Jump to content

Gentlemen, Let's Face It. (Pinpoint)

Balance

118 replies to this topic

#61 SilentWolff

    Member

  • PipPipPipPipPipPipPipPipPip
  • The God of Death
  • The God of Death
  • 2,174 posts
  • LocationNew Las Vegas

Posted 01 June 2014 - 08:42 AM

Pinpoint damage is going nowhere. People that cry about it, never played MW3 or MW4, both of which had pinpoint damage and were great games. The people that complain about it are mostly bad players who want some kind of crutch to play against players with more skill than them. Quit yer crying and l2p or find another game to play.

#62 General Taskeen

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • 5,737 posts
  • LocationCircinus

Posted 01 June 2014 - 08:46 AM

I always wondered what it would be like if aiming and firing was more 'Battletechy' like this Blender BattleTech game:







The targeting interface of each 'Mech blares when the interface is trying to gain a targeting solution. The round circle shrinks when the computer has the best firing solution, ensuring a hit.

Although that might be a bit hardcore for a MW game, Hardcore Mode?

#63 SirLANsalot

    Member

  • PipPipPipPipPipPipPipPip
  • Legendary Founder
  • Legendary Founder
  • 1,540 posts
  • LocationWashington State

Posted 01 June 2014 - 08:52 AM

View PostFupDup, on 31 May 2014 - 03:41 PM, said:

Fortunately, there are a number of convergence solutions out there that do not use randomized mechanics in any form.


HSR is why the convergence system was changed to what we have now. It really was messing with the server load and the hit detection.....

Any "solution" people come up with will not work because the ones that would fix it, cannot be done either because of hit detection issues, or due to how the game was coded originally.


The only way to go about this is to accept that we do have perfect accuracy (based on how well you the human can do) and balance around that. Instead of trying to change a core thing of the game, something that very well could break the entirety of the game itself, you can try to change things AROUND it. Like upping Armor or Internal HP, to compensate for the fact that we can hit what we aim at (its an amazing concept I know). Slowing RoF of weapons is one solution, but one that will hurt the overall game perception to the players and favor high alpha weapons, more then DPS (DPS based game play is a little more fun then just high alpha alpha alpha alpha).

#64 Charli3 Tang0

    Member

  • PipPip
  • Ace Of Spades
  • Ace Of Spades
  • 35 posts

Posted 01 June 2014 - 09:02 AM

View PostReitrix, on 01 June 2014 - 07:43 AM, said:

While I like the idea of Recoil (It exists, however slight and is purely visual for the shooter) the problem is not the ability to place multiple individual shots in a single location. The issue is placing multiple shots at the same instant to the same pixel. Recoil wouldn't have any appreciable effect on the high FLD meta we have now.


Again, we're in fantasy land, so let’s keep going down this rabbit hole to address the high FLD meta. I do not believe it is feasible to fully remove FLD, so let’s make it painful for the pilot.

Extreme Recoil Impact (ERI): Mechs will suffer internal structural damage from repeated use of multiple recoil generating weapons simultaneously.
  • This would again go back to size of mech, number of recoil generating weapons, location of weapons generating recoil, and number of recoil generating weapons fired simultaneously
  • There will be hard ERI caps for each class of mech (Light, Medium, Heavy, Assault). Could even do it per chassis, but trying to keep this sane for now.
  • Simultaneously firing enough recoil generating weapons to exceed the ERI for that mech class will result in internal structural damage to the gyro (Center Torso) and other random areas of the mech based on location of said weapons.

Edited by Charli3 Tang0, 01 June 2014 - 09:02 AM.


#65 FupDup

    Member

  • PipPipPipPipPipPipPipPipPipPipPipPipPipPip
  • Ace Of Spades
  • Ace Of Spades
  • 26,888 posts
  • LocationThe Keeper of Memes

Posted 01 June 2014 - 09:07 AM

View PostSirLANsalot, on 01 June 2014 - 08:52 AM, said:

HSR is why the convergence system was changed to what we have now. It really was messing with the server load and the hit detection.....

Any "solution" people come up with will not work because the ones that would fix it, cannot be done either because of hit detection issues, or due to how the game was coded originally.

