Mass-based turning system

[Planr]

Currently on the game, we have a ship turning system that functions solely based on how long, wide or tall the ship is. This system is okay for now as the game is still in alpha, but it has the major drawback of encouraging building borg cubes to hit a sweet spot in the conflict between ship maneuverability vs ship size.

In a nutshell, it means a 100,000.0 mass borg cube that is 400 meters long turns just as easily as a 50,000.0 mass 400-meter Star Trek ship. It also means that a 400-meter-long pole of hull blocks in every XYZ direction going out from a ship core, with a total mass of 239.5, turns at the same rate as that 100,000.0 mass borg cube. I don't care what anyone says, that's just stupid, plain and simple.
There is also a very low ship turn rate cap around 125 meters in every direction, which just encourages building really small and compact warships that leave little room for aesthetics if one is 100% serious about maximum combat efficiency.

I don't know about you guys but I think this is totally unfair, and we need a better way for the game to determine how easily a ship is able to turn.

The answer is simple: Have the ship's turn rate be based on mass.
The first question of course then is, how do we determine how the ship is able to turn on a particular axis based on just its mass? Well, you can't do it with just that really (I mean you could obviously, but it'd have to be distributed evenly along all lines).
So, to solve that problem, the ship's dimensions would still play a role in the turn rate, but instead of directly limiting it based on how long or wide the ship is, the dimensions as a whole act as ratios to determine how the turn limitations of the ship are distributed. For example, a ship with 50,000 mass that is 400 meters long and 200 meters wide and 100 meters tall would have 400/700 (or 57%) of its turning limitations distributed to the length-based turning axis, and the rest distributed accordingly to the other dimensional turning axes.

Basically this means that two ships with the same box dimensions, but one heavier than the other, would not turn the same rate, but the ratios of their turning abilities are proportionate.

In order for this to work in a game-appropriate fashion, a softcap then has to be set on how the ship's mass reaches toward the minimum maneuverability point of a ship. Currently the game has a relatively hard hardcap on this in the current system, with a 300-meter battleship turning at the same rate as a 1000-meter super-titan. I'm sure i'm not the only one who doesn't like that, right? This is why there should be a slowly-decreasing softcap on the turn rate as ships get bigger and bigger. I don't want it to make all ships beyond a certain size turn the same rate, but at least have a very slight and minimal difference in their turn rate to each other, based on their mass.

On this system, it will allow people of all different sci-fi franchise preferences to build the ships they want with a fair universal maneuverability mechanic for their ship. Those lighter but bigger-box-dimensioned star trek ships would now get the maneuverability they deserve, as opposed to turning like they weigh thousands of mass units more.

For those who understand what I mean here, please let me know what you think.

 

[ltmauve]

So, to solve that problem, the ship's dimensions would still play a role in the turn rate, but instead of directly limiting it based on how long or wide the ship is, the dimensions as a whole act as ratios to determine how the turn limitations of the ship are distributed. For example, a ship with 50,000 mass that is 400 meters long and 200 meters wide and 100 meters tall would have 400/700 (or 57%) of its turning limitations distributed to the length-based turning axis, and the rest distributed accordingly to the other dimensional turning axes.

Erm, what do you mean by "around the length-based turning axis?" If the length is the longest axis (Say, 500m long and 1m wide and high), then turning about the long axis will be the fastest.
Instead it should be (x + y)/(2 * x + y + z) around the z-axis (the two is there because of each dimension being used twice)

 

[Planr]

Erm, what do you mean by "around the length-based turning axis?" If the length is the longest axis (Say, 500m long and 1m wide and high), then turning about the long axis will be the fastest.
Instead it should be (x + y)/(2 * x + y + z) around the z-axis (the two is there because of each dimension being used twice)

When I say "around the length-based turning axis", I am referring to the axis of the ship's turning that is affected by how long the ship is currently. I think it's from side-to-side rather than rolling or up/down.

 

[jayman38]

It would be interesting to do a calculation study, to see how long it takes to calculate the turning moment of 64,000 blocks placed all at once in fully symmetrical, 10-block-in-all-axes placement. It shouldn't be too bad. Each placed block will have something like the following 6 calculations. All numbers and variable names are placeholders, subject to change and flaming argument.

