43 replies to this topic

[Livewyr]

As I posted in another thread...which was discussing this same thing:

They *can* code non-instant convergence.. but here is the problem: (As I understand it from CBT)

Game is server authoritative in everything to reduce hacking. (Everything you do, the server authorizes, from moving, to aiming, to shooting.)

The calculations of constantly making minor adjustments to convergence rate provided too much for the server.
What we had:
Every time your crosshair changed range (essentially every time you move it) a new convergence point is created.
From that point weapon convergence would move (at a fixed rate, increased by the convergence efficiency) from the previous crosshair range, to the new one.
This convergence would be calculated for every single possible point of convergence between the old range, and the new range.

Essentially meaning: Every time you move, anything, the server would be doing countless calculations on your convergence: not just where your range was, and is now, but every range between the two.

This x 16 people at the time x every game on the server.

This caused 2 big problems:
1: Bad server lag issues. (Rubber banding, teleporting, and such)
2: Weapons never quite converging regardless of how long they were on target, for whatever reason. (PPCs were particularly notorious.)

-----------------
So they moved to: Where ever your range is, is where your convergence is. A tiny fraction of the calculations and server load.


This is my understanding of the devs' explanation for removing trained convergence back in CBT. (If a Dev would like to poke through with a more thorough/accurate explanation, that would be great.)

--------------------------
Adding for this thread:
What are you going to propose that achieves a convergence calculation without eating servers alive? (And no, "Targeting" will not work. ECM)

No other game that I know of has a convergence setup, it is either pinpoint, or standard CoF. (And pinpoint is usually done as a constant, or achieved briefly by a hold breath type mechanic such as newer games' snipers.)

This is because most games do not have:
More than 2 weapons to be concerned with.
A range consideration on their weapons.
Different sized targets moving at different speeds. (An FPS equivalent of a light mech.)
[EDIT] Oh, and individual armor amounts with the ability to be killed while still having full armor on all but one spot.

They can simply do a standard CoF for each weapon that increases a standard way while moving, decreases a standard way while aiming, or increases in a standard way with successive automatic fire.

The only way I can see a cone of fire being implemented, is with Autocannons, and making them all burst (like clans).
This will of course ensure that PPC and Gauss reign supreme with no contest.

So what do you do with a weapon system that is hyper accurate at long range? The same they other shooter games do with their long-range hyper accurate weapons (Sniper and Marksman Rifles): You slow down their rate of fire relative to other weapons.

I have not played a game with a fully automatic sniper or marksman rifle. (Though someone else mentioned something like that in BF:BC- and it being completely overpowered.) I do not see why we need long range accurate weapons with the same RoF as short range weapons.

Edited by Livewyr, 18 July 2014 - 08:53 AM.

[TLBFestus]

How about simply removing the arm lock mechanic from the game after your first 25 matches are done?

It won't fix it all, but it would split a fraction of the damage at times, so it would be a start.

[Aresye]

TOGSolid, on 18 July 2014 - 01:29 AM, said:

I'm still baffled as to why people think aiming in this game takes a lot of skill. Then again, I did go from playing a frontline rifle Recon in BF3 to playing this so maybe my perspective is just skewed.

If aiming doesn't take much skill, then why are there so many players that can't hit the broadside of a barn? And yes, I understand that watching other people's screens isn't accurate, but you can tell when a player sucks. The cursor may not line up with what you see, but it isn't hard to tell a player is overcorrecting and missing the majority of their shots.

Likewise, you can usually spot a good player based on how smooth and controlled their crosshair is, coupled with seeing consistent damage done to the enemy mech along with the crosshair hit indicator.

John80sk, on 18 July 2014 - 02:27 AM, said:

I keep hearing players say this, and it's really starting to grate on me. You do realize that any actual weapon system does not strike exactly point of aim point of impact right? This is one of the things that makes marksmanship require MORE skill, not less, as you have to be MORE precise in order to hit your intended target.

Some of the suggestions for fixing pinpoint:

Manual Convergence: Higher skill required. Better players will still continue to dominate, and newer players will have an even harder time learning the game.

Off-center Weapons: Higher skill required for longer ranges. Better players will dominate from long to medium range, and newer players won't even have a chance until they can close to shorter ranges.

Cone of Fire: Lower skill required. Better players will miss shots that should be hits, and newer players will make shots that shouldn't be hits.

I'll believe people that say aiming doesn't take skill when I can actually see people aim well without either standing still, or being part of a high Elo competitive unit.

