I think it’s fair that I give you a complete explanation of the replacement system and what was the situation before that. The original idea was mine and regurgitated many times with the Council, Schine and players (although without mentioning anything concrete).
I’ve spent a lot of my spare time figuring out this system and gathering information from players. I only ask you to read this post from start to end.
Back in Time
Assume that we don’t have docked reactors, or this replacement system for now. We only have the basic reactors.
Why did we give them a soft cap to begin with?
Something called the inverse-cube law. It’s to do with ship size. The ratio of functional blocks (power/weapons) versus non functional blocks (armor/decoration) is completely different for small ships compared to the larger ones.
Non functional blocks are mostly on the outside of the ship, the hull.
Functional blocks are on the inside, the systems.
A small ship usually has 50% of its blocks as hull, the other 50% are systems.
A “big” ship ends up with about 15% of its blocks as hull or less if there is no armor at all, compared to the 85% systems remaining.
You need power for pretty much anything on your ship, so originally the problem with a linear scale is that small ships just didn’t have enough power to be useful. We couldn’t simply increase the linear regeneration without also buffing the bigger ship. We couldn’t simply change the hull vs system ratio either for large ships or small ships.
The solution to this was to make power scale faster at the start and eventually fizzle out to its normal linear component when you reach bigger ship sizes. The threshold was set to 1 mil because around that size, the ratio was beginning to favor systems more than hull and because at that time, ships of that size were considered big. Anything bigger and you would experience severe performance problems.
This isn’t perfect since different ship shapes have different hull vs systems ratios but it worked for many examples. We made power regen scale even better for people that optimized dimensions per group but that came down to the same thing after a while. You didn’t have 5 equal choices, just 1 really. This was unfortunate but inevitable.
Later this soft cap was increased to 2 mil since Starmade could handle bigger ships better than it did 2-3 years ago. Also making people build bigger.
Docked reactors
When these popped up to bypass the soft cap on your ship, it was a good concept. It wasn’t overpowered, since for every new entity where you reached 2 mil regen, you also needed extra blocks to supply it to the main ship + other system blocks to make sure the reactor was usable.
This made power reactors less efficient than the mothership power reactors. Basically making the mothership’s linear regeneration higher (but adding downsides).
The risk of this system was that more of your power is localized into specific areas of high efficiency. The docked reactors gave away their location (navigation markers, it’s easy enough to figure out where they are) allowing any opponent to focus on them first.
A hit reactor rapidly lots its efficiency due to its power lines getting severed or it losing its logic, computer or docked module.
It was an acceptable downside to a system bypassing the original softcap. It allowed bigger ships to enjoy their extra system to hull ratio even more but they had vulnerable spots where people could hurt them the most.
At some point power supply blocks were nerfed a bit to make their efficiency lower to make it more balanced.
I’ve seen people say that it requires a lot of skill to make power reactors, but it’s not. What you need to know is the concept, then it’s child’s play.
Why?
Because docked reactors actually scale linear. The power supplied by a docked reactor isn’t limited by the power reactors till you hit the soft cap, but by the power supply blocks.
Reaching 2 mil regen requires 2000-3000 blocks give or take, depending on how efficient you can make their shape. 1000 blocks sounds like a big difference in efficiency, till you look at the power supply blocks required to supply it all down to the main ship…6600!
First of all, they provide 0.8 power for every 1 power consumed. You can’t make a reactor that provides more than 1.6 mil regen without bringing your efficiency down.
To provide 1.6 mil you need to have about 6600 blocks power supply blocks. That’s at least twice as much than you need power reactors, that’s not even to mention the few power capacity blocks you need to make sure each tick has enough power in storage and game lag doesn’t cut it off. Add some shields to that since well, extra safety is good. Let’s say about 7000-8000 extra blocks on top of the 2000-3000 power blocks.
The limiting factor on docked reactors isn’t the power reactor group (which scales better depending on the pattern you make them => skill based). It’s based on the power supply blocks that are completely linear (put them down wherever you want and there you go => not skill based). Add a simple logic clock to it and you’re pretty much good to go.
Why reactors are so “unique” is that people just don’t realize this is a way to bypass the power softcap when they play the game for the first few months.
Of course there are variants with 2 computers, alternating fire and eliminating any delay between beams but that’s already accounted for in the block count provided above, it assumes perfect power transfer.
If you put this in a graph, compared to the normal reactors you get this.
X-axis: Block count
Y-axis: Power regen/sec
As you can see, the docked reactors are pretty much linear linear even though the power reactor block itself doesn’t scale linear. It starts going down after 1.6 mil but that’s because you can’t go higher than that without also resulting in linear power regen.
The power reactor curve is only 1 group, but it’s the same curve with multiple groups just widened on the X-axis. That would also make the docked reactor curve slightly less linear but barely noticeable.
This is not skill based. There’s no ultimate shape that makes your special engineered reactors 2 times better than an equal sized one of your opponent.
“Make power supply scale non linear too then”. Why would I do that when apparently power reactors are fine right now, a great skill based addition with tons of depth?
Performance
You’ve all heard the reason why we finally decided to move away from these, lag created by them undocking during combat (or on accident). And no, leaving empty space between your reactor and the ship, or none at all doesn’t improve performance at all.
4 big power reactors in a decent sized ship that can fit them, have them undocked in the middle of combat and the game will have to do a ton of calculations to figure out what to do with those undocked structures inside the ship. It will slow down everyone in that sector (and stall the server) till it figures out they have to go outside of the ship or it will never stop. The reactors phase through the solid blocks of a ship but that makes the fewer collision checks even worse (how else would you know it’s outside of the ship).
