

Edited by Thorqemada, 17 September 2015 - 05:07 AM.


#8 Sarlic

    Member

  • PipPipPipPipPipPipPipPipPip
  • Hearing Impaired
  • Hearing Impaired
  • 4,519 posts
  • LocationEurope

Posted 17 September 2015 - 05:10 AM

View PostPraetor Knight, on 14 September 2015 - 11:40 AM, said:

Following up with this idea, here is an example with the Atlas variants that could be considered in distributing the quirks:

Posted Image

This Table shows the totals for the Current Atlas Baseline for Armor and Structure, 2.5x Boost, Boar's Head PTS Quirks, AS7-D PTS Quirks, and how I would reconsider distributing quirks to boost up the Atlas variants for Armor and Structure.

Weapon and Equipment Loadouts would then be where there can be the unique differentiation of the variants should take place, and the variation could be in relation to Sensor differences with mounted Equipment and stock weapons, for example where the Atlas K could have the best long range sensors of the group having a stock long range loadout, with maybe the S having the best short range sensors of the Atlas.

And I would strongly consider sustaining similar mobility / agility profiles, where what needs to be turned off, until it can be addressed, are the Mech Tree Efficiencies in the PTS.


I was working on this already. I only had to apply the differences on armor/structure. But you beat me by it.

Almost each variant needs a unique look on it. The only thing i need to find out what kind of calculation would you use on offensive, defensive and neutral.

Here is a very short list how i see them.

View PostSarlic, on 14 September 2015 - 08:40 AM, said:


I wanted to test the PTS more before i write my butt off for Atlas feedback. I had too little time.
But so far in short this is my opinion: (all very reasonable and logical)

(Specific talking about armor/structure)

AS7-D - Should get both structure and armor, but more structure. It should stand in any situation you can think of: close/mid/long- range. Such a great allround Atlas. It needs to be a wall.

AS7-D-DC - Mix of structure and armor; but more armor. It has ECM already, but it loses two CT laser hps. Thus more armor. It's a stealth/ambush Atlas after all for brawling. King of brawls.

AS7-S - Both structure and armor. It should be neutral so you can decide to outfit as a pure juggernaut, support or w/e. 4 M hps.

AS7-K - Mix of structure and armor, but more armor. Typically longer range / walk with the team Atlas. Can be build as a brawler as you focus fire on medium/longer range with your team who are fighting at the very same time against streakers, boaters and SRMs mechs. But i prefer to be a teamwalker and have a optimal mix of long and medium range.

AS7-RS - Both structure and armor, but more structure. Its the typically peek and shoot Atlas due stacked 2 laser hps in both arms but no CT. Can be outfitted as a brawler, but i personally prefer more to medium range. Any cover is good cover. Losing your arms=half of your fp.

AS-BH - Same as above but then neutral quirks. Stacked laser hps both arms. (Do not own one but its sbout the same as RS, but then brawling range).

Thats how i think of it. In the Atlas guide i have illustrated the playstles and about each variant in a table.

Edited by Sarlic, 17 September 2015 - 05:13 AM.


#9 Nothing Whatsoever

    Member

  • PipPipPipPipPipPipPipPipPip
  • Bad Company
  • Bad Company
  • 3,655 posts
  • LocationNowhere

Posted 17 September 2015 - 05:54 AM

The idea is to get a proper baseline.

The ratio of what is gained is intended to keep pace with weapon performance. Sure the numbers seem to be big, but the base gains I've looked at so far are increasing consistently for the base line of stats. Then from there, mechs like the Jenner should then see quirks to the CT to account for its CT Hitbox.

And as I mentioned before, there can be some differentiation with boosts on different chassis, where it can easily be argued that the Direwolf and King Crab would receive no or very little armor/structure quirks compared to the Atlas, staying with 100 tonners. So, for example, there could be:

Direwolf total armor/structure 760/391 | King Crab total armor ~717/500 | Atlas total armor/structure 768/493

So the actual values for the Dire and K Crab can also increase like the example of the Atlas, but the specific components and relationship to stock armor levels can be a point of consideration for how much they do go up by.




If there is a possibility of dialing back base weapon stats, while retaining the ability to quirk and choose modules or reigning in the agility/mobility of the bigger mechs compared to the Lights, then you can sustain parity in a different way.

Mechs with fewer hardpoints could see bigger weapon boosts that stack with modules, mechs with multiple hardpoints would then see few or no weapon quirks and likely have strengths in other areas.

Also, the big boosts from Mech Tree Efficiencies and how much agility is gained from mounting larger engines is another very important aspect. In the lore the 300 Engine of a Light was at a different price point of an Assault and thus would provide a different performance. So even with a stock 300, a 2xEfficiency 100 tonner easily enough agility to move and stick it's back against a wall and plaster a light. If both mechs turned a tad slower and twisted a hair slower, but the Light retained an agility advantage, then that would be an important matter of parity between the trade-offs of the Light to Assault mech.

Another thing to look at, for example, the Clan Heavies, they all have the same twist, turn, arm speeds, having the 5 ton difference between them, and all of them have better agility than an Orion with an XL 360 Engine! It's a matter of how engines were coded, but that is another important aspect to keep in mind too.



We have been told that mechs are being evaluated on their four main aspects, Firepower, Movement, Survivability and InfoTech, so in this thread I've mostly been detailing aspects of Armor and Structure after seeing the big boosts to the AS7-D that I approve of and felt should be distributed between Armor and Structure; aftering having fought against them and think teamed up with for a match or two in the PTS.

I figure that there's gonna be some tuning needed and more testing, but the increases I'm looking at do have merit to make it to Live, if we do not change base weapon values, leave Heat Capacity where it is, and then continue to see weapon quirks and modules to stack and increase damage output from weapons.

So on the flip side, if we can see changes to base weapon stats and heat capacity, then there is no need to raise armor and structure through an increased base line and too many quirks, (since I still think that certain quirks are warranted due to needy hitboxes).





1 user(s) are reading this topic

0 members, 1 guests, 0 anonymous users