Blood Rose, on 18 July 2014 - 02:52 AM, said:

Umm... Convergence requires MORE skill, not less. And it would also bring back a use for that redundundant 4K XP sink Convergence skill.

Like I said above, convergence requires more skill, and thus will not curb the majority of better players. What it would do however, is make the game even more frustrating for newer players, and those that can't aim very well to begin with.



Fact: Everybody in the game has the exact same aiming system and access to pinpoint weapons.

Pinpoint convergence isn't a problem, because Joe Newb over there has the exact same loadout as me and has every capability that I have. Why then does Joe Newb not prevail and present himself as my equal? Simply put, I can just aim better.

People need to wake up and face the fact that there are players who are downright better than them. Pinpoint isn't some magical advantage that high Elo players have, and low Elo players don't.

[Fut]

Aresye, on 18 July 2014 - 11:53 AM, said:

People need to wake up and face the fact that there are players who are downright better than them.

The entire world would benefit from people figuring this out.

[John80sk]

Aresye, on 18 July 2014 - 11:53 AM, said:

Some of the suggestions for fixing pinpoint:

Manual Convergence: Higher skill required. Better players will still continue to dominate, and newer players will have an even harder time learning the game.

Off-center Weapons: Higher skill required for longer ranges. Better players will dominate from long to medium range, and newer players won't even have a chance until they can close to shorter ranges.

Cone of Fire: Lower skill required. Better players will miss shots that should be hits, and newer players will make shots that shouldn't be hits.

Convergence just won't work, I get that.

Cone of fire however is a perfectly viable option. I'm not talking about a huge spread, but lets say PPC's and Autocannons had a spread roughly the size of a Jenner's CT at 500m. You could still hit a moving Jenner at 500m, you just have to be significantly better at the game to do it.

This is akin to firing an actual bolt gun at 500m on a head sized target. The gun itself will not allow you to hold a grouping better than 5 inches at this range. Meaning you have to be all the more precise to fire on an ~8 inch target. How many of your shots can you honestly say are dead center of a targeted component?

I don't particularly care whether pinpoint stays or not. However, I don't like bad arguments.

[Hellcat420]

TheRealAbray, on 17 July 2014 - 09:21 PM, said:

So, a lot of people agree that there is not a good way to implement an altered convergence system that is fair for a game that is based (at least slightly) on aiming skill.

So instead of changing the max damage that you can do, how about changing the max damage you can TAKE?

So if your right torso takes a 50 or whatever bullshit alpha from a Direwhale, instead of the right torso taking ALL the damage...

There could be a mechanic that limits the max damage any one component can take within a 0.1 second (or something similar) time period? Basically force the poinpoint alphas to splash some damage if they are too large. It sorta mimics the random chance that TT has for what component you hit.

Don't hate the player, hate the game.

that is a bad idea. the way to fix it is to get rid of the instant pinpoint convergence. a lot of people will cry because it will actually take aiming skill when you have to hold your crosshair on the target for a few seconds for your mech to aim its weapons and adjust leadtime. the game as it is requires minimal aiming skill.

Edited by Hellcat420, 18 July 2014 - 12:49 PM.

[Joseph Mallan]

A widening cone of fire when firing more than one weapon at a time. More weapons wider cone. And I am a Alpha striker by choice.

[Lightfoot]

In MechWarrior 4 I had to fire two Light Gauss (12 damage each) and 3x CERPPCs for a total 70 point alpha to take the CT armor of a heavy mech to red-orange, so I always feel MWO mechs are too weak by a patch-worthy amount. I would rather the mechs get more armor than the weapons be gutted with nerfs. Tougher mechs would still play like Battletech. Ghost heat and the Gauss Rifle nerf change MWO into something very different from Battletech.

[Joseph Mallan]

Lightfoot, on 18 July 2014 - 12:59 PM, said:

In MechWarrior 4 I had to fire two Light Gauss (12 damage each) and 3x CERPPCs for a total 70 point alpha to take the CT armor of a heavy mech to red-orange, so I always feel MWO mechs are too weak by a patch-worthy amount. I would rather the mechs get more armor than the weapons be gutted with nerfs. Tougher mechs would still play like Battletech. Ghost heat and the Gauss Rifle nerf change MWO into something very different from Battletech.

My MW:O Atlas has 100(ish) armor on the Front CT... Though you'd hurt me, I can take 70... once.

[Aresye]

John80sk, on 18 July 2014 - 12:39 PM, said:

Convergence just won't work, I get that.

