1094 replies to this topic

[Nemesis Duck]

Why not produce a series of images that show the same scenarios but aiming at the arms instead? Show us the miss rate that it implies and then let's decide.

[Hotthedd]

Imperius, on 14 February 2016 - 09:32 AM, said:

I'll list off a few negatives if this was implemented.

HSR panic attack
Close grouped weapon clusters become meta chassis
Tutorial has to be rewritten
Implemented AI rewrite
All mech rigs readjusted

Or you could just make lasers a little more hot or equipment more sought after.
Increase CT armor on all mechs or lower all DMG numbers by x percent and buff ammo.

Simple fix to increase TTK I mean after all that is the main agenda right?

You are dodging the question. One of the big points your side of the argument keeps making is that the OP's suggestion lowers the skill requirement. If that is the case, then wouldn't everybody choose it in PvP, given the choice of having either?

[pyrocomp]

Imperius, on 14 February 2016 - 09:32 AM, said:

I'll list off a few negatives if this was implemented.

HSR panic attack
Close grouped weapon clusters become meta chassis
Tutorial has to be rewritten
Implemented AI rewrite
All mech rigs readjusted

Or you could just make lasers a little more hot or equipment more sought after.
Increase CT armor on all mechs or lower all DMG numbers by x percent and buff ammo.

Simple fix to increase TTK I mean after all that is the main agenda right?

HSR will not be affected at all. Really. CoF does not touch HSR.
CoF does not depend on weapon location (can be made so, but arms will have it worse), so no bonus from cluster location. For no convergence at all the clustered hardpoints are a bonus, yes.
As this mechinc is pread through almost any other game the tutorial will be redundant as people are already familiar with this thing. You do not have tutorial on how to move mouse in this game for the very same reason.
Also no need to implement any changes to mythical AI as aiming and shooting are not affected, only weapon functioning.
All mechs are reajusted after every change to the weapons made in almost every patch.
So nope.

But, the second part may make sence only if those little were not that little, but TT-crowd will go wild on non-TT values.

[Mystere]

Imperius, on 14 February 2016 - 08:51 AM, said:

What's wrong with the macho man? I think it's relevant to this thread. He was crazy just like this proposal.

If that was your point, a picture of Donald Trump would be more appropriate. People are going to so deserve him when he wins.

[pbiggz]

pyrocomp, on 14 February 2016 - 09:51 AM, said:

HSR will not be affected at all. Really. CoF does not touch HSR.
CoF does not depend on weapon location (can be made so, but arms will have it worse), so no bonus from cluster location. For no convergence at all the clustered hardpoints are a bonus, yes.
As this mechinc is pread through almost any other game the tutorial will be redundant as people are already familiar with this thing. You do not have tutorial on how to move mouse in this game for the very same reason.
Also no need to implement any changes to mythical AI as aiming and shooting are not affected, only weapon functioning.
All mechs are reajusted after every change to the weapons made in almost every patch.
So nope.

But, the second part may make sence only if those little were not that little, but TT-crowd will go wild on non-TT values.

PGI says if they adjust weapons convergence that means having in practice multiple reticles for every mech, and, because of the way HSR is implemented, it would involve a significant rewriting of hit reg code both client and server side, and by extension, it means AI will have to aim differently, requiring a rewrite. Also, the AI is not mythical. The academy is their test bed for it. Go check it out and you'll see bits and pieces of what will eventually be their full blooded AI.

You talk with alot of authority for someone who hasn't seen the source code. I haven't either, im just telling you what Russ told us. Its not going to happen.

But thats not what this thread is about now.

This is.

Edited by pbiggz, 14 February 2016 - 11:18 AM.

[pyrocomp]

CoF does not require multiple recticles as they are not shown and do not exist. CoF can be implemented purely serverside as 4 lines of code rotating recived shot derection (5 numbers, x,y,z,theta,phi) on random angle (I'll skip the lines and direct you to google for rotation matrix yourself). Only after that the shot goes to HSR and hitreg. As every weapons shot is tracked individually it affects nothing. Tracking the convergence point was tiresome and bad for the used implementation of HSR I do remember that. But you still track it as it is now, nothing changes client-side. AI will have to aim the same way allining it's center of the CoF as it is now aligning shots. There is nothing to rewrite. Really.

