1126 replies to this topic

[Alistair Winter]

1000 replies!

Who wants to go through all the posts and look for a consensus?

[Kuritaclan]

Maybee: we agree to disagree.

[Burktross]

Alistair Winter, on 14 April 2015 - 05:26 AM, said:

1000 replies!

Who wants to go through all the posts and look for a consensus?

there is none. Angry neckbeards can't agree with each other.

[Mystere]

Alistair Winter, on 14 April 2015 - 05:26 AM, said:

1000 replies!

Who wants to go through all the posts and look for a consensus?

I'll just give you a TLDR: You will not find any.

Edited by Mystere, 14 April 2015 - 07:22 AM.

[MW Waldorf Statler]

the BT universe is absurd, on the one hand, you can not build a missile guidance technology more, but air-space hunter, you can not write as illiterate lexicons

[pbiggz]

Mystere, on 14 April 2015 - 07:20 AM, said:

I'll just give you a TLDR: You will not find any.

Because arm chair game designers think they know better than actual game designers, and im not talking about paul.

[Mystere]

E Rommel, on 14 April 2015 - 04:49 AM, said:

Ah, I see it's time for an explanation of how Minutes of Angle work. As the name might imply, it's a measure of the angle between the expected trajectory and the actual trajectory. A minute is 1/60th of a degree. Fortunately, some people interested in measuring how accurate their guns were converted this to dispersion for us a long time ago: 1 minute of angle is a dispersion of about 2.9 cm at 100m.

This relationship is linear with distance, because the sides of a triangle with fixed angles have a fixed ratio. 2.9cm at 100m is equal to 29cm at 1000m, which is equal to 1 minute of angle.

So I'm sure you can see why it's unusual that you got roughly 10 times that value from 30cm at 1000m, which is equal to 3cm at 100m.

The relationship between angle and dispersion is not linear of course, it's described by the tangent function, but as long as we are working with differences less than a degree (and we are) we can pretend it is without too much error.

60 MoA would be 174cm at 100m, or 1740cm at 1000m. That is 17.4 meters of dispersion at 1000m, not 2.

I'm not sure where you're getting your "R95" value from either, since you only need two standard deviations to reach 95% confidence. Since I have been trying to base my calculations on standard deviations, a 95% confidence level should double them, not quadruple them.

Granted, for the 88 tables 50% isn't quite 1 SD, but it's close enough to estimate that the 68% value is less than 40cm/68cm respectively (since that relationship is definitely not linear, it drops off quite rapidly).

Well, it looks like I botched the numbers. The 8.8cm KwK 43 has a CEP of somewhere between 30-50cm at 1000m. That is roughly a bit less than 1m R95 at 1000m.

Given this revelation, I'm still fine with my 2m R95 at 1000m idea. Although I can be convinced to drop it to 1m.

And as for running at full speed and firing multiple weapons, I'm sticking to my numbers for that.

E Rommel, on 14 April 2015 - 04:49 AM, said:

As far as "more skill", I don't see how it's any more skillfull when it wouldn't really change playstyles. It would just make some good mechs less good, some bad mechs more bad, and a handful of mechs don't care. Firing an alpha and twisting/covering would still get me more focused damage and less face time, it's just at sub-200 and over-500m ranges some of my damage would spread. Past 600 I don't care, since the meds fizzle before then and LPLs usually just tickle at that range.

I'd also have even more incentive to boat weapons with iddentical ranges, so I can converge them all to the same point. So I don't see what the acolytes of stock builds/schizo builds expect to gain from that. It would just be another nerf to mechs with poor hardpoint placement.

Seems like it would be a waste of effort, overall.

With regard to fixed convergence, yes, you are correct that game play in the 200-500m will more or less stay the same. But, not everybody plays at that range. Small knife fighters (< 150m) become tougher to kill and predatory against Mechs configured for medium to long range convergence, and snipers (800+), especially smaller ones, will become unmolested until the enemy can close in.

I can already hear loud howls of protest from the 200-500m players.

Edited by Mystere, 14 April 2015 - 08:42 AM.