Which is probably why client-side hit detection would probably be something nice to have. No more HSR conflicts, and you no longer get "ghost" damage where you shoot something but the server says "lol he's not really there." We could then have anti-cheat software to deter haxxors.


View PostSirLANsalot, on 01 June 2014 - 08:52 AM, said:

The only way to go about this is to accept that we do have perfect accuracy (based on how well you the human can do) and balance around that. Instead of trying to change a core thing of the game, something that very well could break the entirety of the game itself, you can try to change things AROUND it. Like upping Armor or Internal HP, to compensate for the fact that we can hit what we aim at (its an amazing concept I know). Slowing RoF of weapons is one solution, but one that will hurt the overall game perception to the players and favor high alpha weapons, more then DPS (DPS based game play is a little more fun then just high alpha alpha alpha alpha).

Upping armor/internal HP would achieve higher durability, but ultimately the best way to punch through that durability would still be precision strikes.

Reducing RoF across the board would literally just push this game even more towards poptarting/hillhumping. If staying exposed doesn't let me put additional damage on my enemy, I'm not going to stay exposed. I would gain nothing. I would actually lose something, if anything (because I could be shot during that time). Firing rates only become a problem when they're waaaaaaaaaaay too fast, like when the UAC/5 had a 1.1 second cooldown and 15% jam chance. Then it became a giant anti-mech rapecannon of mass destruction. Maybe a few weapons like [ER]PPCs, the AC/20, and what not could use a slightly longer cooldown, but that's about it.



There are the "burst fire" and "PPC arching" ideas that "work around" convergence (withing actually fixing it), but the afflicted weapons would almost certainly need some kind of compensatory buffs to keep them viable. Otherwise, lasers would just be used because they're dramatically lighter in tonnage and have the same damage spreading.

This just leaves the Gauss Rifle unaffected, which, contrary to popular belief, is still an effective weapon. It only stopped being used as the de-facto ballistic of choice when the charge up got added, so people flocked to ACs for their "snapshotting" needs. Destroying ACs would probably require visiting the Goose Rifle again...

Edited by FupDup, 01 June 2014 - 09:30 AM.


#66 EvilCow

    Member

  • PipPipPipPipPipPipPipPipPip
  • Legendary Founder
  • Legendary Founder
  • 2,243 posts

Posted 01 June 2014 - 09:28 AM

View PostKassatsu, on 31 May 2014 - 01:38 PM, said:

I wouldn't mind seeing dynamic convergence come back, you know, like before HSR where it was possible to miss because your weapons were realigning themselves to your 200 meter target immediately after firing at an enemy 800 meters out. That elite proficiency wouldn't be completely worthless either.


That was not arms realigning, just latency to the server and very unfair for people with higher pings.

#67 Lightfoot

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • Bad Company
  • Bad Company
  • 6,612 posts
  • LocationOlympus Mons

Posted 01 June 2014 - 09:31 AM

View PostPale Jackal, on 01 June 2014 - 06:25 AM, said:


I agree with this. Lasers are already inaccurate enough: as a light, if I've flanked an assault 'mech and am coring into its back, I don't think my lasers need to be non-convergent.

Making Inner Sphere ACs into two-shells per burst, and having PPCs do 66% and 33% of their damage to two components, seems like it would help the problem without requiring any other mechanics to worry about.


See you folks are trying to make the gameplay suit your style. PPCs are not the problem, they are very hot and have a minimum range, any more nerfing would be overkill, so they are not your problem in spite of what you feel.

Ballistics are the top end damage weapons because they have no restrictions or weaknesses. You just bring enough ammo to kill 10-12 mechs and fire as soon as the gun recharges. Every 0.67 seconds to 4 seconds. So if you want a damage nerf, start there with all ACs, not just the ones you don't like. I am thinking a nerf to internal structure damage in the same way Missiles do less armor damage.

But the best way to fix MWO is tougher mechs.

#68 Charli3 Tang0

    Member

  • PipPip
  • Ace Of Spades
  • Ace Of Spades
  • 35 posts

Posted 01 June 2014 - 09:44 AM

View PostLightfoot, on 01 June 2014 - 09:31 AM, said:


...

