Jump to content

Convergence Issue Epiphany...


43 replies to this topic

#21 Khobai

    Member

  • PipPipPipPipPipPipPipPipPipPipPipPipPip
  • Elite Founder
  • Elite Founder
  • 23,969 posts

Posted 14 August 2013 - 09:02 AM

Quote

MWO isnt using battetech armor values. They DOUBLED the armor long ago in closed beta and will probably do the same to internal structure.


MWO is using battletech armor values. They simply doubled them. But doubling them isnt nearly enough to deal with precise aiming or convergence.

In TT a center torso gets hit 20% of the time. In MWO a center torso can get hit 100% of the time. That is a fivefold increase in damage. Yet only a twofold increase in armor.

#22 DaZur

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • Elite Founder
  • Elite Founder
  • 7,511 posts
  • LocationWisconsin

Posted 14 August 2013 - 09:03 AM

View PostDamocles, on 14 August 2013 - 08:59 AM, said:

Is this a typo? Because this would be a huge buff to the AC/20 as its current MAX DMG range is 270m, not 540m.

No typo... max range is 810. Max damage range is 270 as you noted. None of this changes damage output, only effective windows of deviation.

#23 Damocles

    Member

  • PipPipPipPipPipPipPipPip
  • The Hammer
  • The Hammer
  • 1,527 posts
  • LocationOakland, CA

Posted 14 August 2013 - 09:05 AM

View PostDaZur, on 14 August 2013 - 09:03 AM, said:

No typo... max range is 810. Max damage range is 270 as you noted. None of this changes damage output, only effective windows of deviation.

So the AC/20 can only hit max damage under 270m, and with a %accuracy deviation. edit: its a +% accuracy. i see what i was misreading
ugh wtf I need to sleep.

Edited by Damocles, 14 August 2013 - 09:09 AM.


#24 Zultor

    Member

  • PipPipPipPipPip
  • Survivor
  • Survivor
  • 171 posts
  • LocationMinneapolis, MN

Posted 14 August 2013 - 09:07 AM

There should be no 0% deviation no matter what range (or if you are in optimal range) except, maybe, if both the attacker and target are stationary. Linked weapons fired together should still have their hit location "rolled" (for the lack of a better term) separately and each "roll" should have a divergence coefficient.

This would solve so much of the problems where Alpha damage rules the battlefield and make multiple PPC/Guass builds less lethal.

#25 Syllogy

    Member

  • PipPipPipPipPipPipPipPipPip
  • 2,698 posts
  • LocationStrana Mechty

Posted 14 August 2013 - 09:08 AM

Since this is already on my clipboard:

Posted Image


Solution: Move.

#26 DaZur

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • Elite Founder
  • Elite Founder
  • 7,511 posts
  • LocationWisconsin

Posted 14 August 2013 - 09:14 AM

View PostDamocles, on 14 August 2013 - 09:05 AM, said:

So the AC/20 can only hit max damage under 270m, and with a %accuracy deviation. edit: its a +% accuracy. i see what i was misreading
ugh wtf I need to sleep.

LOL! Usually your pretty sentient... wasn't sure where you were going. :)

#27 3rdworld

    Member

  • PipPipPipPipPipPipPipPipPip
  • 3,562 posts

Posted 14 August 2013 - 09:19 AM

View PostKhobai, on 14 August 2013 - 09:02 AM, said:


MWO is using battletech armor values. They simply doubled them. But doubling them isnt nearly enough to deal with precise aiming or convergence.

In TT a center torso gets hit 20% of the time. In MWO a center torso can get hit 100% of the time. That is a fivefold increase in damage. Yet only a twofold increase in armor.


Talk about cherry picking. In TT a CT could get hit 100% of the time, and a player could miss the CT all but 20% of the time.

Don't pretend getting hit only in the CT is common or even expected. How many times have you died with only CT damage?

Generally when I die, I am pretty freaking beat up. So much that I would wager the hit % are not that different from TT to MWO.

Edited by 3rdworld, 14 August 2013 - 09:20 AM.


#28 Dimento Graven

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • Guillotine
  • Guillotine
  • 6,208 posts

Posted 14 August 2013 - 09:20 AM

Here's a thought, maybe someone mentioned this earlier, I've seen alluded to anyway:

REMOVE THE "BEYOND" MAX RANGE DAMAGE CAPABILITY.

