|
|
Post by nerd1000 on Nov 8, 2016 12:10:10 GMT
I've noticed something a bit odd with weapon spread:  As you can see, the bullets are not randomly distributed within their weapon's spread cone- instead they are all at the edges, indicating that every round is fired with the maximum possible deviation from the aim point, in a random direction. Is this correct/intentional? As it is this means that no weapon can hit its target until the gun's spread cone is smaller than the target's cross section, which obviously puts a significant limit on your effective range even when the enemy is flying on a very predictable course. |
|
|
|
Post by dragonkid11 on Nov 8, 2016 12:20:50 GMT
Hmm, this was actually very similar to a spread pattern problem World of Tank have some years ago.
Where the round are more likely to go around the edge instead of the center statistically when the aim circle is at its largest even though the round should go everywhere in the circle equally.
Can't remember how they fix this though.
|
|
|
|
Post by RA2lover on Nov 8, 2016 15:22:27 GMT
World of Tanks used a gaussian curve to determine weapon accuracy. Supposedly, two standard deviation of the shots were supposed to land inside the aiming circle.
In practice, there's always that 3-or-more-sigma shot that gets thrown multiple times outside the circle because the gaussian curve tapers off at infinity. IIRC this was fixed by re-rolling the shot if it missed hard enough.
|
|
|
|
Post by concretedonkey on Nov 9, 2016 6:50:44 GMT
I'm getting very similar patterns of conventional gun barrages fired on a target large enough to see the spread... can't do a shot now however, I'm at work.
|
|
|
|
Post by goduranus on Nov 11, 2016 7:40:50 GMT
I am seeing this pattern for solid slug shots as well. Should be reported as a bug I think.
|
|
khenderson
New Member
my god, it's full of missiles
Posts: 40 
|
Post by khenderson on Nov 13, 2016 17:38:02 GMT
I am seeing this pattern for solid slug shots as well. Should be reported as a bug I think. This might instead be a matter of the barrel flexing. The calculations in the weapon designer appear to underestimate how much the barrel flexes in operation. So if you try to maximize range, or minimize weight for a given engagement range, you end up with an inaccurate weapon. Take the following two railguns as an example. The first one has a rated engagement range against capital ships of 354 km, while the second is only rated at 164. Attachment DeletedAttachment DeletedYet once you set them to ignore range, the second weapon is much more accurate at any range, apparently due to the much more rigid barrel. The shot pattern from the first weapon appears somewhat similar to that in the first post, with the majority of the rounds at the edge of a circle around the target. While the second forms a tighter circle with a moderately more random distribution. Edit: as an experiment I should have done before posting, I increased the barrel thickness on the first railgun dramatically. The range estimate against capital ships dropped to 286 km, but it's now hitting with virtually every round at 900+ kilometers. Attachment Deleted |
|
|
|
Post by n2maniac on Nov 13, 2016 22:49:26 GMT
I am seeing this pattern for solid slug shots as well. Should be reported as a bug I think. This might instead be a matter of the barrel flexing. The calculations in the weapon designer appear to underestimate how much the barrel flexes in operation. So if you try to maximize range, or minimize weight for a given engagement range, you end up with an inaccurate weapon. Take the following two railguns as an example. The first one has a rated engagement range against capital ships of 354 km, while the second is only rated at 164. Yet once you set them to ignore range, the second weapon is much more accurate at any range, apparently due to the much more rigid barrel. The shot pattern from the first weapon appears somewhat similar to that in the first post, with the majority of the rounds at the edge of a circle around the target. While the second forms a tighter circle with a moderately more random distribution. Edit: as an experiment I should have done before posting, I increased the barrel thickness on the first railgun dramatically. The range estimate against capital ships dropped to 286 km, but it's now hitting with virtually every round at 900+ kilometers. Range definition bug! Time to retrofit my 20-30km/s coilguns / railguns. |
|
|
|
Post by qswitched on Nov 13, 2016 23:34:54 GMT
This looks like a bug. The shots should be biased towards the outer edge, but some still should be in the center.
|
|
|
|
Post by ash19256 on Nov 14, 2016 0:10:49 GMT
This looks like a bug. The shots should be biased towards the outer edge, but some still should be in the center. I could get an even distribution with an increased dispersion at extreme ranges, but why would the shots be biased to the outer edge? Logically, there should be an even distribution, or at least one would think so. |
|
|
|
Post by subunit on Nov 14, 2016 20:15:07 GMT
This looks like a bug. The shots should be biased towards the outer edge, but some still should be in the center. I could get an even distribution with an increased dispersion at extreme ranges, but why would the shots be biased to the outer edge? Logically, there should be an even distribution, or at least one would think so. I think it depends on the system. All of the actually-existing slugthrowers I've used tend to string rounds in patterns that are related to the characteristic of the system that is producing the deviation from the point of aim- barrel or stock flex for instance will string rounds along the axis of the flex. I don't know exactly what's being represented in CoaDE, but I suspect some deviation is applied to every round in a random direction (I don't think the guns are modelled so precisely that you could say anything useful about what pattern a particular slugthrower should produce)- so at long range rounds will tend to cluster around a circle whose radius is defined by the median deviation value. |
|
erik
New Member
Posts: 34
|
Post by erik on Nov 15, 2016 7:24:46 GMT
I have noticed this too.
OTOH very high muzzle speed, very accurate rail guns that I've designed seem to fire all rounds along edges of a cone, but some are better distributed. OTOH the 60 mm stock cannon is VERY accurate way, way beyond its listed maximum range. Over 3 times further than it should hit all the rounds come down in perfect line and hit the same spot when aiming at a component; you can see several seconds worth of fire hit the same spot and penetrate armor while over 10 times faster rail gun shells that are supposed to be more accurate too just blast all over the place, many missing target ship entirely.
|
|
|
|
Post by zuthal on Nov 24, 2016 19:46:30 GMT
This looks like a bug. The shots should be biased towards the outer edge, but some still should be in the center. AFAIK, real-life firearms have a roughly gaussian shot dispersion, biased towards the center. Is there any reason it should be different for the guns in CoaDE? Also, if that is implemented like that, I would suggest truncating the gaussian distribution at, say, three standard deviations, and in the gun designer, showing the angular deviation at one standard deviation. |
|
