Analysing the Thargoid Simulation

[Ian Doncaster]

One tick of progress (~2%) on HIP 20485 already.

HIP 23716 has three ticks too - presumably people are staying there and having another go - so it looks like there has been a little bit of threshold reduction to make systems winnable with large effort.

Ebisu has a 5 week timer (assuming accurate) - seven ports. Looking at the other new ones, six ports only seems enough for a 4 week timer.

Ports left	Weeks left
0	1
1	2
2 or 3	3
4, 5 or 6	4
7	5

That pattern would suggest that 11 ports might be enough for a 6 week timer ... but are there any systems which are an exception to the above pattern? I couldn't find any but I can't be certain I checked all of them.

 

[Ralph Vargr]

Today's Galnet also raises the question of what the Thargoid's goal is: it doesn't appear to be a specific system, or if it is, the Thargoids don't know which one either. It might become more obvious once all eight are in and have had chance to take more territory ... or it might be something more abstract like "surround us first". I'll add that to the unanswered questions list, though not with any hope of an answer soon.

Implosion method.

 

[Ned Flandalorian]

Re: thargoid combat bonds, due to all stations in HIP 23716 now being either abandoned, damaged or under attack there is nowhere in that system to hand them in. So they are definitely not having any influence on the progress being made in that system.

 

[Ian Doncaster]

Implosion method.

Possibly, though doing that in 3-D when your initial strike points are spread not much further than a single octant is going to hit the centre way before it hits a lot of other places, so it's not going to look very much like an implosion on time-lapse so much as a wave.

I'm not sure the strategic benefits of it really apply so much when you have near unhindered movement within your enemy's territory, your enemy has no meaningful outside allies to cut them off from, and you have a devastating first strike capability but lesser second-strike capability.

I wonder if what they're trying to do first is build a wall - all of their arrival points are essentially on the Pleiades octant, so if they connect them up, and thicken them inward and outward, they'll essentially require most ships to do a massive detour to the Pleiades (Witch Head, California) to avoid hyperdictions.

 

[Ashnak]

Possibly, though doing that in 3-D when your initial strike points are spread not much further than a single octant is going to hit the centre way before it hits a lot of other places, so it's not going to look very much like an implosion on time-lapse so much as a wave.

I'm not sure the strategic benefits of it really apply so much when you have near unhindered movement within your enemy's territory, your enemy has no meaningful outside allies to cut them off from, and you have a devastating first strike capability but lesser second-strike capability.

I wonder if what they're trying to do first is build a wall - all of their arrival points are essentially on the Pleiades octant, so if they connect them up, and thicken them inward and outward, they'll essentially require most ships to do a massive detour to the Pleiades (Witch Head, California) to avoid hyperdictions.

As German:

"Nobody has the intention to build a wall."

Sorry.

 

[]M[]

As German:

"Nobody has the intention to build a wall."

Sorry.

...but it's what one might feel tempted to do, with increasingly irritating incursions from an apparently hostile species - and the recent evidence that they may be developing technologies that may cause actual harm eventually, ie: to set / enforce some boundaries?

...that said, a wall can be a boundary to space - or the beginnings of a fortress ;]

 

[goemon]

hypothesis
maelstrom creates a cube like expansion cube calculations
with a sphere of influence of maximum coordinates delta of 10 ly per axis

looks like cube calculation with maximum delta on one of the axis being <10 ly.

exampels:
source: Hyades Sector FB-N b7-6

1. Ildano. Distance 21,64 ly (smallest delta -9,72 on z axis)

2. Trianguli Sector ZJ-A b1. Distance 22,18 ly (smallest delta -6,22 on Z-Axis)

 

[Ian Doncaster]

maelstrom creates a cube like expansion cube calculations
with a sphere of influence of maximum coordinates delta of 10 ly per axis

looks like cube calculation with maximum delta on one of the axis being <10 ly.

If one of the three axes has to be <10 LY but the others can be higher, that'd give a six-pronged shape rather than a cube, wouldn't it?

Would a sphere be a better model for that? Then you'd get small delta on one axis when the other two are large coming out automatically rather than needing to be explicitly defined.

 

[Ian Doncaster]

Trying to get some better data on how the difficulty varies between systems week-to-week - has anyone got a screenshot (ideally, timestamped) of HIP 23716 progress from the last hours of last week, please?

 

[Ralph Vargr]

Possibly, though doing that in 3-D when your initial strike points are spread not much further than a single octant is going to hit the centre way before it hits a lot of other places, so it's not going to look very much like an implosion on time-lapse so much as a wave.

I'm not sure the strategic benefits of it really apply so much when you have near unhindered movement within your enemy's territory, your enemy has no meaningful outside allies to cut them off from, and you have a devastating first strike capability but lesser second-strike capability.

I wonder if what they're trying to do first is build a wall - all of their arrival points are essentially on the Pleiades octant, so if they connect them up, and thicken them inward and outward, they'll essentially require most ships to do a massive detour to the Pleiades (Witch Head, California) to avoid hyperdictions.

Forming a contiguous front makes sense. The last three impacts will tell us more.

 

[Ralph Vargr]

...but it's what one might feel tempted to do, with increasingly irritating incursions from an apparently hostile species - and the recent evidence that they may be developing technologies that may cause actual harm eventually, ie: to set / enforce some boundaries?