1. temp_x_moment = abs (ship_loc.x) * abs (block.mass) / 1000
2. temp_y_moment = abs (ship_loc.y) * abs (block.mass) / 1000
3. temp_z_moment = abs (ship_loc.z) * abs (block.mass) / 1000
4. ship.turn_rate.x = ship.turn_rate.x - ( temp_y_moment ) - ( temp_z_moment )
5. ship.turn_rate.y = ship.turn_rate.y - ( temp_x_moment ) - ( temp_z_moment )
6. ship.turn_rate.z = ship.turn_rate.z - ( temp_x_moment ) - ( temp_y_moment )

I threw the block.mass into the equations in case different blocks get different mass, and used the "abs" function to get the absolute values for everything, not only to get the absolute distance from the origin of the ship, but also in case some wiseguy tried to set a negative mass on popular blocks.

I guess I could have used more advanced math to get the combined vector of the two moments to adjust the turn rate, but square root calculations are hideously computationally-expensive.

 

[ltmauve]

When I say "around the length-based turning axis", I am referring to the axis of the ship's turning that is affected by how long the ship is currently. I think it's from side-to-side rather than rolling or up/down.

Well, both the yaw and pitch are affected by the length of the object. Only when rolling does the length not matter.[DOUBLEPOST=1411255029,1411254942][/DOUBLEPOST]

I guess I could have used more advanced math to get the combined vector of the two moments to adjust the turn rate, but square root calculations are hideously computationally-expensive.

Actually, since you need the distance squared for moment-of-interia calculations, it's much easier.
Xmoment += mass * (y * y + z * z);

 

[FunnyBunny14]

I built a ship with a huge cross shaped reactor while testing the jump drives. It had 16 mass and it turned as slow as you would expect a 1000 mass ship to turn.
Totally agree.

 

[NeonSturm]

Agree. Even suggested this earlier.

Would this work?

Roll = (1- (High + Width) / Total)^2
Pitch = (1- (High + Length) / Total)^2
Yaw = (1- (Length + Width) / Total)^2​

The only issue I see is that optimized ships with the new thrust system will have to be very Tall or very Wide to rapidly turn in the opposite direction if they want to keep the ability to escape quickly.

 

[Malacodor]

If I understand correctly your system still discourages antennas which add to a ship's dimension and favours bulky shapes. Also, I think it's better to limit angular acceleration instead of angular velocity.

 

[jayman38]

If I understand correctly your system still discourages antennas which add to a ship's dimension and favours bulky shapes.

If mass is changed per block, things like antennas and light sticks will be super-low mass, so the mass multiplier will be almost nothing. so their reach into extended dimension will not matter much, so antennas won't end up adding much to the turning reduction as much as they now do.

 

[Planr]

If I understand correctly your system still discourages antennas which add to a ship's dimension and favours bulky shapes. Also, I think it's better to limit angular acceleration instead of angular velocity.

The impact of having antennaes would still be extremely minimal; it would be contributing barely any extra mass to the ship so it wouldn't be affecting the turn rates much at all, plus it would only change the turn rate penalty distribution for the Z-axis of the ship by only a tiny amount (that's assuming your antennaes are pointing forward).

 

[lemdoran]

i also think a rewamp can be good, as some pretty long but not so big ships can ba so painfully to turn that you better do a big cube to preserv the maximum combat efficiency, with a lack of style and beaty...

Making mass the determinant factor will make more interest to classy - long but light ships.

 

[NeonSturm]

With a ship >2km minimum radius, you can make a distance-holder like:
H . . . H . . . H . . . H
. . T . . . . . . . T . .
. . . . . T . . . . . . .
. . . . . . . . . . . . .
H . . . H . . . H . . . H
. . . . . . T . . . . . .
. . . T . . . . . . . . .
. . . . . . . . . T . . .
H . . . H . . . H . . . H
. . . . . . . . . . . . .
. T . . . . T . . . . T .
. . . . . . . . . . . . .
H . . . H . . . H . . . H
And use free floating drone-sized sniper turrets, as evasion with passive push is a viable strategy again ((as your core is just out of range))

 

[Planr]

With a ship >2km minimum radius, you can make a distance-holder like:
H . . . H . . . H . . . H
. . T . . . . . . . T . .
. . . . . T . . . . . . .
. . . . . . . . . . . . .
H . . . H . . . H . . . H
. . . . . . T . . . . . .
. . . T . . . . . . . . .
. . . . . . . . . T . . .
H . . . H . . . H . . . H
. . . . . . . . . . . . .
. T . . . . T . . . . T .
. . . . . . . . . . . . .
H . . . H . . . H . . . H
And use free floating drone-sized sniper turrets, as evasion with passive push is a viable strategy again ((as your core is just out of range))

in just what way does this even remotely relate to the thread...