The slow convergence is really a heck of... well, it's on opposite side of simple in this case. Yes. Thats why in the whole offender - long range high alpha instant pinpoint convergence - the pinpoint is usually discussed. Ranges and damages are... well people whine really loud if you touch stats. Limit alpha damages more lights at small ranges and some brawlers and is somewhat illogical (even with limitations on the power draw you still can fire everything even it's fianlly charged). Instant is here due to technicals (yes, I remember when convergence was delayed). And we have to have convergence or the game becomes something unplayable. So kill pinpoint at long ranges to bring down the whole chain. There is no need to kill pinpoint on middle or small ranges (maybe later it will becom apparent, but not now).

I haven't seen the source code, but there are general ways of doing certain things. I agree that PGI could've invented their own way (no news here), but doubt they did it completely different. Some things are just so basic as you learn to code those things that way while learning how to code (and I haven't seen different implementations in any code I came across both in games and in scientific codes).

Edited by pyrocomp, 14 February 2016 - 11:33 AM.

[Mystere]

Imperius, on 14 February 2016 - 09:32 AM, said:

I'll list off a few negatives if this was implemented.

HSR panic attack
Close grouped weapon clusters become meta chassis
Tutorial has to be rewritten
Implemented AI rewrite
All mech rigs readjusted

Or you could just make lasers a little more hot or equipment more sought after.
Increase CT armor on all mechs or lower all DMG numbers by x percent and buff ammo.

Simple fix to increase TTK I mean after all that is the main agenda right?

Or, instead of touching each and every affected weapon, mech, and other pieces of equipment, just have convergence on lock and fixed (but user-adjustable) convergence in the absence of one.

All the necessary code -- with the exception of user adjustment of convergence distance -- is already there and should be usable unless PGI's software development assets look like this:

Edited by Mystere, 14 February 2016 - 12:29 PM.

[pyrocomp]

Mystere, on 14 February 2016 - 12:28 PM, said:

I hope those are labled at connectors and some map is there taped to it somewhere... Or come in, unplug 10 random cables, exchange positions and... 3-4 hours of checks and all that.

[Mystere]

pbiggz, on 14 February 2016 - 11:17 AM, said:

PGI says if they adjust weapons convergence that means having in practice multiple reticles for every mech, and, because of the way HSR is implemented, it would involve a significant rewriting of hit reg code both client and server side, and by extension, it means AI will have to aim differently, requiring a rewrite. Also, the AI is not mythical. The academy is their test bed for it. Go check it out and you'll see bits and pieces of what will eventually be their full blooded AI.

You talk with alot of authority for someone who hasn't seen the source code. I haven't either, im just telling you what Russ told us. Its not going to happen.

Citation please. I'd like to see what PGI actually said.

pbiggz, on 14 February 2016 - 11:17 AM, said:

But thats not what this thread is about now.

This is.

I already said this guy is more suitable for your purposes:

Edited by Mystere, 14 February 2016 - 12:37 PM.

[AEgg]

oldradagast, on 14 February 2016 - 07:13 AM, said:

Because zero convergence that you have to manually adjust is unworkable, as demos prove, and would be immediately "solved" with macros that would put us right back to where we are today, except anyone without a macro would be dead meat.

The slight scatter from a cone of fire at long ranges is the only logical solution. It does not fundamentally break any core game mechanics or add another level of complexity, but it reduces the overwhelming effectiveness of long-range, pinpoint alphas. It also adds a bit of the damage scatter that Battletech rules require to work and offers more design space: if you have a target lock, the cone narrows and your shots become more accurate, but if you're almost overheating, the cone wides a bit, and your shots become less accurate. It makes sense and is rather like Battletech.

Fixed convergence isn't unworkable, its simply something you have to learn to master.

You don't let users adjust fixed convergence in-game, it can only change in the mechlab. Pick your convergence distance, and live with it. You aim under it, your shots split either side, you aim past it, they cross and then split either side. Even if it was adjustable in game, you could not macro it, as macros cannot read information from the screen, they can only repeat keystrokes (so they can't know what the current convergence values you want are).

Letting the user set their convergence values both fixes the always-pinpoint problem while also opening up a lot more skill and strategy, since you want to position yourself where you are shooting at your convergence range.

The idea of perfect convergence when locked and fixed or no convergence elsewhere is also viable, but it comes with a slew of other problems (i.e. ECM).

