- Joined
- Apr 30, 2015
- Messages
- 22
- Reaction score
- 10
StarMade Mandatory Ship Aesthetics rehaul.
Hi everybody. Me and several of my friends were playing SM for a while now, pointing out problems with it and discussing possible solutions for them. In this post, i want to show some possible things that i think are wrong with SM now and to adress some ways of fixing them.
Me: i'm a 5th year student (doing my 1-st year Master's studies) of CS, game enthusiast since 2000; was playing competitively in mmorpgs, fps, racing games; doing stuff in computational math (plasm modelling) and linked to it algorithm optimizations.
Disclaimer: written in an airport, extremely long.
Sadly, it's not TLDRable either.. The best possible TLDR that can be made by human being was posted by @Tobie.
I wish some of you will stay with me through this text.
Also note: i do not elaborate on the "optionality" of this system in this post. Game mode, config -- those are all options, but this post is long enough.
Ability to change the way some checks are implemented requires scripting of sorts, and admin-defined scripting inside of "on hit" event handlers can be devastating for perfomance. Thus, ability to configure everything is a suggestion in itself, and it can be quite tricky to implement.
I see little beauty in a ship whose hull seems smooth only because it's 1500 blocks wide. It's one thing if effort went in this sleek hull and thoughtfuly placed wedge lines that not only hide corners but also take shape of a penis, but it's completely different when you just mashed blocks around until it looks smooth through squinted eyes.
Another problem with gigantism is that it's okay for EVE to have titans, because there are 40 thousand people online every second to swarm it with merlins. On a SM server with an average online of 50 a guy with a kilometer-long rogue titan with 60 turrets and 2200 drones will be unstoppable by the rest of the players combined. In the end there'll be three guys meeting on thursdays to check who's got the biggest gun out there, and the rest will be annihilated on the spot.
Now i will talk about the design rehauls i want to propose. I will talk about weapon systems, shield systems, combat system management, drone systems and maybe something else i come up with. Please note that any of those proposals wont work with current shields/weapons/scaling factors, so it's kinda pointless to assess the effect of an isolated design update.
Well, i have a flight in an hour, gotta get up and go. I hope what i wrote will be of some use and that at least overall "mildly forcing into obeying" idea will find its use during the design of this game. I would like to hear about weak points of my ideas and of their possible solutions, as i understand that no idea can be without flaws -- but i hope the collective mind of this forum will help fixing those.
Hi everybody. Me and several of my friends were playing SM for a while now, pointing out problems with it and discussing possible solutions for them. In this post, i want to show some possible things that i think are wrong with SM now and to adress some ways of fixing them.
Me: i'm a 5th year student (doing my 1-st year Master's studies) of CS, game enthusiast since 2000; was playing competitively in mmorpgs, fps, racing games; doing stuff in computational math (plasm modelling) and linked to it algorithm optimizations.
Disclaimer: written in an airport, extremely long.
I wish some of you will stay with me through this text.
Also note: i do not elaborate on the "optionality" of this system in this post. Game mode, config -- those are all options, but this post is long enough.
Ability to change the way some checks are implemented requires scripting of sorts, and admin-defined scripting inside of "on hit" event handlers can be devastating for perfomance. Thus, ability to configure everything is a suggestion in itself, and it can be quite tricky to implement.
- Ship design is not a step you have to spend time on. People are not encouraged to build aesthetically pleasing ships. Flying box always will be functionally better. It's not that i think people should spend several days just planning their ship: it's that i think that people who do spend this time should have better ships. Right now any aesthetics is counter-functional.
- Combat is too shallow. In current system, there are no ship specializations and they can not show up on their own.
- Current system encourages gigantism.
Another problem with gigantism is that it's okay for EVE to have titans, because there are 40 thousand people online every second to swarm it with merlins. On a SM server with an average online of 50 a guy with a kilometer-long rogue titan with 60 turrets and 2200 drones will be unstoppable by the rest of the players combined. In the end there'll be three guys meeting on thursdays to check who's got the biggest gun out there, and the rest will be annihilated on the spot.
- Aesthetically pleasing ships.
