A ship design concept

[ToddoftheYear]

Layers:7 Surface: 198 Volume:231

Surface = Capacitors + Ship Core = 198
Volume = Shield Rechargers = 377
Your math is wrong.

Funny you didn't bother to even put up a radius on it to compare to.

Octahedrons have the highest volume to surface area ratio in Starmade. I didn't include radius because I wasn't comparing radius.

 

[GRHayes]

Surface = Capacitors + Ship Core = 198
Volume = Shield Rechargers = 377
Your math is wrong.

Either you have something off screen or the game is simply wrong. It wouldn't be the first time block count has been off in this game.
The center layer is the only one that is not duplicated
Volume:
1*2+
2*2*2+
3*3*2+
4*4*2+
5*5*2+
6*6*2+
7*7=231
Surface:
1*2+
4*2+
8*2+
12*2+
16*2+
20*2+
24*2+
28=198

Don't believe me do it manually yourself.

 

[ToddoftheYear]

Either you have something off screen or the game is simply wrong. It wouldn't be the first time block count has been off in this game.
The center layer is the only one that is not duplicated
Volume:
1*2+
2*2*2+
3*3*2+
4*4*2+
5*5*2+
6*6*2+
7*7=231

lel

Don't believe me do it manually yourself.

Seems fine to me.

Surface 6 Volume 1
Surface 18 Volume 7
Surface 38 Volume 25
Surface 66 Volume 63
Surface 102 Volume 129
Surface 146 Volume 231
Surface 198 Volume 377
Surface 258 Volume 575
Surface 326 Volume 833
Surface 402 Volume 1159
Surface 486 Volume 1561
Surface 578 Volume 2047
Surface 678 Volume 2625
Surface 786 Volume 3303
Surface 902 Volume 4089
Surface 1026 Volume 4991
Surface 1158 Volume 6017
Surface 1298 Volume 7175
Surface 1446 Volume 8473
Surface 1602 Volume 9919
Surface 1766 Volume 11521
Surface 1938 Volume 13287
Surface 2118 Volume 15225
Surface 2306 Volume 17343
Surface 2502 Volume 19649
Surface 2706 Volume 22151
Surface 2918 Volume 24857
Surface 3138 Volume 27775
Surface 3366 Volume 30913
Surface 3602 Volume 34279
Surface 3846 Volume 37881
Surface 4098 Volume 41727
Surface 4358 Volume 45825
Surface 4626 Volume 50183
Surface 4902 Volume 54809
Surface 5186 Volume 59711
Surface 5478 Volume 64897
Surface 5778 Volume 70375
Surface 6086 Volume 76153
Surface 6402 Volume 82239
Surface 6726 Volume 88641
Surface 7058 Volume 95367
Surface 7398 Volume 102425
Surface 7746 Volume 109823
Surface 8102 Volume 117569
Surface 8466 Volume 125671
Surface 8838 Volume 134137
Surface 9218 Volume 142975
Surface 9606 Volume 152193
Surface 10002 Volume 161799
Surface 10406 Volume 171801
Surface 10818 Volume 182207
Surface 11238 Volume 193025
Surface 11666 Volume 204263
Surface 12102 Volume 215929
Surface 12546 Volume 228031
Surface 12998 Volume 240577
Surface 13458 Volume 253575
Surface 13926 Volume 267033
Surface 14402 Volume 280959
Surface 14886 Volume 295361

 

[GRHayes]

lel

Seems fine to me.

First, off my apology. You are correct on the calculation. I forgot the matrix was angular and not strait lined.
That said the radius of the Octohedron is far greater than that of the sphere. You are right though it does have a lower surface area to volume.

 

[JinM]

So this 2 page talk is about what shape is more efficient for turret designs? Because I think the circle is more efficient as design choice.

Just to point my stance to this matter out. But I am not able to mathematically proof this (I don't want to). :D

But I want to point out, that the mathematical approaches you two guys are taking are not suitable:

Yes, the wedged square has more volume with the same surface touching blocks - but it also has a bigger diameter. It's just natural that a equally shaped object of a bigger radius has a bigger volume than that one of a smaller radius.