But the best way to fix MWO is tougher mechs.


Or introduce a new mechanic to make projectiles have a weakness... such as recoil...

Making mechs tougher only makes lasers less desirable.

#69 Turist0AT

    Member

  • PipPipPipPipPipPipPipPip
  • FP Veteran - Beta 1
  • FP Veteran - Beta 1
  • 1,311 posts

Posted 01 June 2014 - 10:02 AM

View Poststjobe, on 01 June 2014 - 02:56 AM, said:




Ya, with this kind of war machines, weapons should hit ware i aim them (Marksman is a robocop in this case). You mention poor fire controll, i dont know what you mean about that. Im not into BT lore (I come from Heavy Gear). Beside this game barelly follows BT, wich is probably a bad thing.

I accept that mechs generate heat, thats fine, i buy that. What i mean is heat penalty in form of ghost heat. Educate me on ghost heat in the BT universe.

And the last point.

I buy what you say about armour being effective. Im talking about weapons not armour.

When i read about BT, and it says this mech is packing a PPC, it sounds like HOLY SH!T PPC of doom and his gona F you up. In this game. Its like: wow you only got one of those? lol, noob. I want to be scared when i scan an enemy and see his loadout. To make me think twice about engaging him and how/if i should engage him. Right now, i allready know what everyone is packing, couple of this, couple of that.

Edited by Turist0AT, 01 June 2014 - 10:13 AM.


#70 Mcgral18

    Member

  • PipPipPipPipPipPipPipPipPipPipPipPip
  • CS 2019 Top 8 Qualifier
  • CS 2019 Top 8 Qualifier
  • 17,987 posts
  • LocationSnow

Posted 01 June 2014 - 10:10 AM

View PostSilentWolff, on 01 June 2014 - 08:42 AM, said:

Pinpoint damage is going nowhere. People that cry about it, never played MW3 or MW4, both of which had pinpoint damage and were great games. The people that complain about it are mostly bad players who want some kind of crutch to play against players with more skill than them. Quit yer crying and l2p or find another game to play.


I still don't understand thus argument...there really isn't a way to make MWOs targeting any simpler.

It doesn't take any "skill" to target a location. If there were separate targeting reticules it would at least rake some skill to pull off the same performance.

It seems to me it's only the bads who don't want to lose their crutch.

#71 DeathlyEyes

    Member

  • PipPipPipPipPipPipPip
  • The Messenger
  • 940 posts
  • Google+: Link
  • Facebook: Link
  • Twitter: Link
  • Twitch: Link
  • LocationMetaphorical Island somewhere in the Pacific

Posted 01 June 2014 - 10:16 AM

Pinpoint damage will always be an issue regardless of whatever is implemented short of anything relying on random chance. Certain mechs, the highlander and dragon slayer in particular have weapons clustered near each other. No solutions seem to target this. The best solution is to reduce their advantage by helping other weapons out. For example reducing laser duration and srm spread would both help combat the advantages of ppc, gauss and autocannons. I hate to burst anyone's bubble but a battletech simulator makes for a bad game based off chance. Instead let the game remain skill based. If people people want the games to last longer further buff armor.

Edited by SLDF DeathlyEyes, 01 June 2014 - 10:17 AM.


#72 Aresye

    Member

  • PipPipPipPipPipPipPipPipPip
  • Heavy Lifter
  • Heavy Lifter
  • 3,462 posts

Posted 01 June 2014 - 10:31 AM

The TTL argument in a nutshell:

Focus fire with teammates is OP. Mechs shouldn't die in 5-10 seconds.


I mean, TTL is great if you're a light pilot. You can rush straight into 2 lances of heavies/assaults, get 3-4 shots off, drop an artillery strike, and run away with barely taking any damage.

Edited by Aresye, 01 June 2014 - 10:34 AM.


#73 Pht

    Member

  • PipPipPipPipPipPipPipPipPip
  • 2,299 posts

Posted 01 June 2014 - 10:35 AM

View PostEddrick, on 31 May 2014 - 02:03 PM, said:


I'm not sure if the idea was origenaly his. But, here is where I heard it from. From the PPC Damage Delivery thread, about PPC Damage Arcing.