Being an engineer myself, beauty of the ship for me is visible amount of effort that went into its building. X-wing replica? Check. EVE ship replica under 200 blocks length, built by hand and carefuly adjusted to look sleek by smart use of wedges and pentas? Hold my beer sir, i have to write an appendage for my testament. A script that generates ship blueprints from a 3ds file? Prometheus you are, giving fire for the people in need. USS Enterprise 4000 blocks wide, built by a script? Errm. Huh. I guess. No, it's awesome if you're still in school, or if you learned scripting or 3d modelling because of it, if only just basics. But i will never rate such ship higher than a 9x9x9 handmade drone. - A way to reward skilled players in a combat.
Sure, anybody can play anything, and we do not want this to turn into another EVE. What i mean is that there have to be some combat moves besides repositioning, turning and shooting. - Degrading functionality during combat. I want to be able to damage other people cannons or go for their engines, and to do so knowingly. This means that somehow the game has to make ship design mandatory - i should be able to see where are the engines on your ship and where are the cannons or drone bays.
- No giant unstoppable juggernauts of destruction, armed with every system in the game. This is the most vague of all of my wishes, but it is the only one that can be fixed through number tweaking.
Well, by mildly forcing people into obeying these rules. Note the 'mildly' and 'forcing' parts, they are the key of this post.
Why do i feel people should be forced? Well, because any MMO is a competitive game. Even in PVE of WoW there were in-raid fights about who's got the bigger DPS, and in PVP part of it or of any other game people won't build a beautiful thing if it's less powerful than Brett's Assholination Cuboid. Anyone building a beautiful ship will have to be either protected by the cubes or destroyed by them; and only way to have a beautiful yet useful ship is to either abuse OP weapon (kh khh heatmissiles kh kh automated circuit kh kh) (btw awesome design! as a software engineer i rejoice at the sight of an automated abuse :[*]or abuse environment (said heatseekers ignoring the only ship around the user that actually damages him) (btw great tactical thinking!).
Why do i feel people should be forced? Well, because any MMO is a competitive game. Even in PVE of WoW there were in-raid fights about who's got the bigger DPS, and in PVP part of it or of any other game people won't build a beautiful thing if it's less powerful than Brett's Assholination Cuboid. Anyone building a beautiful ship will have to be either protected by the cubes or destroyed by them; and only way to have a beautiful yet useful ship is to either abuse OP weapon (kh khh heatmissiles kh kh automated circuit kh kh) (btw awesome design! as a software engineer i rejoice at the sight of an automated abuse :[*]or abuse environment (said heatseekers ignoring the only ship around the user that actually damages him) (btw great tactical thinking!).
Weapons have to be cuboids. I believe that cannon >::::::::::::::< is visually recognizable as one and is aesthetically pleasing, and cannon > ::::...< is a very tiny weenie on a very big nutsack.
Why cuboids? Because this is the only computationally easy shape to check against. Because it allows for easy and fast calculation of its dimensions, which are used below. Also, thanks to weapons being cuboids it's possible to mark the whole side of a weapon as its output, so now you can fire 2x2 cannonballs and giant rockets.
Geometry of the weapon system should affect it's characteristics, but all the weapons are broken in their own ways as they are now, so here's what i'd change:
How would geometry of a weapon affect it? Well, for example (effects can be different than proposed, this is just to illustrate the point of my idea):
Now in this system we have:
Also note: i do NOT believe in sublinear scaling of weapon blocks. People will just start utilizing double-chessboard pattern that is now used by all useful salvagers.
Clarification for interaction of such system with slave/effect systems can be found here
Why cuboids? Because this is the only computationally easy shape to check against. Because it allows for easy and fast calculation of its dimensions, which are used below. Also, thanks to weapons being cuboids it's possible to mark the whole side of a weapon as its output, so now you can fire 2x2 cannonballs and giant rockets.
Geometry of the weapon system should affect it's characteristics, but all the weapons are broken in their own ways as they are now, so here's what i'd change:
- AMCs have unlimited range (well, limited, but by server settings and the despawning heuristics, not gameplay-wise)
- AMCs have scatter, so you still wont hit anything past certain distance except by chance.