[Almond Brown]

Burktross, on 13 April 2015 - 12:02 PM, said:

We'll call it recoil.

Lasers with recoil? Interesting...

[pbiggz]

Almond Brown, on 14 April 2015 - 09:57 AM, said:

Lasers with recoil? Interesting...

If you use my system, you get to pick which weapons generate reticle bloom and which dont, and how much.

[Almond Brown]

E Rommel, on 14 April 2015 - 04:49 AM, said:

Ah, I see it's time for an explanation of how Minutes of Angle work. As the name might imply, it's a measure of the angle between the expected trajectory and the actual trajectory. A minute is 1/60th of a degree. Fortunately, some people interested in measuring how accurate their guns were converted this to dispersion for us a long time ago: 1 minute of angle is a dispersion of about 2.9 cm at 100m.

This relationship is linear with distance, because the sides of a triangle with fixed angles have a fixed ratio. 2.9cm at 100m is equal to 29cm at 1000m, which is equal to 1 minute of angle.

So I'm sure you can see why it's unusual that you got roughly 10 times that value from 30cm at 1000m, which is equal to 3cm at 100m.

The relationship between angle and dispersion is not linear of course, it's described by the tangent function, but as long as we are working with differences less than a degree (and we are) we can pretend it is without too much error.

60 MoA would be 174cm at 100m, or 1740cm at 1000m. That is 17.4 meters of dispersion at 1000m, not 2.

I'm not sure where you're getting your "R95" value from either, since you only need two standard deviations to reach 95% confidence. Since I have been trying to base my calculations on standard deviations, a 95% confidence level should double them, not quadruple them.

Granted, for the 88 tables 50% isn't quite 1 SD, but it's close enough to estimate that the 68% value is less than 40cm/68cm respectively (since that relationship is definitely not linear, it drops off quite rapidly).

As far as "more skill", I don't see how it's any more skillfull when it wouldn't really change playstyles. It would just make some good mechs less good, some bad mechs more bad, and a handful of mechs don't care. Firing an alpha and twisting/covering would still get me more focused damage and less face time, it's just at sub-200 and over-500m ranges some of my damage would spread. Past 600 I don't care, since the meds fizzle before then and LPLs usually just tickle at that range.

I'd also have even more incentive to boat weapons with identical ranges, so I can converge them all to the same point. So I don't see what the acolytes of stock builds/schizo builds expect to gain from that. It would just be another nerf to mechs with poor hardpoint placement.

Seems like it would be a waste of effort, overall.

Well said, and, if convergence was settable in Mechlab, a player would then be "forced to build" around it every time, and of course, once (if) implemented, you could not have it any other way.

A game can't have both "almost instant (auto)" and a "settable" convergence configuration. You would have to settle on one or the other, from the design side of things. So if MechLab settable turned out to be actually a stupid thing, then going back may not be an option for the Dev team.

On the run "convergence" changes would likely be abuseable. So nope to that.

Edited by Almond Brown, 14 April 2015 - 10:22 AM.

[Almond Brown]

pbiggz, on 14 April 2015 - 10:09 AM, said:

If you use my system, you get to pick which weapons generate reticle bloom and which dont, and how much.

Sorry I must have missed that part. So each weapon I load on my Mech has a different "bloom effect" value"? What happens when I fire multiples, are those individual values "additive" or "multiplicative" and how does that system distinguish between values of weapons when more than one is fired and optimal ranges are very different?

I would love to see your Chart of different "bloom effect" values based on each weapon and range class and how they interact in varied fired clusters. We can work on how possible Mech speeds and Movement also effect the "bloom effect" values separately ok? Thanks

Edited by Almond Brown, 14 April 2015 - 10:32 AM.

[pbiggz]

Almond Brown, on 14 April 2015 - 10:29 AM, said:

Sorry I must have missed that part. So each weapon I load on my Mech has a different "bloom effect" value"? What happens when I fire multiples, are those individual values "additive" or "multiplicative" and how does that system distinguish between values of weapons when more than one is fired and optimal ranges are very different?