Thanks. Precious few people seem to actually have sources! ;)

---

Obviously I can't speak for IC88, but he didn't actually seem to be saying "take aim completely away from players" - or anything that had that meaning. How exactly the backend handles the srms I don't know, but I do know how the TT handles it.

Properly done (removing everything from the tt that represents pilot's gunnery/piloting skills), players would still aim; in fact players would have to use all the skills they use atm, even lag shooting.

Unless, however, "aim" is defined as "directly controlling exactly where all direct fire weapons, which are of the same velocity and fired at the same time, converge." Than, yes. you'd be right. I just don't see any good reasons to define "aim" that way.

View PostTechnoviking, on 31 May 2014 - 02:04 PM, said:


Here is my problem with any of this.

Right now, as a pilot you are trying to pin point fire your shots to a wounded torso/arm leg etc. This takes effort, precision, and concentration, especially if you are moving.


Yes.

Quote

If pin point doesn't matter... what are you doing instead? What video game action are you doing when where your shots can't pinpoint?


Great question. I can only speak for myself; even though I believe that MWO is what it is and likely won't change here (a shame).

What do you do? Well, were the 'mechs performing like they do in the lore - you would be making far more meaningful decisions about piloting and gunnery; and than attempting to skillfully carry them out. In other words, you'd be piloting an armored combat unit that has far more affect on the outcome of combat - affect which the player controls.

The above is ambiguous, so here's some meat on the bones; in the lore, the pilot has to control the tracking at all times (for weapons other than missiles); and does so by controlling the reticule on his main HUD. The pilot also has choose which weapons are shot and has to trigger the chosen weapons. The 'Mech is forbidden by design and programming from doing these things.

As the pilot is doing this, the 'mech's computers are crunching numbers for aimpoints and than sending those aimpoints to each individual weapon (all of which are aimable; the aren't fixed, they're turreted/etc), trying to get each weapon properly converged.

The pilot directly controls tracking, firing, and what is fired. Than there's the 'mech's part of the job - getting the weapons to hit what the pilot is tracking with the reticule, and the pilot controls this too; in indirect but meaningful fashion.

The 'mech's part is what is missing - and all the pilot's skill in putting their 'mech into the right condition to make the shot the pilot is asking of it are missing too. On the opposing side, a lot of the choices and the skill required to make your own 'mech harder to hit while hitting the other guy are out - simply not possible - too.

Quote

People just want it to go away without replacing a skill/intensity/meaning thing to do.


Not everyone.

View Poststjobe said:

Pinpoint damage by itself isn't a huge issue; it becomes a huge issue in combination with instant convergence and front-loaded damage only available to a select few weapon systems (ACs/PPCs/Gauss).


Pinpoint is also a huge issue when combined with discreetely hittable locations, some of which can disable/destroy a target, instead of when it's used versus a simple hitpoints pool; and especially in a combat system that was never meant to have pinpoint in it.

Quote

My preferred solution to this is to make these weapon systems spread their damage just like all the other weapon systems; make ACs burst-fire and PPCs beam-duration weapons (there's plenty of lore backing this implementation, it's simply the way these weapons are supposed to behave).


What would they do with gauss? That's just a big single round going down range.

View PostFupDup, on 31 May 2014 - 03:41 PM, said:

Fortunately, there are a number of convergence solutions out there that do not use randomized mechanics in any form.


Such as?

#74 C E Dwyer

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • Legendary Founder
  • Legendary Founder
  • 9,274 posts
  • LocationHiding in the periphery, from Bounty Hunters

Posted 01 June 2014 - 10:42 AM

I've felt for months now that the only practical way (note practical not best way) to limit the effect of pin point is limit the number of weapons mechs can mount.

The actual number of hard points won't change, just each hard point has a limit to each type of weapon it can carry.

Pin point at range is deadly because of arm lock and locations of weapons the classic example is the Dragon Slayer

2 arm mounted ac5 and right next door 2 energy mounts that can take ppc's

Simply disallow any energy weapon larger than a medium laser in the right torso the left arm can still take ppc or large lasers, and pin point at range and also jump sniping is drastically reduced.

This could be done for every mech across the board making more point to owning more mech variants.