...that said, a wall can be a boundary to space - or the beginnings of a fortress ;]

Festung Thargoidia?

 

[GraphiteGB]

Think the more important question is do they have an AI that knows how to make a Globe not just a ring.
be knida silly to ring the bubble but allow people to jump 50 lys and then just go over there invasion space...

 

[nemolomen]

My guess is that the AI will expand toward Colonia leaving the bubble a burning wreck as it eats more and more. IF it has the capacity to do that then there is no need for envelopment...

 

[goemon]

If one of the three axes has to be <10 LY but the others can be higher, that'd give a six-pronged shape rather than a cube, wouldn't it?

Would a sphere be a better model for that? Then you'd get small delta on one axis when the other two are large coming out automatically rather than needing to be explicitly defined.

the thing with sphere calculation is it is heavy to calculate for algorythms, which i think was the prime candidate for why expansion does cube calculations.
but more data is needed ...

 

[yttrbio]

You can already see where all maelstroms will arrive if the predictions are correct. They roughly form 2 diamonds at a slightly obtuse angle, between the bubble and the galactic rim, like a big set of jaws that are going to chomp down on the bubble. So it looks less like an envelopment and more like a wall on one side.

 

[Reldas]

Trying to get some better data on how the difficulty varies between systems week-to-week - has anyone got a screenshot (ideally, timestamped) of HIP 23716 progress from the last hours of last week, please?

Is this what you need? HIP 23716, 3mins 26 seconds before the weekly tick.

 

[Ian Doncaster]

the thing with sphere calculation is it is heavy to calculate for algorythms, which i think was the prime candidate for why expansion does cube calculations.
but more data is needed ...

True, but it's not that heavy, and it only needs doing weekly for eight sites which can themselves be efficiently trimmed to a mostly right set with a cube first if necessary.

Is this what you need? HIP 23716, 3mins 26 seconds before the weekly tick.

Brilliant, exactly what I needed. Thank you.

 

[]M[]

True, but it's not that heavy, and it only needs doing weekly for eight sites which can themselves be efficiently trimmed to a mostly right set with a cube first if necessary.

This kind of efficiency is no doubt necessary on the scale of the galaxy model itself, with multiple levels of sectors / sub-sectors - but yes, in terms of calculating just 8 sets of coordinates, that’s a handful of CPU cycles and mere fractions of a second.

 

[bobbyroberts99]

Thargoid Invasion (orange)​

In a Thargoid Invasion state, the system is under intense Thargoid attack. Supercruise traffic may be interdicted by Thargoids, hyperdictions will occur, and stations in the system will become AX combat zones. Stations will gradually (unknown: in what order, and when in the invasion process?) become damaged and abandoned via this process, and it appears possible from detail in the system map to tell if a station is at risk this week.

In inhabited systems, this state lasts for at least 2 weeks and often more, depending on the number of ports in the system, and can be resisted by destroying Thargoids, delivering supplies, evacuating civilians to rescue ships, and completing missions.

If sufficient activity (currently: an implausibly large amount) takes place in the system, the attack is repelled and the system enters the Recovery state. Otherwise, it either moves to Control or the current battle to protect stations is lost, extra stations are damaged or abandoned, and the progress to repelling the attack resets.

The political simulation appears to be frozen with no influence or state movements during an Invasion (Unknown: are inbound expansions still allowed? what about outbound if the state was already active? conflicts which were active when the invasion started?). Powerplay Control or Exploit do not appear to be directly affected at this stage.

Here's an alternative hypothesis to the current conventional wisdom that we win the system and the thargoids leave the first time we 100% the progress bar in an invasion system or that we lose all progress if we don't get to 100%.

Instead, each week we fight to defend a station. If we 100% the bar, the station in question stays in our control and does not progress to the Burning state. If we fail to 100% the bar, that station transitions to Burning. After the Thursday tick, if there are still unattacked stations, we fight over another one (or maybe two). We do this each week until there are no stations to defend. Then during the last week of the invasion, we fight for control of the system. If we 100% the final bar, we keep control of the system and it enters Recovery. If we lose the final bar, the system goes to Thargoid Infestation with zero population, and in a week we will be able to challenge for control again by killing thargoids. During the Recovery phase, any stations that we successfully defended are returned to full operation and any that we failed to defend will need to be repaired.

Here's another way to look at it: failing a progress bar does not reset all of our progress except if we lose the final week. We win or lose objectives if we fail to 100% the bar in the weeks preceding the final battle for a system.

FDevs UI and explanations are for sure unclear, incomplete and opaque (but maybe less so than the BGS?). All is not lost. Right now we're still learning (part of the fun IMO). I'm not sure if this is how the system works but thought that I'd throw out an alternative way to look at the situation that, to me anyway, is less demoralizing than losing all progess after throwing everything at a system and just losing all that effort. We just lost a station in HIP 23716 which we can get back later if we want it badly enough. This system lets us pick our targets. Defending Alert state systems seems like a good way to avoid weeks-long invasions. Thoughts?

 

[Ian Doncaster]

Thoughts?

It's definitely not the obvious interpretation of the UI for Invasion, but then neither is "the timer is actually a series of independent 1-week timers for different sets of stations". Hopefully we'll find out with HIP 23716 this week.

That approach would have the slightly annoying property that if you wanted to keep all the stations you'd need to fight hard for N weeks, and then you could still lose the entire system (and therefore all the stations anyway) by losing the last week.