Seriously. Why are ERPPC/Gauss such an effective damage combo? Because the stupid, the slow, the unaware, the hard headed can be killed at 1620 meters due to the "max range" mechanic that MWO has implemented.

In the ERPPC/Gauss the maximum range that THAT particular combo would normally have is 660meters, the LONG RANGE of the gauss (ERPPCs having a Long Range of 810).

In TT Long Range was THE maximum range. Any 'mech beyond the LONG RANGE, was not even targetable.

So, eliminate the x2/x3 MAX range BS.

Problem solved, until the stupid, the slow, the unaware, and the hard headed still end up getting killed because of stupid, slow, unaware, hard headed mistakes that they will CONTINUE to make, and would CONTINUE to make even if you lowered the damage to 1 point every 10 seconds...

#29 Joseph Mallan

    ForumWarrior

  • PipPipPipPipPipPipPipPipPipPipPipPipPipPipPip
  • FP Veteran - Beta 1
  • FP Veteran - Beta 1
  • 35,216 posts
  • Google+: Link
  • Facebook: Link
  • LocationMallanhold, Furillo

Posted 14 August 2013 - 09:23 AM

View PostUnbound Inferno, on 14 August 2013 - 08:20 AM, said:

I agree.

If you swap that +3% to a -3% then you'd get the result you are looking for.

Ok great minds think alike. I was thinking the same thing, Best accuracy up close, getting worse as you get further out.

#30 Khobai

    Member

  • PipPipPipPipPipPipPipPipPipPipPipPipPip
  • Elite Founder
  • Elite Founder
  • 23,969 posts

Posted 14 August 2013 - 09:24 AM

Quote

Talk about cherry picking. In TT a CT could get hit 100% of the time, and a player could miss the CT all but 20% of the time.


Yes. But statistical average is that the CT gets hit 20% of the time. The armor values in TT are weighted based on statistical average. Streaks of luck happen, sure, but theyre irrelevant to the discussion, since FASA did not determine armor values based on streaks of luck.

But those TT armor values have absolutely no place in MWO because we arnt rolling dice. We choose exactly where we want to hit and hit locations are grossly underarmored because of it.

MWO needs its own armor values. PGI should datamine the game and find out how often each hit location actually gets hit then redistribute max armor based on those findings. The result is that center and side torsos would get armored up while arms and legs would become much more vulnerable (although light mechs would have proportionally more leg armor since their legs are actively targeted).

Quote

Ok great minds think alike. I was thinking the same thing, Best accuracy up close, getting worse as you get further out.


That is generally how ALL cone of fire systems work. Accuracy decreases inversely proportional to range.

Edited by Khobai, 14 August 2013 - 09:37 AM.


#31 DaZur

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • Elite Founder
  • Elite Founder
  • 7,511 posts
  • LocationWisconsin

Posted 14 August 2013 - 09:35 AM

View PostJoseph Mallan, on 14 August 2013 - 09:23 AM, said:

Ok great minds think alike. I was thinking the same thing, Best accuracy up close, getting worse as you get further out.

For most mechs unless they're provisionally kitted for ranged support where their minimum range is fairly large, the ability to hit opposing mechs inside the minimum and to large extent the optimum windows is literally a given.

You're effectively already inside their "personal space"... :lol:

increasing accuracy inside this range is the equivalent of standing at the waters edge and taking a step out into the water to make sure you don't miss the water when you throw a rock. :)

#32 Joseph Mallan

    ForumWarrior

  • PipPipPipPipPipPipPipPipPipPipPipPipPipPipPip
  • FP Veteran - Beta 1
  • FP Veteran - Beta 1
  • 35,216 posts
  • Google+: Link
  • Facebook: Link
  • LocationMallanhold, Furillo

Posted 14 August 2013 - 09:38 AM

Yeah Mins would give you a small hour glass effect. :)

#33 Cycleboy

    Member

  • PipPipPipPipPip
  • Bad Company
  • 183 posts
  • LocationWisconsin

Posted 14 August 2013 - 09:42 AM

View PostJoseph Mallan, on 14 August 2013 - 09:23 AM, said:

Ok great minds think alike. I was thinking the same thing, Best accuracy up close, getting worse as you get further out.


I'm thinking he meant that +/- swap as correct mathmatical function... at range you are +1% deviation (pulling too far convergence) and under range is -3% deviation (not pulling in far enough). Those would be the correct math numbers applied to the angle of attack vector applied to each weapon.