Had this system been used before ghost heat there wouldn't have been any need to introduce it.

If this is used pin point is still there but much reduced, bring back a brawling element and making pulse lasers more desired for pin point at closer ranges, and possibly the most effective way of reducing jump sniping as the most reward for least risk

#75 DeathlyEyes

    Member

  • PipPipPipPipPipPipPip
  • The Messenger
  • 940 posts
  • Google+: Link
  • Facebook: Link
  • Twitter: Link
  • Twitch: Link
  • LocationMetaphorical Island somewhere in the Pacific

Posted 01 June 2014 - 11:02 AM

View PostCathy, on 01 June 2014 - 10:42 AM, said:

I've felt for months now that the only practical way (note practical not best way) to limit the effect of pin point is limit the number of weapons mechs can mount.

The actual number of hard points won't change, just each hard point has a limit to each type of weapon it can carry.

Pin point at range is deadly because of arm lock and locations of weapons the classic example is the Dragon Slayer

2 arm mounted ac5 and right next door 2 energy mounts that can take ppc's

Simply disallow any energy weapon larger than a medium laser in the right torso the left arm can still take ppc or large lasers, and pin point at range and also jump sniping is drastically reduced.

This could be done for every mech across the board making more point to owning more mech variants.

Had this system been used before ghost heat there wouldn't have been any need to introduce it.

If this is used pin point is still there but much reduced, bring back a brawling element and making pulse lasers more desired for pin point at closer ranges, and possibly the most effective way of reducing jump sniping as the most reward for least risk

Not a real solution as people already have content configured in ways that this would conflict with. Paid content might I remind you. Realistically the best solution is to reduce convergence in other mechs so that mechs like the Dragon Slayer no longer have a huge advantage. This increases the viability of other mechs on the battlefield.

#76 C E Dwyer

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • Legendary Founder
  • Legendary Founder
  • 9,274 posts
  • LocationHiding in the periphery, from Bounty Hunters

Posted 01 June 2014 - 11:11 AM

View PostSLDF DeathlyEyes, on 01 June 2014 - 11:02 AM, said:

Not a real solution as people already have content configured in ways that this would conflict with. Paid content might I remind you. Realistically the best solution is to reduce convergence in other mechs so that mechs like the Dragon Slayer no longer have a huge advantage. This increases the viability of other mechs on the battlefield.


Well that's the whole point, the mech content would have to be reconfigured, it would change play styles and make thing work on the whole for the better.

It cures boating of weapons which people complain about, it makes mech more multi weapon platforms which is what mech warrior/battle tech is supposed to be and removes it from the other stompy robot games.

it makes jump sniping far less effective, which most people that use don't like

as for paid content

I own every not just every hero mech but every mech currently available in game, with a bay for each and I'm more than willing to have the mechs load out changed if it will improve play and make it more dimentional.

I remember the bitching about the T-59 being nerfed in World of tanks, (paid content), those nerfs made the game a much better one

#77 ShadowbaneX

    Member

  • PipPipPipPipPipPipPipPip
  • 1,089 posts

Posted 01 June 2014 - 11:27 AM

View Postwolf74, on 31 May 2014 - 02:06 PM, said:

I will Give Player Pin point convergence, but they have to Control the Convergence Point. (AKA you the player must Pick How Far out the weapons cross in front of you. Also to help Keep LRMs from becoming OP, they would Get the Streak Treatment for Location lockon. For LRM Fired with out a Lock a High Arc Flight than Drop at the Set Convergence point, these would aloud LRM to do a Over the Hill shot.


Great idea for X-Wing vs. TIE Fighter, but if you've got an enemy targeted (ie, you've hit 'R') then why wouldn't the on-board weapon systems not be able to calculate the weapons convergence for you?

Personally, I'd like it if there was some play in how the shots landed. I was playing last night against some good (or lucky) gunner that managed to sink 2 separate salvos into my CT at stupidly long range. This dropped my Cataphract's CT armour to nearly nil while the rest of the mech was pristine, despite torso twisting. If there was a bit more play shots could have strayed slightly and some damage would have hit the side torsos instead. Being able to aim should give you an advantage, but you shouldn't have a 30-point alpha all hit the exact same point all the time...there should be some, minor, deviation.