Cone of fire however is a perfectly viable option. I'm not talking about a huge spread, but lets say PPC's and Autocannons had a spread roughly the size of a Jenner's CT at 500m. You could still hit a moving Jenner at 500m, you just have to be significantly better at the game to do it.

This is akin to firing an actual bolt gun at 500m on a head sized target. The gun itself will not allow you to hold a grouping better than 5 inches at this range. Meaning you have to be all the more precise to fire on an ~8 inch target. How many of your shots can you honestly say are dead center of a targeted component?

I don't particularly care whether pinpoint stays or not. However, I don't like bad arguments.

Bad arguments eh? Well, first I'll preface this by saying I'm an active duty helicopter crewchief, qualified instructor for both .50cal and 7.62 crew served weapons, and qualified level III tactical crewman in all mission areas. I've spent the last 6 years studying (and using) electrooptics, aerial ballistics, radar/IR theory, weaponeering, RF propogation, and more.

What you are implying with aimpoint errors in comparison to your bolt rifle is an example of interior ballistics, or in other words, "Factors that influence the trajectory of a projectile prior to exiting the barrel that cannot be compensated for by the gunner." Such factors include barrel twist rate, barrel erosion, projectile weight, counterclockwise vs clockwise rotation, barrel length, and weapons mount stability.

Implementing a cone of fire for ACs, machine guns, and LBX makes sense, as no matter what kinds of stabilizing systems are used on mechs, they cannot fix internal ballistic factors. The gauss is kind of special as the projectile and speed negate the majority of ballistic factors at common engagement ranges.

PPCs and lasers on the other hand, are directed energy weapons. They do not have internal factors and aren't subjected to external factors like gravity or wind, but they are influenced by external factors such as humidity, reflection, refraction, and atmospheric propogation. For all intents and purposes however, the ranges in BT are too short to be influenced by these factors to any noticeable degree.

Directed energy weapons are commonly boresighted to perfectly match the crosshair in which a pilot uses to aim, often incorporating some form of gimbal that automatically slews the weapon/s for perfect convergence. Today we already have the technology to hold a laser focused on a single point smaller than 1m from MILES away, on a moving, vibrating, and unstable platform.

So if you're going to throw the realism argument out there, basically it would make sense for ballistic weapons to have convergence errors due to internal ballistics, but PPCs and lasers would still be pinpoint.

Even if we incorporated a cone of fire at longer ranges for all weapons, how would that increase the skill required to hit a light at range? You can't compensate for RNG mechanics. They're unpredictable by their very definition. A good pilot cannot predict where their shot will go at longer ranges because it's completely random. They could have the perfect lead with PPCs, but the RNG makes it so they miss. On the other hand, a pilot could have an incorrect lead, but the RNG makes it so they hit. Thus, good players who know how to lead their shots well will end up getting screwed, while bad players who can't aim will occasionally be rewarded, which is a textbook definition of lowering skill cap.

[TOGSolid]

Honestly, just fixing the heat system would do a lot to fix this game. If everyone was capped at 30 heat a lot of problem builds would just vanish overnight.

[John80sk]

Aresye, on 18 July 2014 - 01:43 PM, said:

Bad arguments eh? Well, first I'll preface this by saying I'm an active duty helicopter crewchief, qualified instructor for both .50cal and 7.62 crew served weapons, and qualified level III tactical crewman in all mission areas. I've spent the last 6 years studying (and using) electrooptics, aerial ballistics, radar/IR theory, weaponeering, RF propogation, and more.

What you are implying with aimpoint errors in comparison to your bolt rifle is an example of interior ballistics, or in other words, "Factors that influence the trajectory of a projectile prior to exiting the barrel that cannot be compensated for by the gunner." Such factors include barrel twist rate, barrel erosion, projectile weight, counterclockwise vs clockwise rotation, barrel length, and weapons mount stability.

Implementing a cone of fire for ACs, machine guns, and LBX makes sense, as no matter what kinds of stabilizing systems are used on mechs, they cannot fix internal ballistic factors. The gauss is kind of special as the projectile and speed negate the majority of ballistic factors at common engagement ranges.

PPCs and lasers on the other hand, are directed energy weapons. They do not have internal factors and aren't subjected to external factors like gravity or wind, but they are influenced by external factors such as humidity, reflection, refraction, and atmospheric propogation. For all intents and purposes however, the ranges in BT are too short to be influenced by these factors to any noticeable degree.