- Missiles have finite flight time and turning speed
- Beams have 'defocusing' in the sense that their damage falls off as a function of distance, but they can be active for pretty much all the time
- long AMC has less scatter (and effectively longer range because of that)
- wide AMC has more damage
- long missile is faster and has larger range
- wide lockon missile has larger locking ellipsoid
- long beam has less defocusing (reliable long-range damage)
- wide beams are easier to aim and affect larger surface (duh)
- very long AMC sniper cannons that just won't let their hosting ship turn fast enough to counter fighters (think those ships railguns). Yet those railguns are hard to hit from afar, as they have low signature (think arrow pointing at you) and can move sideways quite fast (but not turning)
- powerful, but short guns that do not increase size of the hosting ship enough to slow its turning speed (think those ships fighters)
- longrange rocket destroyers that can keep several fighers at range, but a larger group of fighters can dodge/knockdown some of the rockets and make it into close range, where fighters will have too much orbiting speed to be locked on with a small 1x1output reticle
- shortrange rocket destroyers that will be destroyed by the railguns, but have easy time locking onto fighters thanks to giant reticle of their 10x10x1 rocket arrays
- wide beams for close-range fighter and drone suppression
- long range beams for sustained but low damage, useful for slowly burning through titan's hull
- Motherships with Loads Of Drones That Are Balanced. More info lower below.
- recognizable weapon silhouette. I can selectively try to destroy your weapons now, so you are less of a threat to my fleet, instead of just randomly blasting in your general direction or core-drilling with beam turrets.
- Core-drilling disappeared! With long range sniper rifle you will have hard time aiming for exact same point over and over; with a short range rifle you will have scatter negating drilling; with a long beam you have low damage and with a wide beam it's damage is distributed around.
Clarification for interaction of such system with slave/effect systems can be found here
My main problem with shields is that right now people do not care where are they. Shields are usually what people fill holes with, much like sealant. Or some people (including myself) are just pile them up in some corner of the ship and let them collect the dust here.
So, we have to force people into distributing their shields around the ship.
What this will achieve?
So, we have to force people into distributing their shields around the ship.
- Make shields only affect neighbouring blocks. Say, only those who have Hamming metrics lower than 10 starting from specified Output block, similarly to how weapons have output blocks.. Or 10^15: numbers all need tweaking.
- If a block is hit, it looks up what shield groups affect it. This lookup can be cached, as it only changes when either the said block or shield block of said group is destroyed. The damage is distributed across these shield groups.
- There'll have to be a ship hp hud element: say, simplified transparent rotating silhouette of the ship, with shield groups highlighted. The more damage group sustained, the redder it is. Full shields are green, for example.
- regeneration ratio is split across damaged shield groups.
- You now have to think where to place shields and which systems to protect first
- You can selectively bombard other ship's engines. Or cannons. Or cabin. Or drone bays. Or... whatever you want.
- You can defend against a focus. Building a drone ship? Better reinforce those bays. Freighter? Who needs engines when warpdrive is charging. Destroyer? I have 20 missile silos, but destroying my engines will make me unable to catch up to enemy snipers, better protect those thrusters.
This part is quite simple:
What this is achieving?
- drone control system with drone control system units.
bigger drone control - more "allowed weighted collective range". For example, AWCR of 1500 allows for 5 drones 30tonnes each, 10km far away from the ship or 10 mass-15 drones 2.5 km away. - You send more drones away? Watch out for reactor. Those antennas are not for looks: the farther your drones are and the more of them are out, the bigger the power usage. Also if your drone system is damaged, your farthest drones deactivate, so you have to protect the carrier somehow.
- preemptively nerfs giant 2000-droned carriers.
- opens niche for anti-fighter drone ships that are protecting sniper ships.
All systems have some sort of "best-shape". This "best shape" should provide mechanical benefits to the group, or it won't be used. Your ship is nice? Well mine is powerful. A single Brett guy on a server will either make other people leave or force them into same brickships as his own.
For example, powercaps can be forced into cylindrical shapes using following two rules:
What this achieves:
Power generators:
What this achieves:
Engines:
other systems can have similar notes, and those were used only to illustrate the idea of forcing players into certain "recognizable shapes" for their systems.
But wait, i've just created a monster! With shields that have to be distrubuted, players will try to minimize surface area. So we will have Brett's Assholination Spheroids roaming the skies while last of humans are racing to death on flying carts through the deserts and canyons of a certain planet in a binary system.
How to fix? Relatively easy.
-= Intersystem interference =-.