As it is now, it converges on to whatever you're reticule is over. If you're directly on a target all is good and your salvo should land on target. If they're moving off parallel to you though and you have to lead your target, it will converge on the point that you're leading. This means that part of a weapon group can hit while the other one will go wide.

If the system was able to (slowly) adjust it's convergence point to your locked target, it'd mean that you can land leading shots a bit better and make it so that there's a bit more spread in damage with PPCs & ACs, making them more like missile salvos and beam weapons.

#78 stjobe

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • Legendary Founder
  • Legendary Founder
  • 9,498 posts
  • LocationOn your six, chipping away at your rear armour.

Posted 01 June 2014 - 12:08 PM

View PostTurist0AT, on 01 June 2014 - 10:02 AM, said:

Ya, with this kind of war machines, weapons should hit ware i aim them (Marksman is a robocop in this case). You mention poor fire controll, i dont know what you mean about that. Im not into BT lore (I come from Heavy Gear). Beside this game barelly follows BT, wich is probably a bad thing.

Yes, that is a bad thing. BattleTech has 30 years of lore behind it, and as such, certain things are expected to be in certain ways.

For example, the galaxy is on the tail end of roughly 300 years of rather harsh interstellar war that has made a lot of high-tech weaponry and control systems lost tech (or "lostech", as it is known). The very 'mechs that we pilot can only be made in automated factories - the knowledge on how to construct them has been lost. Same with their components, which is where the old BattleTech saying "kill the meat, save the metal" comes from. A highly trained MechWarrior is a valuable asset for sure, but pales to insignificance in comparison to his or her 'mech.

'Mechs are also rare; it's common that they're several hundred years old, patched and cobbled together and handed down from parent to oldest child as the family heirloom. Owning a 'mech automatically sets you apart and above from the common people of the Inner Sphere. Not exactly nobility, but a Mechwarrior is indeed privileged.

With a lot of high-tech stuff being lostech, the battlefield in BattleTech is rather more akin to medieval knights or WWI air combat than 20th century mechanized warfare; 'mechs are so valuable one side might decide to forfeit and give up instead of risking their 'mechs. And when fighting does occur, it's usually to first blood, not to the death (and certainly not to the complete destruction of the 'mechs!). Exceptions exist, of course, but the general rule is indeed to "spare the metal" - if nothing else, they can be used for spare parts (which are always needed) after their pilots are ransomed back to the enemy.

As for perfect aim; according to lore, the battlefield is inundated with all kinds of counter- and counter-countermeasures to the point where the antiquated lostech sensor and guidance system has to struggle to keep up; this is the in-universe explanation to both the random to-hit roll and the short ranges that a board game designed to be played on a kitchen table necessitated.

So no, a BattleTech game with perfect accuracy is a contradiction in terms; your patched-up, 200-plus-year-old, replaired-counless-times-with-parts-from-different-systems 'mech just doesn't provide enough accuracy to put two bullets through the bullseye. That's simply not how BattleTech works.

View PostTurist0AT, on 01 June 2014 - 10:02 AM, said:

I accept that mechs generate heat, thats fine, i buy that. What i mean is heat penalty in form of ghost heat. Educate me on ghost heat in the BT universe.

There's no Ghost Heat in the BattleTech Universe, but there is rather severe heat penalties for running hot. Translated to MWO terms, those penalties start somewhere between 30-50% heat, and you'd be lucky to get to 100% heat without shutting down, being slowed to a stop, having absolutely no chance of hitting anything, or just plain exploding.

None of those things exist in MWO, and it's one of the defining aspects of BattleTech; that you struggle as much against your own heat as against the enemy. In MWO, that aspect is simply missing.

Hence why some of us think it's rather hilarious (as well as sad) that they just added "a BattleTech game" to the MWO logo; it's missing so many simply key elements from the BattleTech universe that adding that phrase now feels like some sick joke.

View PostTurist0AT, on 01 June 2014 - 10:02 AM, said:

I buy what you say about armour being effective. Im talking about weapons not armour.