#34 Joseph Mallan

    ForumWarrior

  • PipPipPipPipPipPipPipPipPipPipPipPipPipPipPip
  • FP Veteran - Beta 1
  • FP Veteran - Beta 1
  • 35,216 posts
  • Google+: Link
  • Facebook: Link
  • LocationMallanhold, Furillo

Posted 14 August 2013 - 09:46 AM

View PostCycleboy, on 14 August 2013 - 09:42 AM, said:


I'm thinking he meant that +/- swap as correct mathmatical function... at range you are +1% deviation (pulling too far convergence) and under range is -3% deviation (not pulling in far enough). Those would be the correct math numbers applied to the angle of attack vector applied to each weapon.

Ssssssssssh!
I'm on an Island here. :)

#35 DaZur

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • Elite Founder
  • Elite Founder
  • 7,511 posts
  • LocationWisconsin

Posted 14 August 2013 - 09:56 AM

View PostCycleboy, on 14 August 2013 - 09:42 AM, said:


I'm thinking he meant that +/- swap as correct mathmatical function... at range you are +1% deviation (pulling too far convergence) and under range is -3% deviation (not pulling in far enough). Those would be the correct math numbers applied to the angle of attack vector applied to each weapon.

View PostJoseph Mallan, on 14 August 2013 - 09:46 AM, said:

Ssssssssssh!
I'm on an Island here. :)

I'm using Polish math... If I have to hand out translational cards for effect it becomes awfully cumbersome and tedious. :lol:

#36 Ph30nix

    Member

  • PipPipPipPipPipPipPipPipPip
  • 2,444 posts

Posted 14 August 2013 - 10:03 AM

how about they just do whatever MW3 and MW4 did with convergance?

#37 Taemien

    Member

  • PipPipPipPipPipPipPipPip
  • Legendary Founder
  • Legendary Founder
  • 1,576 posts
  • LocationNorth Carolina

Posted 14 August 2013 - 11:00 AM

View PostmiSs, on 12 July 2013 - 02:28 PM, said:


Answer from Paul: Weapon convergence is a tough nut to crack. We want to keep the number of random “dice rolls” to a minimum, and network synchronization can become unpredictable when trying to determine a convergence point that may or may not be moving. It will be necessary to make the convergence point calculation server authoritive and that can cause a desync due to the fact that the simulation runs at different frequencies on the server and client.

While this is something we’ve wanted for a while, it’s becoming more and more apparent that it is going to take some serious engineering time to address. Currently, the engineers who would be working on this are already tasked with other high priority features and investigations. Pulling them off their current schedule will have both short and long term negative effects on the game as a whole, so the chances are very low that they will be able to address convergence any time soon. Convergence will always be something we will keep on the drawing board but as to when it can be tackled is not known at this time.


http://mwomercs.com/...vs-42-answered/

Alot of you have no background in communications and networking. Some of you do, but for some reason are missing how it works in an online game. So I'll explain what the dev is speaking about in that quote.

When he talks about desynchronization this is what he means in the context of MWO. Basically what you see on your screen isn't exactly what you get. A mech moving at 120kph to your left is probably, at least on their screen, 2-3 mech lengths or more to the left. You're effectively seeing a ghost image.

How HSR works with that is it knows the image you see isn't real, but is what you can perceive so it retroactively compensates your shot ahead for consistency. This way you can fire and hit what you see on your screen rather than having to shoot ahead (aka before HSR, and MW3).

However it works in reverse too. See the above statement assumes you are not moving. But you probably are, so the process has to work in reverse it has to account for your movement and compare it to the movement of everyone else on the map to make sure there is an impact, or intersect of weapon fire and mobile object (enemy or teammate).

Pretty complex thing. One they currently have a bug in but are working at getting that sorted out.

As for engineering time, this relates to the game engine. Not so much the one you see, but the networking one that works behind the scenes. See a mech being hit on your screen is not an event that simply says, "if graphic looks hit then apply damage somewhere" its more like something that looks like the quadratic formula that you had to do in algebra. The point is, the graphical hit has nothing to do with what happens behind the scene. In fact the graphics are simply showing a output of what actually happened behind the scene. As said, what you see isn't always what you get.

Because its that complex, convergence makes that equation a helluva lot more complex and quite frankly it has to be researched, tested, and made. The equation doesn't exist. Remember those philosophers that sat around drawing lines and stuff in the sand that gave us our current equations to find the circumference of a circle? Thats pretty much what the engineers have to figure out, something new, at least in respect to MWO specifically.