I would love to see your Chart of different "bloom effect" values based on each weapon and range class and how they interact in varied fired clusters. We can work on how possible Mech speeds and Movement also effect the "bloom effect" values separately ok? Thanks

Working on something now.

Edited by pbiggz, 14 April 2015 - 10:50 AM.

[pbiggz]

Try to break this. No really. That's how you test to see if a system will work. You create it, then you try to destroy and see if it holds up.

ps: i realize even a 2 degree cone of fire at 300 meters would be huge. This is just proof that the math works. You sub in better values if you actually want to implement it.

Edited by pbiggz, 14 April 2015 - 01:26 PM.

[Gorgo7]

Convergence is excellent the way it is. The real issue is the low velocity of LRMs. They should be much quicker.

Edited by Gorgo7, 14 April 2015 - 01:21 PM.

[pbiggz]

Gorgo7, on 14 April 2015 - 01:21 PM, said:

Convergence is excellent the way it is. The real issue is the low velocity of LRMs. They should be much quicker.

Convergence is NOT excellent the way it is, and this discussion wasn't about LRMs.

[Kuritaclan]

Actually this issnt convergence it is a CoF modifer by weapons fired. If you apply to every weapon this
calculation angle you creat in first place value that could determine a radius of a circle for a maximum displacment. After this you need to define what angle it need to be replaced (value between 1 and 360)
To use your example with 2 ppc + 2x ac5

PPC1 is Phi1 with 30°
PPC2 is Phi2 with -45° (315°)
and so on
now it would be randomly distributed on the circle. To make it a scatter group you need to rng the radius for every single weapon. For example you can separate the maximum deplacment into 100 parts and make another diceroll (1 to 100). Also this is not the hit zone - since balistic deplacment by gravity (the game sim that) is not included. At 500 to 1000m the bullet drop quit a bit, mostly recognizable with the ac20. This would after the traveltime mean that if the roll got max deplacement at angle -45° it is quit outside of the circle. Your spread get bigger by such effects. At ~200m means in mwo i made two pictures. Also note that pixels do not be fully accurate because of resolution and sampling of my mouse movement there is some failure within.


As you can see while a LBX 10 has a optimal range of 540m the spread is that high that is not far away to miss if you are above 200m range. If now for those who think we need such a deplacment for Lasers PPC and ACs, than i welcome you all to Spreadwarrior Online.

Edited by Kuritaclan, 14 April 2015 - 02:54 PM.

[operatorZ]

Gorgo7, on 14 April 2015 - 01:21 PM, said:

Convergence is excellent the way it is. The real issue is the low velocity of LRMs. They should be much quicker.

8/10...almost got me...gj

Kuritaclan, on 14 April 2015 - 02:41 PM, said:

Actually this issnt convergence it is a CoF modifer by weapons fired. If you apply to every weapon this
calculation angle you creat in first place value that could determine a radius of a circle for a maximum displacment. After this you need to define what angle it need to be replaced (value between 1 and 360)
To use your example with 2 ppc + 2x ac5
PPC1 is Phi1 with 30°
PPC2 is Phi2 with -45° (315°)
and so on
now it would be randomly distributed on the circle. To make it a scatter group you need to rng the radius for every single weapon. For example you can separate the maximum deplacment into 100 parts and make another diceroll (1 to 100). Also this is not the hit zone - since balistic deplacment by gravity (the game sim that) is not included. At 500 to 1000m the bullet drop quit a bit, mostly recognizable with the ac20. This would after the traveltime mean that if the roll got max deplacement at angle -45° it is quit outside of the circle. Your spread get bigger by such effects. At ~200m means in mwo i made two pictures. Also note that pixels do not be fully accurate because of resolution and sampling of my mouse movement there is some failure within.

As you can see while a LBX 10 has a optimal range of 540m the spread is that high that is not far away to miss if you are above 200m range. If now for those who think we need such a deplacment for Lasers PPC and ACs, than i welcome you all to Spreadwarrior Online.

interesting! tell me MOAR!!