When the ship finally poops out the reactors, you’ve gone through 2 minutes of extreme lag spikes and ping increase all around the server. Depending on ship size, the server might never recover and give up entirely.
This is a severe issue, you can only improve physics up to a point. They had to go and looking at this curve, the solution is simple…
Replacement block
The graph shows that that at any given point below 1.6 mil regen, the power regen scales pretty much linear with the block count of a reactor...each reactor has its own softcap…
We replace the docked reactor with an onboard block that has a soft cap per group! Let them scale linear too!
Unfortunately, the linear scale wouldn’t work since that would mean the soft cap wouldn’t work. No group based softcap would mean everyone could spread them out over the ship as single blocks or as small clusters.No one actually uses thousands of docked reactors because it’s tedious to make and dock. We had to stay to the core idea of a few medium to big reactors.
Which means we have to keep the softcap per group idea, and get rid of the linear regen.
The downside of docked reactors is that the group was focused on 1 spot and that it was easy to destroy (once you found it on the ship, see navigation markers), as I mentioned before. Another reason why we couldn’t allow everyone to spread them out since that moves away from the docked reactors.
The middle ground is to make them scale faster till they hit the softcap. Reaching peak efficiency at about 1.9 mil regen/sec. Which they do after about 9500 blocks.
I’m simplifying numbers here since that’s easier to read but currently it’s peak efficiency is 193 power per block in a group. They start at 25 power per block.
Even if you make a line of this, it would be 9500 blocks long which is simply still easy to hit on a big ship, even on accident.
The only downside about this complete system is at this stage...There’s no navigation marker, it’s impossible to know where they are. You could spread them around in non cubic shapes, it would take ages of random fire to bring most reactors down to individual groups where they’re worthless. Or even in cubic shapes, you would never know you’re hitting an important spot.
And here comes the explosive part. Because they’re part of a group it’s easy to make them scale properly with randomized explosions. The best thing about this is that it promotes you to keep them more in a small dimension group than spreading them out as long lines. The explosion power doesn’t decrease, and it will would create a lot more system damage if 50 explosions went off along a long line instead.
Why do they have internal storage cap? Because power reactors also had a battery variant. These blocks were 90% power capacity and a little regen. They were used in bursts, to give your power capacity back to 100% after firing your first alpha volley. After that they were pretty much useless for a while till they were full.
These internal storage ones, scale worse at the start compared to the normal power capacitors, but they do scale faster. Going beyond the reactor size of peak efficiency could give you a large internal storage at the cost of a bigger to hit target and less efficiency. The explosive power does become weaker past the peak efficiency so really big reactors wouldn’t lose that much from 1 explosion event.
Its internal regen is probably too high but those are values we can tweak along the way. The important part is that we get the functionality of 2 types of reactors into 1, and the build style determines which one is going to be the most effective.
At the end, what we get is:
- A system that promotes keeping groups into smaller dimensions (they look like docked reactors)
- A system that has a better efficiency than linear regen (better efficiency than actual docked reactors even)
- A downside that makes them lose efficiency when hit (docked reactors share the same disease, except that they are more easily found).
- Little to no performance hit. Explosions are set at the start of an explosion event, reducing explosion impact (which is negligible in small radiuses).
Now before you get worried about the explosive system…
It’s scaled in such a way that the explosions happening don’t take out much besides the reactor group itself.
The explosions are numerous but localized. They only happen in the group.
The explosions happen every 2 seconds.
A single hit to a reactor would trigger an explosion event. It would bring its regen down from 1.9 mil back to about 100K in best case scenarios in about 100 seconds. Meaning that a hit reactor is still more efficient than linear power regen for at least a full minute. (It actually is always better than a linear reactor)
Not only that, but if you put armor layers inside the reactor, you’ll reduce the localized explosion damage to the reactors a lot. A simple sandwich armor system increased its mass by about 50%, but after the explosion event it had 500K power regen left. Still an efficiency of 114 power per block!
- How you shape the reactor will reduce the chance of it getting hit from a certain angle.
- How you armor the exterior of the reactor (more at the front facing side) will determine how lucky someone has to get to trigger an explosion event.
- How you put armor inside the group determines how much you have remaining after it gets hit
- How deep or where the reactor is located can reduce the chance of it ever being hit
- Using a few big reactors, or many smaller (but more spread around) reactors will change how fast you can lose your entire power regen during combat.
The system looks simple from a distance, but you can actually get a decent amount of variation without having to teach the player all about logic and power supply beams.
Now it’s true that at this point, modular ship design is discouraged/eliminated without power supply beams.
The solution to that seems simple though, with this separate block we can just have their regen inherit up and down chains without a problem. As long as we never do that for the basic regen there’s nothing that circumvents the initial non linear growth of the basic power reactors since it’s per group anyway, not per entity.
It’s certainly something we can look into for next release.
I think that’s about as far as I’m willing to take it here. I can make a bigger post on its own in a separate thread if that is required.
For now this is scary, and for some a punch in the stomach for all the time you’ve spent on figuring out docked reactors. Not only I believe that this will work, but so does the council and all of Schine.
The config values may require tweaking, but that’s where your input comes in. It’s important that you’ve used the new block and experimented with it enough to grasp the positives and negatives. Only then can you see that it’s either not working as it should, or is just fine as it is.