Every system decreases its functionality for every block of other systems in its bounding cuboid, which for this check is treated as increased for every several blocks of the system. This is computationally easy, cached operation that needs recalculation only on block destruction; the cache is held by the ship rather than by the block so it takes little memory. Block destroyed? Lookup which group it belonged too, lookup systems that were intersecting with said group, then re-evaluate their power. Seems quite possible even for large ships and populated servers.
What this achieves:
For example, powercaps can be forced into cylindrical shapes using following two rules:
- power of a group increases linearly in block count
- power of a group decreases with surface area, at a speed greater than linear (x^2? x^1.5? I don't know: it needs tweaking. Maybe not a polynomial at all?).
- ships will have to feature distinctly visible power capacitor banks, or they will have to provide excess shielding and armor plating to hide them.
- bigger ships will resort to multiple power storages rather than making single giant one, as starting at some point growing surface area will be voiding all the extra power from your powercaps in the same group.
- power of a group increases linearly in surface area
- power of a group decreases (linearly or somehow) in overall spatial size (sum of dimensions).
- no more giant long power-rods in the middle of the ship
- complex reactor designs
- the closer they are to the edges of the bounding cuboid of the ship, the bigger ship's turning rate along the axis those engines are acting around.
But wait, i've just created a monster! With shields that have to be distrubuted, players will try to minimize surface area. So we will have Brett's Assholination Spheroids roaming the skies while last of humans are racing to death on flying carts through the deserts and canyons of a certain planet in a binary system.
How to fix? Relatively easy.
-= Intersystem interference =-.
Every system decreases its functionality for every block of other systems in its bounding cuboid, which for this check is treated as increased for every several blocks of the system. This is computationally easy, cached operation that needs recalculation only on block destruction; the cache is held by the ship rather than by the block so it takes little memory. Block destroyed? Lookup which group it belonged too, lookup systems that were intersecting with said group, then re-evaluate their power. Seems quite possible even for large ships and populated servers.
What this achieves:
- BAS are now required to leave a lot of space in between systems; and the bigger are those systems, the bigger gaps are required. If a lot of space inside the sphere is wasted, it loses its "useful volume"; and with that, it's no longer the best shape for a ship. Now people will try to build long ships or wide ships or multi-hull ships.
- drones can still be densely-packed, as for small system it's interference cuboid is the same as its bounding cuboid.
- with systems now not all mixed in, and with gaps between them, if you know the design of a ship your enemy is flying, you can try to destroy known subsystems. Also you know some "basic placements", such as "engines are on the sides", "long guns shoot far", "long ships turn slow", so you can to some degree assess the ship functionality by the looks alone. In clan wars, this allows for purposeful misinformation by making underoptimal, yet unrecognizable ships. Say, what if i build a long ship with a giant cannon in the middle? You'd think i am a sniper; but as you got nearby i'd launch drones at you as you forgot to check my bottom for drone bays. Reconnaisance and industrial espionage are relevant now.
CSM right now does not exist and power system is just meh, i cant even elaborate on that. So what do i want to change?
What this proposal achieves?
- All systems can now be toggled on and off. If system is online, it drains MASSIVE amount of energy constantly. Amount drained depends on the size of all togled on systems of the same type (so people won't start making chessboard-pattern weapons), and power consumptions grows at rate greater than linear (N log N may be okay).
- Any online system can be toggled off for no cost and instantly.
- Any offline system can be toggled on. Depending on the size and type of system, its activation time differs. Say, the larger the cannon - the longer it toggles on (up to several minutes for titans), but enormous warp drive can be toggled on much faster than barely operating on. Well, at least if you have enough power for the warpdrive to function at all, as it should be quite prohibitive to have WD and anything else active at the same time, unless the ship is specifically fitted for that role.
- Power usage during the actual usage (shooting / warpjump / cloaking) is significantly decreased.
- If for some reason (damaged reactors or full volley with engines on) the power for actual usage is not enough, system immediately goes offline (with the need for reactivation).
- Extra gameplay depth to combat. A lot of people on this forum want to see warp drive disabled for some time after shots were fired, but i believe that's sort of artificial limitation. Actually, i do not believe in quantitative limitations, as they feel fake and do not add depth to the gameplay. Consider:
Bob farms pirates on asteroid belts. Josh-the-PVP-God jumps in. Bob has guns active, but decides to run away.