[Troutmonkey]

Quote

Anybody advocating a CoF (which I am not entire against) needs to simply ask themselves how stupid said system would look on a Hunchback-4P. The answer is very.

What's the point of installing weapon arrays if each element has huge independent accuracy errors?

If they have a shared gimbal then that just obsoletes every other design with widely spaced hardpoints.

Its lose-lose, don't even go there.

Heh, CoF looks fine. It's Fixed/Manual convergence that looks stupid.

It doesn't really matter anyway, as under my proposal (link in sig) 6 MLAS would only have the slightest CoF anyway and 5 MLAS probably would have none.

Edited by Troutmonkey, 14 February 2016 - 03:43 PM.

[AEgg]

Troutmonkey, on 14 February 2016 - 03:41 PM, said:

Heh, CoF looks fine. It's Fixed/Manual convergence that looks stupid.

-snip-

Your example mech used a very poor choice of fixed convergence, i.e., far too close. How much of an impact convergence has depends on how high the angle is, and that angle depends on 1. how far apart your arms/weapons are and 2 what the ratio of that to your total range is.

As the convergence point gets further away from the weapons, the portion of that range which is near pinpoint goes up.

Eyeballing it the target mech is what, five times your width away from you? Assuming mechs are ~5m wide, that means it's like 25m from you, and your convergence is set at like 10-15m for weapons with a 250m range or more. You'd probably set your convergence at a range of about 300-400m for medium lasers, at which point, the angles involved are much smaller. You'd be doing badly at very close range with that, true, but if you pick your convergence, that's what you get. You'd set your smaller weapons to lower ranges, but still nothing as small as five to one.

Fixed convergence, you would have to simply learn to aim with it. Want to fire a hunch of lasers on your right shoulder?, aim to the left of your target.

[Troutmonkey]

AEgg, on 14 February 2016 - 03:57 PM, said:

Your example mech used a very poor choice of fixed convergence, i.e., far too close. How much of an impact convergence has depends on how high the angle is, and that angle depends on 1. how far apart your arms/weapons are and 2 what the ratio of that to your total range is.

As the convergence point gets further away from the weapons, the portion of that range which is near pinpoint goes up.

Eyeballing it the target mech is what, five times your width away from you? Assuming mechs are ~5m wide, that means it's like 25m from you, and your convergence is set at like 10-15m for weapons with a 250m range or more. You'd probably set your convergence at a range of about 300-400m for medium lasers, at which point, the angles involved are much smaller. You'd be doing badly at very close range with that, true, but if you pick your convergence, that's what you get. You'd set your smaller weapons to lower ranges, but still nothing as small as five to one.

Fixed convergence, you would have to simply learn to aim with it. Want to fire a hunch of lasers on your right shoulder?, aim to the left of your target.

New players can't even figure out how to walk properly, how do you think they're going to figure out manual convergence? Something like what I showed is going to be a common outcome. Other players most likely won't be bothered with constantly adjusting it for perfect aim and will just set it really high. It's not a fun mechanic, and it doesn't give out balanced results. All in all it's a bad mechanic that doesn't fix the issue and creates dozens more. Fixed and Zero convergence suffer from nearly all the same issues. They're just bad mechanics and shouldn't be considered at all.

Delayed convergence and variations of CoF (like my TCL proposal in sig) are the only two viable options left, and delayed convergence has been ruled out by PGI because of server issues

Edited by Troutmonkey, 14 February 2016 - 04:06 PM.

[AEgg]

Troutmonkey, on 14 February 2016 - 04:04 PM, said:

New players can't even figure out how to walk properly, how do you think they're going to figure out manual convergence? Something like what I showed is going to be a common outcome. Other players most likely won't be bothered with constantly adjusting it for perfect aim and will just set it really high. It's not a fun mechanic, and it doesn't give out balanced results. All in all it's a bad mechanic that doesn't fix the issue and creates dozens more. Fixed and Zero convergence suffer from nearly all the same issues. They're just bad mechanics and shouldn't be considered at all.

Delayed convergence and variations of CoF (like my TCL proposal in sig) are the only two viable options left, and delayed convergence has been ruled out by PGI because of server issues

True, anyone who doesn't know what they're doing would have big problems with convergence, I'll give you that.

We know we can't have delayed convergence, and cone of fire has it's fair share of problems too (being random, mostly).