The radius of the wedged square does not go from its inner point to one of the sides, but to its corners. This is not a practical model to compare effectiveness of booth shapes. If you wanted to compare an octahedron vs a sphere...well either you take the octahedron's corners as point of refference for the radius, or you say the octahedrons virtual radius is the distance from the center to a corner plus the center-to-side distance divided by 2.

No matter what you say, the actual "radius" (and only circles and spheres and stuff like that are having a radius!) of any shape that is not round, is the distance from its center to the farthest point of it. But if you are disagreeing with me on that I think we can't talk anymore. (I say that, because in this forum only the guys who disagree acutally respond and I feel like everyone writes here just to disagree - including me. =) )

I have to add, that this picture would have a totally different block count result, if you turned the square 45 degrees. But that's not the point of your talk guys, I know. I just wanted to remind you.

 

[jayman38]

The problem with hard-corner shapes vs. simulated circles is one of host concavity. As the sub-entity grows in size, you have to allow for the largest "radius". For simulated circles, the radius of the concavity remains relatively close to the circle diameter, plus approximately three blocks. For even-sided polygons, it's the line between opposite points.

The concavity diameter versus polygon flat-to-flat distance becomes larger with size, leading to larger and larger "spaces" and openings that can be exploited by opponents. That's not really a big problem, because the concavity is generally going to be well-protected anyway, and is typically the root entity, which normally has plenty of shielding.

Generally, for larger diameters, while you might get more efficiency with a polygon cross-section, you'll simply get a few more internal blocks with a simulated circle.

If you are building a titan, maybe the concavity doesn't matter, but for part-time builders like me, who only build small, concavity size makes all the difference.

 

[ToddoftheYear]

So this 2 page talk is about what shape is more efficient for turret designs?

Someone was wrong on the internet and I was compelled to correct that. My arguments have nothing to do with building ships or turrets.

 

[NeonSturm]

Wow, the thread goes into perma-math state :D

Cubes have a terrible volume to surface area.

We have not-even cubic tetra packs for drinks in super-markets.
We could save so much waste if we made these at least cubic or give them a hexagonal cross-section! But who cares?

ToddoftheYear GRHayes : The surface is not the visible surface sides, but how many blocks you need to cover it.
3 sides need between 1 and 3 blocks for cover, depending on if you have a concave or flat/convex shape.

ToddoftheYear posted a "perfect shape" for a boxel game volume/surface ratio when you look at "how many blocks to cover it.
But it is "not perfect" if you use the "average visible surface area from all angles" to determine the best shape.

I'd say ToddoftheYear's shape is overall better for turrets, but you could smooth some corners for visual appearance or to reduce the collision volume and turning radius or performance.​

 

[Crashmaster]

It wouldn't be the first time block count has been off in this game.

Pretty sure I remember you being wrong that time too.

Block count is off

 

[GRHayes]

So this 2 page talk is about what shape is more efficient for turret designs? Because I think the circle is more efficient as design choice.

Just to point my stance to this matter out. But I am not able to mathematically proof this (I don't want to). :D

But I want to point out, that the mathematical approaches you two guys are taking are not suitable:

Yes, the wedged square has more volume with the same surface touching blocks - but it also has a bigger diameter. It's just natural that a equally shaped object of a bigger radius has a bigger volume than that one of a smaller radius.

The radius of the wedged square does not go from its inner point to one of the sides, but to its corners. This is not a practical model to compare effectiveness of booth shapes. If you wanted to compare an octahedron vs a sphere...well either you take the octahedron's corners as point of refference for the radius, or you say the octahedrons virtual radius is the distance from the center to a corner plus the center-to-side distance divided by 2.