When i read about BT, and it says this mech is packing a PPC, it sounds like HOLY SH!T PPC of doom and his gona F you up. In this game. Its like: wow you only got one of those? lol, noob. I want to be scared when i scan an enemy and see his loadout. To make me think twice about engaging him and how/if i should engage him. Right now, i allready know what everyone is packing, couple of this, couple of that.

Thank PGI for that. Because they chose (or simply were not competent enough to do otherwise) to implement perfect accuracy, instant convergence, and front-loaded damage, they also broke the whole armour system; the BattleTech armour system is designed for spread-out damage (in TT with random hit location rolls) and it simply breaks when you can duct-tape several weapon systems together and have them all hit the same spot.

They tried to fix it by increasing IS and Armour values, but that's only treating the symptoms. They need to remove at least instant convergence and front-loaded damage (and probably perfect accuracy as well) if they want the armour system to work.

If they do, they can go back to TT values and a single AC/20 will again be a scary thing indeed.

Edited by stjobe, 01 June 2014 - 01:43 PM.


#79 Lykaon

    Member

  • PipPipPipPipPipPipPipPipPip
  • 3,815 posts

Posted 01 June 2014 - 12:40 PM

View PostcSand, on 31 May 2014 - 01:25 PM, said:


Seriously though

Pinpoint is not that big a deal. Keep moving, use your torso twist and play smarter


And this helps how?

Torso twist? doesn't matter with front loaded damage the entire volume of damage is applied in a singular instant.You are hit then you twist to do what? you were already hit with the full volume of damage.

Or did you mean twist first? like before the enemy shot? yeah like you can protect your vitals that way when all the enemy are using the same FLD groupfire alpha tactic turned away from one may as well be facing another....useless tactic when everyone is doing the FLD alpha bandwagon.Not like I can't wait until you must face to fire to unload on you so twisting is a farce.

Also.twisting is a tactic that any mech under 70 tons can't expect to see having any effect.When the instantly applied damage volume is genneraly also equal to your total armor values on most body locations twisting only means you get first choice of what body part is blown off first.
So best case scenario I may lose a left arm before my main weapon or engine is cored if I'm piloting anything under 60 tons.

Play smart? like huddle in cover the entire match? or did you mean poptarting? being first to engage is first to die when the majority of a pug force is huddling behind a rock waiting to get arty strikes dropped on them.

Playing smart is perpetuation of an extremely narrow meta that has become for many painfully dull and predictable at best.

At it's worst the narrow meta is a barrier to new users it doesn't say anywhere in any tutorial or weapon description that to "play smart" is to only use a tiny portion of the mechs and weapons in a specificly perscribed manner.

Pin point or as I call it front loaded pinpoint is the single most damaging mechanics failure MWo has currently.If any one thing were to be corrected it must be the FLD pinpoint issue

This issue is at the core of so many issues that fixing it will vastly improve the state of the game on a whole.

Match maker balanced on Elo? worthless waste of time if the team is balanced on skill then obviously the team with more meta will win so meta is the deciding factor the team with more FLD pinpoint wins.Match maker is still fail unless this FLD is fixed.

Weapon Balance? you can not balance weapons as singular systems without first balancing damage application mechanics.A laser will never stack up to an AC as long as one has the superior damage application method.

Role warfare? No point to trying to accomplish this as long as the prevailing meta applies.The meta will dictate what weapons and chassis are used most and the role is always be the mech with the FLD weapons.

Seriously look at the common threads on these forums and focus on the cause it is almost always at it's roots a FLD related issue.

Is X weapon over powered or under powered?
Do mechs die to easy?
Jumpjets/poptarting?
Chassis balancing?
Teams vs Pugs?
Complaints about mechanics implimented to control the FLD pinpoint issue (weapon "balancing" Ghost Heat Etc.)

Easily more than half of the "issues" are because of FLD pinpoint it needs to go!

#80 Radiant Mass

    Member

  • PipPip
  • Big Brother
  • 29 posts

Posted 01 June 2014 - 12:45 PM

This proposal covers more than the Pinpoint and FLD problem.

It is quite extensive. Just a head up to those who don't like to read a whole lot.





34 user(s) are reading this topic

0 members, 34 guests, 0 anonymous users