But what if 30secs into warp drive activation Bob notices that Josh's ship is actually vastly inferior to his own?- old/"forbidwarp": Bob decides to jump away. He tanks Josh for 4 mins and jumps away.
- activated systems: Bob decides to jump away. He tanks Josh for 4 mins and jumps away.
Proposed system significantly increases the cost of "redirecting power". It's still a viable combat maneuver, but now the decision "Disable cannons, get outta here" weights much more. You can not back down on your decision in a large enough ship, you have to plan ahead of time. In the old system, decisions are significantly cheaper. I believe that a rush of emotions during the decision making process contributes to the game, and what's better way to influence decision making if not by increasing responsibility of player for his decisions?- old: Bob starts shooting
- activated systems: Bob deactivates his jump drive and turns cannon on. It will be active in 3 minutes. Or he can disable its activation and activate jump drive again. In 4 minutes, as he resetted the process.
- Ability to balance jump ships and cloak ships. Now i can make a ship with a very fast-charging jumpdrive, but due to enormous jumpdrive energy requirements i will have to ditch shields, weapons and basically everything. First guy with a warp scrambler i run into wont even notice he hit something until he checks his inventory and discovers a whole bag of my goodies. Same goes for cloakers: i believe it's possible in such system to create a permacloaked ship that will slowly move into enemy positions, hide behind some moon, deactivate cloak, deactivate engines, activate heavy antenna array and scan the hell out of them so his mates can plan attack. Or maybe shieldless, yet cloaked drone reconnaisance ship which uses multiple drone-based scaners, hiding himself from enemy radar while her engines are offline? Seems quite OP, until somebody with a radar of his own notices an invisible ship with no weapons and shields. Oh, of course he can turn them on! But it will take precious time which he does not have; during which he will have to sit with no invisibility and no shields and no engines. Enemy can ram him into sun and he won't be able to do a thing. But that's the cost of specialized ships, and i think it's great.
- combat depth
- power management
- targeted damage
- mandatory ship design
- systems now have to be visually distinguishable, at least on a blueprint
- gigacannon impossible because of NlogN power consumption growth
- checkerboard cannon is not a loophole for power consumption, as it is collective size of all simultaneously active missiles/cannons/beams that matters, not size in a group
- drones
- their behaviour configurations can be easily linked to their mothership, thus preventing balance issues of drone swarms.
- system flexibility: because of their now-centralized nature, mothership-based effects can be added (increased shields, partial energy transfer, reduced cooldowns, autorepair while in dock, name it).
- existence of drone control system allows for drone blueprints being stored in the mothership, so it can automatically build drones and recall them after the fight
- ship roles
- fighters are good versus everything but drone ships and destroyers
- long range cannons are good versus big, slow-moving targets
- destroyers who have weak defences and low speed, but are able to fight off fighters and big targets depending on their missile configurations
- drone ships that stand ground againist fighters
- titans that require BIG reactors for their large guns (yet low-power vessels just wont have enough power to fire big cannons, so if you want to have a big cannon - you have to build a titan).
- preemptive balance tweaks
- spherical ships are not advantageous over long ones
- drone spam is limited because of drone system power consumption
- players are not able to circumvent cannon power consumption rate by making several small cannons (because power consumption is based on the total amount of blocks across all groups)
- core drilling is prevented because cannons now have scatter and beams defocus over long range
- a number of balance issues that i quite probably overlooked. But i think such system is still better than the one currently implemented.
Well, i have a flight in an hour, gotta get up and go. I hope what i wrote will be of some use and that at least overall "mildly forcing into obeying" idea will find its use during the design of this game. I would like to hear about weak points of my ideas and of their possible solutions, as i understand that no idea can be without flaws -- but i hope the collective mind of this forum will help fixing those.
- I got home and fixed formatting however i was able to. Sorry for everybody whose eyes fell victim to it. I guess you can ally with those geometry guys out there who are after me :D
- Clarified how shield metrics are calculated, under "shields". Added reference to clarification of how proposed system interacts with the concept of slaves/effects. Added reference to the best possible TLDR ever, thank you very much Tobie .
- Added "optionality" part in the disclaimer.
Last edited: