A revelation on the redundancy of System Integrity

[Scypio]

What am I missing? What have I 'forgotten?'

reducing systems profiles to the minimum possible so they can't be hit.

I suppose i misunderstood you here then.

 

[dwwojcik]

Polish cavalry charging German tanks

For the record, they never actually did this. The Polish cavalry charged and successfully routed a unit of German infantry to cover the withdrawal of other Polish forces. They never faced tanks; they only were forced to retreat once German armored cars with machine guns showed up.

 

[Agame3]

Oups. Sorry.

 

[Captain Fortius]

I think it would actually encourage theoretical sketti ships (which no one dreads or care about and haven't even been seen on a server since 2017).

Making systems 1-shot would make the best defense not getting hit at all. Which is what spaghetti is all about - reducing systems profiles to the minimum possible so they can't be hit.

How might ultra fragile systems discourage spaghetti (which to re-iterate, even before integrity, really didn't need a lot of discouragement anyway and was never a problem)?

They weren't a problem for me, yet for some obscure reason they were taken more seriously than they should have been.

But come to think of it, if they never get hit, then the CURRENT system is no spagetthi deterrant eaither! :D

So this new one wouldn't make the situtation worse on that front. (I talked about the perfect solution a couple of times before; making free-floating block groups die off - but as we can see, none of the higher ups seemed to like, or even notice that.)

 

[MacThule]

So I detect maybe a developing theme building towards almost a sort of consensus that it would be appropriate for system integrity operate in a way that is less about shape and more about damage to a system once it is in place and operating? Fragile systems in general instead of one particular vulnerability, something along those lines?

Which would translate to less of a "core drilling" situation and more of a situation where drilling out any system enough to shut it down is probably going to put the enemy in your hands unless you lose an equivalent system during the same period. I think it could be a great system... with enough thought and refinement. Trying for too much realism (i.e. any shot to a system, regardless of how minor, is likely to cripple it) might steal a lot of fun from the game, but I think a scale of penalties to performance and a rising chance for complete shutdowns and critical failures (boom) would be realistic, tension-building, and a way better means to introduce random, unknown factors into combat than recoil is.

If systems could shut down at any point along the line from being damaged, combat becomes excitingly unpredictable but in a very fair way. It's a dovetail between demanding precision targeting and demanding a respect for the fact that no amount of precision, planning, engineering, etc is always going to result in assured outcomes but rather outcomes become a function of the intersection between percentage chances of failure, skills and tactics being used, and the kind of damage actually happening. Maybe 90% of the time getting a single good shot into the enemy's thruster doesn't do much, but that 10% chance that you'll kill a whole engine group makes targeting it worth the while because even without a total system shutdown you're still eroding capabilities by targeting systems.
[doublepost=1527538155,1527537791][/doublepost]

They weren't a problem for me, yet for some obscure reason they were taken more seriously than they should have been.

But come to think of it, if they never get hit, then the CURRENT system is no spagetthi deterrant eaither! :D

So this new one wouldn't make the situation worse on that front. (I talked about the perfect solution a couple of times before; making free-floating block groups die off - but as we can see, none of the higher ups seemed to like, or even notice that.)

Actually, far from obscure, the reason is because a loud minority who wanted to cling to the old system used the phantom menace of flying spaghetti monsters to try and convince everyone that new power was "terrible" and could never work because of things like spaghetti (even though spaghetti originated with developments that were already occurring under power 1.9). The result is we got checks and balances put in place for a problem that was never a serious problem. Having a few determined people with alts making a huge ruckus for weeks on end can distort reality though. Look at the US White House... things happen.

 

[Captain Fortius]

Yes! Perfectly good insight. (on all counts)
I'd play that game.

 

[Agame3]

Maybe 90% of the time getting a single good shot into the enemy's thruster doesn't do much, but that 10% chance that you'll kill a whole engine group makes targeting it worth the while because even without a total system shutdown you're still eroding capabilities by targeting systems.

So, multiple outputs one behind the other, for making that 10% as certain as possible is OK?

Spoiler

Really, engineering is not all about fiability, but fiability is a pretty important part of engineering, just like efficiency. (Engineers are the one who design real life systems and that is for a reason.) Efficiency = the ratio effect/effort. Either increasing effect for the same effort either decreasing effort for the same effect. The most damage done to an enemy with the least clicks or thinking effort during battle is a form of efficiency. Of course, finding the most efficient solutions is an effort by itself, unfortunately...

Fiability = lowering the chances of failure. Usually it needs studying all (limited by engineer's imagination) possibilities of failure. From a Fiability point of view, a honest battle, one that gives you 50% chances of failure, is unacceptable. 50% is huge. It's an engineer's job to lower that chance until it gets negligible. Exploit/bug use or not, fiability and efficiency are what matters. Moral considerations are the concern of the users, not of the designer.

I met in real some people who are not engineers, but think like one. In this game I met even more. Yet, I see so few community content made by them... IDK why. It should be the norm, not the exception...

 

[Zoolimar]

I met in real some people who are not engineers, but think like one. In this game I met even more. Yet, I see so few community content made by them... IDK why.

There is enough content like this, just most people don't see a reason to load up it on the dock I think. And most of them seem to have left after power 2.0 was made due to the fact that there is nothing left to do for them. At least for now. I've seen one of the Zyrr's small ships, which was not a full spaghetti but still had a very low density - it looked a little like an empty square with pillars at the corners. No armour, no decor. Just systems and multi-beam turrets.

It burned through ~30 layers of girders and shield capacitor blocks on top of an advanced armour plate in around 2 seconds. On a moving ship with 2 TWR.

 

[Agame3]

There is enough content like this, just most people don't see a reason to load up it on the dock I think. And most of them seem to have left after power 2.0 was made due to the fact that there is nothing left to do for them. At least for now. I've seen one of the Zyrr's small ships, which was not a full spaghetti but still had a very low density - it looked a little like an empty square with pillars at the corners. No armour, no decor. Just systems and multi-beam turrets.

It burned through ~30 layers of girders and shield capacitor blocks on top of an advanced armour plate in around 2 seconds. On a moving ship with 2 TWR.

I am glad to hear that. (I mainly play an outdated version, probably at some point I will make my server for that version.) I mean I find this kind of tables for power 1, Legacy Power Systems - StarMade Wiki and probably soon for pow2, but nothing at all about diminishing returns and power needed for different weapon setups (number of weapon modules needed to do 1M damage per hit or per second, optimal setup regarding how many groups for a certain number of modules give the most damage per hit, knowing there are both diminishing returns for too large weapon groups but also for too many small weapon groups). Neither in Weapon 1, Weapon 2 or the incoming Weapon 3.

I find lots of threads about what block is best used as a toilet in RP (For the love of Christ!) and what turret looks best but none about ways to make a warhead torpedo miss its target in first pass to go around the frontal docked armor of an enemy ship and hit it from behind or from the side. I find way more threads with complains about certain useful blocks used as useless decorations that are now obsolete and impossible to get than threads about side-launching missile batteries used to circumvent the same frontal armor on needle ships.

And then everyone complains everyone else plays the game for PvP and only PvP, lol! Where are the actual PvP forums? About devices, systems, tactics, not about what faction killed who on a server nobody but 10-20 people ever heard of.

This thread is a nice one, sorry for derailing it a bit. But where are all the others? (And where is the math in this thread, actually?)

 

[MacThule]

So, multiple outputs one behind the other, for making that 10% as certain as possible is OK?

You could I guess, but it might be wasted effort this early in the development of new weapons.

And it might not help anyway. Again, consider that ~400 outputs across 50 AI entities don't do more than a couple dozen hits on a 20K mass ship rolling around at very low speed over a 2-3 sector area for 10 and 15 minute stretches (while I made food in the other room) over multiple test sessions.

I see this less as a challenge to up the outputs to 4,000 and increase the number of AI entities to 200, and more as a need to improve the AI's ability to handle new cannons or improve the function of new cannons themselves. I seriously doubt cannon are intended to require a vast array of waffled and layered outputs to get any hits in at all.

 

[Agame3]

Actually in a virtual world with inertial dampeners found by default and not by design in everything, where gravity is created as needed by 1 cubic meter cheap module and antimatter is obtained straight from energy in other modules... I do not think recoil is still a thing. Recoil is unbelievable and unnecessary, AI is fine.

Edit: You know I meant AI does not need to get the burden of dealing with recoil, not that AI was ever smarter than a starving bacteria. Sorry for saying AI is fine.

 

[Scypio]

Actually, far from obscure, the reason is because a loud minority who wanted to cling to the old system used the phantom menace of flying spaghetti monsters to try and convince everyone that new power was "terrible" and could never work because of things like spaghetti (even though spaghetti originated with developments that were already occurring under power 1.9).

Mmmh.. Take someone and your best ship, then download the fair and balanced, then give it to the random guy you found and then just look who win. Spaghet didn't got used much because you needed to get rid of the usual way of thinking when building and designing your ship (and that's hard) and the few that actually came to this idea agreed to don't use them. But again, don't chit chat, do what i said in my first sentence and look yourself. I always blame myself for never thinking about spaghet myself while i do loved vertical ship for the exact same reason. They are just one step further...

I met in real some people who are not engineers, but think like one. In this game I met even more. Yet, I see so few community content made by them... IDK why. It should be the norm, not the exception...

Mmmmh... Because they are the upper layer of the community, the ones that push the limits. Most ppl here just want to build a pretty ship and feel it useful against pirates. And well, when you build a ship that literally takes you month, you don't share it like that.

 

[Nosajimiki]

I think I'd go a little tamer and just say that systems should suffer soft failures like shutdowns or delays when damaged, and that damage should leech to them from adjacent hits even thru shields, scaling in leech distance by system size. You wouldn't want to cram shield blocks randomly in walls, especially near the outer hull, because your shields would get messed up the moment those walls would be hit.
That's essentially the Star Trek system of ship combat and system disabling, and ultimately I think it's the system that works best for making silly designs crammed with systems fragile (altho StarMade can do without the explosivity) in a way that feels intuitive, along with "target their ____" even while shields are up.

Soft failure, sure, but shield leech is terrible from a game design perspective. If you have enough of an advantage against an enemy, you should be able to walk away unscathed without having to go through the annoyance of repairs. Current shield models already over-emphasize this enough. Don't punish the guy with the overwhelming force for daring to attack a 500 mass fighter, punish the fighter for daring to solo a battleship 1000x his size, because I can promise you that replacing the fighter will be less annoying that putting that big ship in a shipyard and hoping you have the exact block types you need pre-loaded into it to fix the scratches.

Especially since most of these engagements are just gonna be as dumb as you flew too close to a pirate base and it took a few pot shots at you.
[doublepost=1527626379,1527624399][/doublepost]

Therefore, turning rate of a PvP ship is one of the most important characteristics of it.

Not really true.
Reason #1: Lag turning - Battle causes lag that allows even big ships to turn fast enough to point into pretty much any enemy.
Reason #2 Range Control - Most PvP ships used some kind of stop weapon meaning you could not get within 5k of one without being ground to a halt and held there until they decided to let you go. At that range there was no way to capitalize on poor turning radiuses
Reason #3 turrets - Even if you could momentarily get around a ship, most PvP ships could focus at least 75% of their firepower in nearly any direction anyway

So this new one wouldn't make the situation worse on that front. (I talked about the perfect solution a couple of times before; making free-floating block groups die off - but as we can see, none of the higher ups seemed to like, or even notice that.)

Because that theory was disproved by gird designed spaghetti which would be harder to break up than solid systems further strengthening the spaghetti meta.

Actually, far from obscure, the reason is because a loud minority who wanted to cling to the old system used the phantom menace of flying spaghetti monsters to try and convince everyone that new power was "terrible" and could never work because of things like spaghetti (even though spaghetti originated with developments that were already occurring under power 1.9). The result is we got checks and balances put in place for a problem that was never a serious problem. Having a few determined people with alts making a huge ruckus for weeks on end can distort reality though. Look at the US White House... things happen.

Not alts, I can personally verify from in game experience that nearly all of those people posting were people's primary accounts. Don't make assumptions because you played on the safe servers where people didn't play to win.

And then everyone complains everyone else plays the game for PvP and only PvP, lol! Where are the actual PvP forums? About devices, systems, tactics, not about what faction killed who on a server nobody but 10-20 people ever heard of.

This thread is a nice one, sorry for derailing it a bit. But where are all the others? (And where is the math in this thread, actually?)

PvPers tend to share less, for 2 reasons: For some it is because it is about the competitive advantage. For others, it is about the desire to prevent toxic designs from spreading. But if you want to learn about some of starmade's historical exploits and metas, the trident wiki has a few good ones, but even this is a pretty incomplete list:

Complex Armor
Disruptor Systems
Docked Battery
Logic Drive

You could I guess, but it might be wasted effort this early in the development of new weapons.

And it might not help anyway. Again, consider that ~400 outputs across 50 AI entities don't do more than a couple dozen hits on a 20K mass ship rolling around at very low speed over a 2-3 sector area for 10 and 15 minute stretches (while I made food in the other room) over multiple test sessions.

I see this less as a challenge to up the outputs to 4,000 and increase the number of AI entities to 200, and more as a need to improve the AI's ability to handle new cannons or improve the function of new cannons themselves. I seriously doubt cannon are intended to require a vast array of waffled and layered outputs to get any hits in at all.

Depends a lot on how those outputs are laid out and what kind of armor you are dealing with.

 

[MacThule]

Don't make assumptions because you played on the safe servers where people didn't play to win.

Excuse me?