If you honestly think its easier than that... by all means hit the We're Hiring button below because apparently you know something they (and I) don't and you can get paid for your time.

Right now there is far more pressing matters then simply devoting code time to this. They need UI 2.0, they need to address weapon balance, they need to improve new player experience, and of course community warfare... all of which are not negatively impacted by no convergence changes. In otherwords, with the current pin point accuracy mechanics, those additions to the game can still go on. Especially since every other MechWarrior game made had it and did fine.

#38 Dimento Graven

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • Guillotine
  • Guillotine
  • 6,208 posts

Posted 14 August 2013 - 11:16 AM

View PostTaemien, on 14 August 2013 - 11:00 AM, said:


Alot of you have no background in communications and networking. Some of you do, but for some reason are missing how it works in an online game. So I'll explain what the dev is speaking about in that quote.

...
Great explanation, only there is a mechanism in operation already for convergence. PGI has admitted that. At one point when asked why switching from a far target to a close target and quickly firing results in the shots going on either side of the close target the response from PGI was along the lines of you haven't given the firing computer time to recalculate the weapon convergence. Wait a few moments, then fire.

So there is a convergence mechanism in the engine. Whether PGI is willing/able to fully utilize it is another question.

My understanding is that, typically at the most basic level you have two range variables: ACTUAL RANGE and MEASURED RANGE. ACTUAL RANGE is exactly what it sounds like the actual range between you and what you're targeting. MEASURED RANGE is a value used for simulating 'firing computers', the value can be adjusted at a given rate based on how long you hold your targeting reticule over the target, and other factors can be used to adjust the speed and accuracy of that measurement giving you a fudge factor. So that, no interference and with enough time and care the player can fire accurately, however, if there are mitigating circumstances, quickly switching from different ranged targets, heat affects slowing down targeting computer, heat affects throwing measurements off, etc. etc., you can simulate de-convergence by changing the MEASURED RANGE and having the weapons converge in front of or behind the target, partially or even completely missing the target.

#39 Taemien

    Member

  • PipPipPipPipPipPipPipPip
  • Legendary Founder
  • Legendary Founder
  • 1,576 posts
  • LocationNorth Carolina

Posted 14 August 2013 - 05:22 PM

View PostDimento Graven, on 14 August 2013 - 11:16 AM, said:

Great explanation, only there is a mechanism in operation already for convergence. PGI has admitted that. At one point when asked why switching from a far target to a close target and quickly firing results in the shots going on either side of the close target the response from PGI was along the lines of you haven't given the firing computer time to recalculate the weapon convergence. Wait a few moments, then fire.

So there is a convergence mechanism in the engine. Whether PGI is willing/able to fully utilize it is another question.

My understanding is that, typically at the most basic level you have two range variables: ACTUAL RANGE and MEASURED RANGE. ACTUAL RANGE is exactly what it sounds like the actual range between you and what you're targeting. MEASURED RANGE is a value used for simulating 'firing computers', the value can be adjusted at a given rate based on how long you hold your targeting reticule over the target, and other factors can be used to adjust the speed and accuracy of that measurement giving you a fudge factor. So that, no interference and with enough time and care the player can fire accurately, however, if there are mitigating circumstances, quickly switching from different ranged targets, heat affects slowing down targeting computer, heat affects throwing measurements off, etc. etc., you can simulate de-convergence by changing the MEASURED RANGE and having the weapons converge in front of or behind the target, partially or even completely missing the target.


That is not PGI's doing however. This is something already in the CryEngine. They have to work alongside that, or bypass it to get convergence to work. That involves recoding it from the ground up, hence the engineering issue.

#40 Dimento Graven

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • Guillotine
  • Guillotine
  • 6,208 posts

Posted 15 August 2013 - 06:53 AM

View PostTaemien, on 14 August 2013 - 05:22 PM, said:

That is not PGI's doing however. This is something already in the CryEngine. They have to work alongside that, or bypass it to get convergence to work. That involves recoding it from the ground up, hence the engineering issue.
First off, I am going to do something that you should NEVER do on a forum and that's admit ignorance:

I am completely ignorant as to the full convergence mechanism in the CryEngine.

That said, it can't be THAT hard can it? A well written engine would have that as an exposed function, allowing manipulation wouldn't it?





1 user(s) are reading this topic

0 members, 1 guests, 0 anonymous users