What about "almost perfect" convergence? It's not my idea of course, but the basic principle is that instead of always having perfect convergence right on your target, what if you always had convergence to a point 20m behind your target? (probably scaling with range so at 50m the point was 5m off but at 1000m it was 50 or 100m off.)

[pyrocomp]

AEgg, on 14 February 2016 - 04:23 PM, said:

True, anyone who doesn't know what they're doing would have big problems with convergence, I'll give you that.

We know we can't have delayed convergence, and cone of fire has it's fair share of problems too (being random, mostly).

What about "almost perfect" convergence? It's not my idea of course, but the basic principle is that instead of always having perfect convergence right on your target, what if you always had convergence to a point 20m behind your target? (probably scaling with range so at 50m the point was 5m off but at 1000m it was 50 or 100m off.)

Hows that is different from the proposed CoF? With respect to usual 'wepons should fire where I'm aiming them'. Other than that CoF punishes long ranges and constant shift punishes low ranges. Will the KGC be able to hit a facehugger from his arm-mounted weapons?

[AEgg]

pyrocomp, on 14 February 2016 - 04:26 PM, said:

Hows that is different from the proposed CoF? With respect to usual 'wepons should fire where I'm aiming them'. Other than that CoF punishes long ranges and constant shift punishes low ranges. Will the KGC be able to hit a facehugger from his arm-mounted weapons?

Imperfect convergence is still 100% predictable, assuming the amount it is off by is always the same or always varies in the same way (i.e. a constant 50m or always 5% of the range to target).

Cone of fire makes long range terrible, cylinder of fire, you can guess but can't really be sure where weapons will hit. Offset convergence and you know exactly where your weapons will hit, but it's not all in exactly the same place.

I believe it would be possible to have an offset convergence value that always gives your weapons the exact same spread pattern on the target (calculate the range to target and use whatever offset gets the desired amount of spread from the furthest apart weapons). For example, if you fired two ML from left and right torso, they would always hit 1m apart on the target, no matter how far away the target was.

[Troutmonkey]

AEgg, on 14 February 2016 - 04:33 PM, said:

Imperfect convergence is still 100% predictable, assuming the amount it is off by is always the same or always varies in the same way (i.e. a constant 50m or always 5% of the range to target).

Cone of fire makes long range terrible, cylinder of fire, you can guess but can't really be sure where weapons will hit. Offset convergence and you know exactly where your weapons will hit, but it's not all in exactly the same place.

I believe it would be possible to have an offset convergence value that always gives your weapons the exact same spread pattern on the target (calculate the range to target and use whatever offset gets the desired amount of spread from the furthest apart weapons). For example, if you fired two ML from left and right torso, they would always hit 1m apart on the target, no matter how far away the target was.

The thing about my proposal is that it doesn't harm long range any more than short range, and it only harms high alphas. The idea is to punish high alpha with a degree of inaccuracy so they can't focus on the same component with every weapon.
Other suggestions that mess with convergence harm regular game play and unfairly punish all players even if they are shooting a "reasonable" amount of weaponry. It's one of the reasons I'm against any sort of accuracy penalties for movement or heat.

[Mystere]

AEgg, on 14 February 2016 - 12:59 PM, said:

The idea of perfect convergence when locked and fixed or no convergence elsewhere is also viable, but it comes with a slew of other problems (i.e. ECM).

One of the free analog inputs on my HOTAS throttle, or a mouse scroll wheel, can deal with that just fine.

[Mystere]

Troutmonkey, on 14 February 2016 - 03:41 PM, said:

Heh, CoF looks fine. It's Fixed/Manual convergence that looks stupid.

Hold on a second! So far I haven't dissed on your CoF idea. Don't make me start now after 50 pages.

[Troutmonkey]

Mystere, on 14 February 2016 - 06:10 PM, said:

Hold on a second! So far I haven't dissed on your CoF idea. Don't make me start now after 50 pages.

haha I can't keep track of who is for or against my ideas now. Either way each idea should be evaluated on it's own merits, not who's advocating for them. I just can't get behind Zero, Manual or Fixed convergence at all as it just fundamentally breaks gameplay without specifically addressing the problem.

It's a sledgehammer solution that "solves" the problem while smashing everything around it in the process.
"look, pinpoint alpha's are solved!"
"what about all this mess you just made?"
"but, but, I solved it!"

Edited by Troutmonkey, 14 February 2016 - 06:28 PM.