Directed energy weapons are commonly boresighted to perfectly match the crosshair in which a pilot uses to aim, often incorporating some form of gimbal that automatically slews the weapon/s for perfect convergence. Today we already have the technology to hold a laser focused on a single point smaller than 1m from MILES away, on a moving, vibrating, and unstable platform.

So if you're going to throw the realism argument out there, basically it would make sense for ballistic weapons to have convergence errors due to internal ballistics, but PPCs and lasers would still be pinpoint.

Even if we incorporated a cone of fire at longer ranges for all weapons, how would that increase the skill required to hit a light at range? You can't compensate for RNG mechanics. They're unpredictable by their very definition. A good pilot cannot predict where their shot will go at longer ranges because it's completely random. They could have the perfect lead with PPCs, but the RNG makes it so they miss. On the other hand, a pilot could have an incorrect lead, but the RNG makes it so they hit. Thus, good players who know how to lead their shots well will end up getting screwed, while bad players who can't aim will occasionally be rewarded, which is a textbook definition of lowering skill cap.

What is it with people pulling their wangs out around here for no apparent reason? Yet another of my pet peeves. Careful who you talk down to though, you might end up looking foolish.

I wasn't making an argument for what's realistic or not in our little world of giant robots. My point was saying a cone of fire does not detract from the amount of skill involved, just as an actual rifle's imperfect grouping does not detract from the skill involved in using said weapon system. Also, I'm pretty sure we're discussing PPFLD weapons here, not lasers.

A smart shooter knows the limitations of his weapon systems, and compensates. Sure RNG wise a bad player might make a shot he otherwise would have missed, but the likelihood of him hitting his intended target is actually lower. Currently that "bad pilot" has a very wide margin of error, with CoF that margin of error is lowered.

A smart pilot on the other hand would understand that at certain ranges he is unable to hit a specific component and fire center mass, or not aim for a small component that he may not be able to hit.

An exceptional pilot can fire his weapon at its max effective range, and strike the targeted component despite the possibility that if he's more than a ********* off he could miss. This requires both the presence of mind to know what that range is, and the marksmanship to make it happen. This is a thinking man's shooter no? This may reduce the importance of a single skill (twitch shooting), but only because it expands the skills necessary to be successful.

Welp, think that's enough devil's advocate for tonight.

[Nanoswimz]

There is no more 6ppc stalkers, so crying about high alpstrikes is a bad point.

[TheKGB]

Convergence has always been the real answer to the question in my mind.

The issue is pinpoint damage, the fact that we can cause all of our damage to one location.

I think it's as simple as applying a few degrees of inaccuracy dependent on how fast the mech is moving/falling. The variance doesn't need to be huge, just enough to spread damage out among the torso sections a little more equally. For those worried about "skill" (which as stated before, convergence actually adds more skill in the form of throttle management and counterplay in the fact that you must remain stationary to guarantee your pinpoint alpha.) The area within which certain weapons spread could be in relation to their placement on the body. In this way, you could guess that a weapon on your left arm would stray more left within the your cone of fire, and you could lead accordingly.

This type of system has been proved by counterstrike and others, and really solves a big part of the gauss/ppc pinpoint alpha problem that makes mediums and lights so akin to paper and often dissatisfying to play. Without instant convergence, we might even be able to do away with the silly ghost heat system, and increase the TTK of many mechs. Not only that, but with a varied convergence based on movement/falling speed, we can create more interesting piloting decisions and open up various opportunities for counterplay.

After that, LRM's could use a few changes, and the heat system needs to have a few more effects (maybe high heat effects your convergence? who knows!) but i really do think adding a little inaccuracy not only increases the plausibility and realism of the game, but works well for gameplay.

Edited by TheKGB, 19 July 2014 - 05:20 AM.

[Jolly Llama]

TheRealAbray, on 17 July 2014 - 09:21 PM, said:

So, a lot of people agree that there is not a good way to implement an altered convergence system that is fair for a game that is based (at least slightly) on aiming skill.

So instead of changing the max damage that you can do, how about changing the max damage you can TAKE?

So if your right torso takes a 50 or whatever bullshit alpha from a Direwhale, instead of the right torso taking ALL the damage...

There could be a mechanic that limits the max damage any one component can take within a 0.1 second (or something similar) time period? Basically force the poinpoint alphas to splash some damage if they are too large. It sorta mimics the random chance that TT has for what component you hit.

Don't hate the player, hate the game.

You will never stop high alpha damage. I have seen sooooo many nerfs to try and stop this. It is impossible. Players will always try to do the most damage in the shortest time possible.