[pbiggz]

Kuritaclan, on 14 April 2015 - 02:41 PM, said:

Actually this issnt convergence it is a CoF modifer by weapons fired. If you apply to every weapon this
calculation angle you creat in first place value that could determine a radius of a circle for a maximum displacment. After this you need to define what angle it need to be replaced (value between 1 and 360)
To use your example with 2 ppc + 2x ac5
PPC1 is Phi1 with 30°
PPC2 is Phi2 with -45° (315°)
and so on
now it would be randomly distributed on the circle. To make it a scatter group you need to rng the radius for every single weapon. For example you can separate the maximum deplacment into 100 parts and make another diceroll (1 to 100). Also this is not the hit zone - since balistic deplacment by gravity (the game sim that) is not included. At 500 to 1000m the bullet drop quit a bit, mostly recognizable with the ac20. This would after the traveltime mean that if the roll got max deplacement at angle -45° it is quit outside of the circle. Your spread get bigger by such effects. At ~200m means in mwo i made two pictures. Also note that pixels do not be fully accurate because of resolution and sampling of my mouse movement there is some failure within.

As you can see while a LBX 10 has a optimal range of 540m the spread is that high that is not far away to miss if you are above 200m range. If now for those who think we need such a deplacment for Lasers PPC and ACs, than i welcome you all to Spreadwarrior Online.

Now you're rambling about rng. Nowhere in my math is there RNG, it even states that if you bothered to read it and I know you didn't.

Why are you so militantly opposed to this idea? Are you scared its going to make it too hard for you? Its not.

[Kuritaclan]

operatorZ, on 14 April 2015 - 02:45 PM, said:

interesting! tell me MOAR!!

I uploaded the pics now to show what a degree is in mwo

pbiggz, on 14 April 2015 - 02:49 PM, said:

Now you're rambling about rng. Nowhere in my math is there RNG, it even states that if you bothered to read it and I know you didn't.

Why are you so militantly opposed to this idea? Are you scared its going to make it too hard for you? Its not.

I do not rambling about rng. I add up your concept with rng, to simulate a hit distribution. You also can define the postions for like 4 weapons at the circle on lets say 0°, 90°, 180°, 270°, if you don't wanna add in rng. And if you do not adjust the radius you have the maximum range between all 4 hitpoints for those 4 weapons. However with rng the 4 weapons of the exampe do the same, what the lbx does to its shrapnel.

I'm agianst it because there are weapons wich do spread damage like a lbx or a srm - and there are weapons like ppc, who "should" hit what i aimed for (and because of velocity of it and relative movement of you and the enemy it does not hit 100% the part you aimed for) If i fire to much of them i get a heat penality.

Look at the pictures if you make the radius of this circle to short it does not matter for lets say 4 lasers, since the beam duration add in a natural displacment, however if you mage the radius to big it is a shotgun effect you have with the usage of more weapons. You add in modifiers, and they can be used to adjust. But to be honest i don't see how this will help. Alpha would be over in this case yes, but if you can not alpha your lets say 5+ lasers why bring them if you have a better alpha with natural spread weapons what do more damage with less penality like heat.

This is 200m! And the picture shows an Atlas, you know there are light mechs in the game. They would profit by such a change, since you may not miss on big mechs, but probably on lights. And again if you balance the cone on lights it does may not mean much to assault mechs, since their parts have a bigger surface your cone apply to on the same range.

Now we are also talking of ppc and weapons named Extreme Range Large Laser. Now say me what is the correct degree of spread you would acept at 1000m that alpha is no longer available, but you may at least hit with one weapon? - you know how small the cone becomes, when you this little circle use on the low range, the spread is nearly not existent.

I appreciate what you have done, but i don't think it is a good solution. Another point is you have a linear scaling of weapons in your formula - if you wanna do something like that you may need a "ghost" if/then query, so that 2 weapons of the same kind are not hit with penality of this spread, but 4 get kicked hard for example - maybe an expotenital increase would help too.

Edited by Kuritaclan, 14 April 2015 - 03:34 PM.

[Gorgo7]

Re: The OP

No, do not get rid of convergence.