 

[Malacodor]

The impact of having antennaes would still be extremely minimal; it would be contributing barely any extra mass to the ship so it wouldn't be affecting the turn rates much at all, plus it would only change the turn rate penalty distribution for the Z-axis of the ship by only a tiny amount (that's assuming your antennaes are pointing forward).

If I understand correctly now, such an antenna would increase the turn rate about the Z-axis, since width and height become relatively smaller as the length increases without adding significant mass. This system favours antennas extremely.

 

[Commodore]

I would suggest that the turn rate uses a logarithmic formula such as: http://www.wolframalpha.com/input/?i=log-linear+plot+2%2F%28ln%28x%2B1%29+%2B+2%29%2C+x%3D1+to+1000000

Note that the vertical axis is not shown getting to zero. This kind of formula makes a meaningful difference in turning rate between ships at all scales. A 750,000 block ship would be able to turn inside a 1,000,000 block ship, providing a bit of a counter for the "more mass = win" rule.

I think cube ships should be discouraged, meaning that while dimension based penalties are realistic, we're better off without them.

 

[RedDragonflyxx]

This looks ok but I don't think its a good solution. I think it should be based on the actual amount of mass being moved and use that to affect acceleration. That way the antenna's length doesn't matter at all, just the distance it is being moved and the mass. I'm not going to look up the physics behind this tonight but I think you get the picture: The closer you're blocks are to the center of mass the less energy is needed to move it.

 

[aceface]

@Planr mass does affect turning speed just not as much as before.

 

[Planr]

If I understand correctly now, such an antenna would increase the turn rate about the Z-axis, since width and height become relatively smaller as the length increases without adding significant mass. This system favours antennas extremely.

No, you misread greatly. The lengths in each direction for XYZ determines how the turnrate drawback pool determined by the ships mass would be distributed. Adding more blocks along the Z axis would divert more of those drawback numbers for turnrate from the other dimensions, meaning your ship would turn slightly slower along the Z-axis and faster along the X and Y axises.

Of course there will also need to be additional regulating equations for discouraging people from building needle or wall-shaped capital ships just so they can turn really fast in a particular direction. I was thinking the efficiency of dimensional ratios vs distribution would be a particular factor. If one dimension of the ship is significantly longer than the others, than the drawback percentages for that long dimension would steadily decrease in terms of percentage returns.

This looks ok but I don't think its a good solution. I think it should be based on the actual amount of mass being moved and use that to affect acceleration. That way the antenna's length doesn't matter at all, just the distance it is being moved and the mass. I'm not going to look up the physics behind this tonight but I think you get the picture: The closer you're blocks are to the center of mass the less energy is needed to move it.

I'm not sure how this relates to the thread, as you're talking about acceleration and the energy required to move the ship. I'm talking about simply turning the ship.

Also, the system you are talking about sounds good in theory but it sounds like it would require an immense amount of constant calculations

 

[Malacodor]

The closer you're blocks are to the center of mass the less energy is needed to move it.

Also, the system you are talking about sounds good in theory but it sounds like it would require an immense amount of constant calculations

The amount of calculations should actually be manageable, as explained here.

Adding more blocks along the Z axis would divert more of those drawback numbers for turnrate from the other dimensions, meaning your ship would turn slightly slower along the Z-axis and faster along the X and Y axises.

This doesn't make sense at all, with more blocks along the Z-axis a ship should turn slower about the X- and Y-axis, not faster.

 

[ltmauve]

Adding more blocks along the Z axis would divert more of those drawback numbers for turnrate from the other dimensions, meaning your ship would turn slightly slower along the Z-axis and faster along the X and Y axises.

Um, no. The Z-axis acceleration depends on the mass distribution in X and Y. The Y-axis acceleration depends on the mass distribution in X and Z. The X-axis acceleration depends on the mass distribution in Y and Z.
The Z-axis rotation is roll. The Y-axis rotation is yaw. The X-axis rotation is pitch.
Think about it. If you rotate weights (like the exercise kind) along the bar, it doesn't matter how close together or far apart they are. If you rotate it perpendicular to the bar, it does matter where the weights on the bar are.