[3rdworld]

Are lasers converging a problem?

[Lykaon]

TheRealAbray, on 17 July 2014 - 09:21 PM, said:

So, a lot of people agree that there is not a good way to implement an altered convergence system that is fair for a game that is based (at least slightly) on aiming skill.

I have a potential solution and it retains aiming skill.

First let's look at the "problem" at it's basic level.

To much damage being placed in a tight grouping in a very brief period of time.

It would seem initially to be an issue exclusivley with convergence but,in fact this issue is linked directly to other mechanics as well as convergence.Since altering convergence mechanics appears to be off the table (due to potential complexity in execution) We should examine the other mechanics that contribute heavily to pinpoint front loaded high volume damage.

And those mechanics are?

Weapons that use a pin point front loaded damage application method. PPC/ER-PPC Inner Sphere ACs and gauss.We have already seen this mechanic partially addressed with the release of clantech.The Clan ACs do not use a pinpoint front loaded damage mechanic and the clan ER-PPC uses splash damage mechanics to reduce the pinpoint.

And,more importantly Group Fire mechanics that allow a single weapon group to be fired instantly and apply 30+ pinpoint FLD damage.

A slight alteration of the Group Fire mechanics can assist in diffusing high concentrations of damage in brief time periods.

Mainly this alteration is a sort of enforced chainfire mechanic but is altered to be functional with laser weapons.

A group of weapons when fired will have a brief delay between the first weapon and the next in succession.

Currently lasers when chainfired start to discharge only after the previous laser weapon has completed it's burn duration.This would mean that a group of 6 medium lasers would take 5 seconds to fully discharge.

Laser one begins it's one second discharge at second 0
Laser two begins it's discharge only after laser one is finished on second 1
Laser two begins it's discharge only after laser two is finished on second 2
ETC and so forth with each laser in turn only begining it's one second burn after the previous has finished it's one second burn.

Now my suggestion is to have a very brief delay in the discharge of a second and subsequent weapons contained in a single weapon group.Let's say 0.2 seconds

Now the group of 6 medium lasers will fully discharge in 2 seconds.

Laser one discharges at second 0.0 and burns for 1 second
Laser two discharges at second 0.2 and burns for 1 second
Laser three discharges at second 0.4 and burns for one second
Laser three discharges at second 0.6 and burns for one second
ETC and so on until all six lasers have discharged and completed their burn cycles finishing in 2 seconds.

Idealy this groupfire alteration would be accompanied with alterations to Autocannons altering the fire patterns to burst fire.

Gauss rifles need some special considerations.

Here are some options

Remove the charge mechanics and have gauss rifles fire three "shots" in rapid succession with each doing 5 damage.Let's say we had a 0.1 second delay between the "shots".This would have a gauss discharging in slightly under one third of a second applying 15 damage to a target and if patiently aimed apply most if not all to a specific target area.However,because of the "burst" mechanic poptarting gauss snipers would spread damage slightly when firing upon moving targets.The combined motion of the shooter and target will disperse the damage avoiding easy concentrations of damage.
It may also be a need to limit weapon groups to include no more than a single gauss rifle to avoid twin gauss mechs from becoming rapid fire triple AC10 mechs mechanicly speaking.

Option two is less intensive and requires a weapon group to include no more than a single gauss rifle and retains the charging mechanics.This would leave a 2 PPC 1 Gauss poptart sniper with a 0.4 second discharge period (from the group fire alteration) that also includes a gauss charging cycle.If someone is good enough to place all 35 damage into your CT with these hoops to jump through they deserve to be allowed to do it.

[Dirkdaring]

TOGSolid, on 18 July 2014 - 01:29 AM, said:

I'm still baffled as to why people think aiming in this game takes a lot of skill. Then again, I did go from playing a frontline rifle Recon in BF3 to playing this so maybe my perspective is just skewed.

Ok hotshot, post a video of you playing and lets see your amazing aiming.

[TOGSolid]

Quote

Remove the charge mechanics and have gauss rifles fire three "shots" in rapid succession with each doing 5 damage

That would feel really weird. Just slowing the fire rate of it would do a lot to keep it in the realm of being only really useful at long range.

Edited by TOGSolid, 19 July 2014 - 04:09 PM.

[Yokaiko]

TOGSolid, on 19 July 2014 - 04:04 PM, said:

That would feel really weird. Just slowing the fire rate of it would do a lot to keep it in the realm of being only really useful at long range.

Sounds alot like a clan AC20