No matter what you say, the actual "radius" (and only circles and spheres and stuff like that are having a radius!) of any shape that is not round, is the distance from its center to the farthest point of it. But if you are disagreeing with me on that I think we can't talk anymore. (I say that, because in this forum only the guys who disagree acutally respond and I feel like everyone writes here just to disagree - including me. =) )

I have to add, that this picture would have a totally different block count result, if you turned the square 45 degrees. But that's not the point of your talk guys, I know. I just wanted to remind you.

You are correct the farthest point of an Octohedron or any other non-round object is the fartherst point usually a corner. In the case of an Octohedron the radius is easy to determine by the layer count from any corner to the thickest layer.
[doublepost=1490649478,1490649037][/doublepost]

Pretty sure I remember you being wrong that time too.

Block count is off

Nope it was an actual bug where the system wasn't counting all the power blocks put down. T753 Block count and mass is off
[doublepost=1490649715][/doublepost]

Wow, the thread goes into perma-math state :D

We have not-even cubic tetra packs for drinks in super-markets.
We could save so much waste if we made these at least cubic or give them a hexagonal cross-section! But who cares?

ToddoftheYear GRHayes : The surface is not the visible surface sides, but how many blocks you need to cover it.
3 sides need between 1 and 3 blocks for cover, depending on if you have a concave or flat/convex shape.

ToddoftheYear posted a "perfect shape" for a boxel game volume/surface ratio when you look at "how many blocks to cover it.
But it is "not perfect" if you use the "average visible surface area from all angles" to determine the best shape.

I'd say ToddoftheYear's shape is overall better for turrets, but you could smooth some corners for visual appearance or to reduce the collision volume and turning radius or performance.​

It does have a lower surface volume it however has a much larger spacial volume or radial area it needs.

 

[Crashmaster]

Nope it was an actual bug where the system wasn't counting all the power blocks put down. T753 Block count and mass is off

I guess I was confused by the explanation. Thanks.

 

[GRHayes]

Someone was wrong on the internet and I was compelled to correct that. My arguments have nothing to do with building ships or turrets.

Well thanks for correcting me. Sorry I was an ass in response.
[doublepost=1490651378,1490651327][/doublepost]

I guess I was confused by the explanation. Thanks.

No problem. That said I have been wrong in the past on here. I think the last time had to do with power calculations and the changes going on.

 

[KiloZulu]

Well, I like the design.

I'l have to submit and post my ship that uses a similar concept when I get home tonight. As others have posted here, you do have to deal with interference where the turret is blocked by the hull, but that's an issue for any turret. I kept my own ship small and maneuverable so you can keep your enemy out of those dead zones.

 

[Nickizzy]

But the weapons can still be blocked horizontally.

You can build "Satellites". Not only turrets similar to a moon for a planet, but you can also make them wiggle on their flight path to evade pathing prediction.

If they have a huge shield capacity, they can even dive into the fire range to fire pulse-slaved guns and reload behind the ship.

Just off set them 5 blocks up. The actual idea(For me) is to have turrets that can fire up/down and left/right all at once.

 

[GRHayes]

Well, I like the design.

I'l have to submit and post my ship that uses a similar concept when I get home tonight. As others have posted here, you do have to deal with interference where the turret is blocked by the hull, but that's an issue for any turret. I kept my own ship small and maneuverable so you can keep your enemy out of those dead zones.

You are correct it is an issue with any turret even satellite turrets still have areas the hull obscures thus you need them on both sides of a ship not just one.

That said so far from basic testing by spawning a number of pirates in it takes about 30% the time to take them out with these than conventional turrets. I would say there are two main factors that are make that possible. Rather than having two smaller turrets I can have the combined fire power of both those in one weapon. The biggest factor probably is the fact that the turrets can keep tracking the targets as it pass the hull and continue firing. Conventional turrets will continue to track the target and not move onto another target unless it moves out of range so you can literally have your ship surrounded and the turrets be locked onto targets and not a single conventional turret firing. If the conventional turret was to break lock when out of site and go on to the next target it would be an improvement. That said I think this system would still win primarily because the damage level would be greater per hit plus the increased defensive strengths also.