Empire of the Stars - Comments Thread

[MinuteMan]

Hey Steve,

How does this campaign compare to your previous ones in terms of scale?
Amount of NPR's, amount of colonies, fleets, etc.

It seems to me that it's starting to get rather big. 
Is the performance still fine after the civilian line updates?

[Steve Walmsley (OP)]

Hey Steve,

How does this campaign compare to your previous ones in terms of scale?
Amount of NPR's, amount of colonies, fleets, etc.

It seems to me that it's starting to get rather big. 
Is the performance still fine after the civilian line updates?

This is the largest one I have documented for C#, although I had a larger one for which I didn't post an AAR (500+ known systems, 7 swarm races, etc.).1

Performance is fine - at least I haven't noticed anything except that saves are taking longer.

[Bughunter]

Great read as always, I love the short, direct to the point format of Steves AAR:s.

23rd October 1983
After an NPR clears the orbital defenders, maybe they should consider the loot on the ground theirs and take offense if you move in with ground forces and snipe it for yourself? On the other hand if you add in every special case like this you will soon have a very complex AI especially considering different goals will sometimes conflict.

30th December 1983
Mechs.. should have seen it coming in a japanese campaign but it just hadn't crossed my mind, brilliant 

[Kurt]

Just finished wading through your massive set of updates.  Great writeups! The massive scale of your campaign continues to impress. 

The communications issue continues to intrigue me.  Your assumption of instantaneous comms across the length and breadth of your empire simplifies matters greatly, not only allowing the coordination of forces across multiple systems, but allowing you, the gamer, to not have to deal with light speed delays, or comms that depend on courier traffic.  It makes it easier for you to control your empire and perceive the big picture, and also gives your campaign a more "modern-day" feel, as you have nearly instantaneous comms between ships and worlds.  Whereas in my campaign, with no FTL comms, communication is limited to light speed within a system and to courier or other ship speeds between systems. 

My campaign is much smaller than yours at this point, but communication takes much longer, obviously.  The times (by courier) are as follows:

Terra to Victory Base in Groombridge (14.6 billion kilometers): 17 days
Terra to Shield Base in Monoceri (9.32 billion): 10.9 days
Terra to Bastion Base in Chi Draconis, beyond the frontier(12.1 billion kilometers): 14.1 days
Terra to Omega Ceti system, new swarm location (22.6 billion kilometers): 26 days

Compared to your campaign, my campaign has a more World War I or even 19th century feel to it, as comms are limited and frontline commanders have great latitude and independence to do what they see fit without referring to higher command. 

To be honest, I'm not sure I could manage this setup in a campaign the size of yours.  The complexity of tracking who knows what and when they knew it would likely become unmanageable in short order. 

I noticed that in your first post from yesterday, one of your carriers launched an attack using its missile equipped fighters.  I was particularly interested in this as this is a major issue I'm dealing with in my campaign.  The Imperial Terran Fleet's initial design concept was large beam-armed combatants, supported by smaller specialized escorts and patrol craft.  The demands of engaging the swarm have forced me into deploying carriers with small fast fighters equipped with missiles.  I have struggled to design carriers that carry sufficient fighters and large enough magazines to rearm to the fighters for additional attacks.  Currently my carriers have magazines large enough to rearm their fighters twice, giving them a capacity for three strikes before needing to reload at a colony or munitions ship.  I keep thinking I'm doing something wrong and that I should be able to have larger magazines, but after reading about your carrier strike I realized that I'm probably just channeling a previous version of Aurora in thinking I could cram more magazines into the design.     

[Steve Walmsley (OP)]

Just finished wading through your massive set of updates.  Great writeups! The massive scale of your campaign continues to impress. 

The communications issue continues to intrigue me.  Your assumption of instantaneous comms across the length and breadth of your empire simplifies matters greatly, not only allowing the coordination of forces across multiple systems, but allowing you, the gamer, to not have to deal with light speed delays, or comms that depend on courier traffic.  It makes it easier for you to control your empire and perceive the big picture, and also gives your campaign a more "modern-day" feel, as you have nearly instantaneous comms between ships and worlds.  Whereas in my campaign, with no FTL comms, communication is limited to light speed within a system and to courier or other ship speeds between systems. 

My campaign is much smaller than yours at this point, but communication takes much longer, obviously.  The times (by courier) are as follows:

Terra to Victory Base in Groombridge (14.6 billion kilometers): 17 days
Terra to Shield Base in Monoceri (9.32 billion): 10.9 days
Terra to Bastion Base in Chi Draconis, beyond the frontier(12.1 billion kilometers): 14.1 days
Terra to Omega Ceti system, new swarm location (22.6 billion kilometers): 26 days

Compared to your campaign, my campaign has a more World War I or even 19th century feel to it, as comms are limited and frontline commanders have great latitude and independence to do what they see fit without referring to higher command. 

To be honest, I'm not sure I could manage this setup in a campaign the size of yours.  The complexity of tracking who knows what and when they knew it would likely become unmanageable in short order. 

I noticed that in your first post from yesterday, one of your carriers launched an attack using its missile equipped fighters.  I was particularly interested in this as this is a major issue I'm dealing with in my campaign.  The Imperial Terran Fleet's initial design concept was large beam-armed combatants, supported by smaller specialized escorts and patrol craft.  The demands of engaging the swarm have forced me into deploying carriers with small fast fighters equipped with missiles.  I have struggled to design carriers that carry sufficient fighters and large enough magazines to rearm to the fighters for additional attacks.  Currently my carriers have magazines large enough to rearm their fighters twice, giving them a capacity for three strikes before needing to reload at a colony or munitions ship.  I keep thinking I'm doing something wrong and that I should be able to have larger magazines, but after reading about your carrier strike I realized that I'm probably just channeling a previous version of Aurora in thinking I could cram more magazines into the design.     

It's the first time I have done comms this way. Usually I assume I can transit using the stable jump network, but not outside it. This approach does make life easier for large campaigns.

I used to build carriers with two reloads too, but I realised the battle was often won or loss bt the time I needed the second reload, so I decided for this campaign to go with a single reload, and use colliers if necessary.

[Kurt]

<snip>

It's the first time I have done comms this way. Usually I assume I can transit using the stable jump network, but not outside it. This approach does make life easier for large campaigns.

I used to build carriers with two reloads too, but I realised the battle was often won or loss bt the time I needed the second reload, so I decided for this campaign to go with a single reload, and use colliers if necessary.

This does raise an interesting facet of campaign play style.  The NPR's and Spoilers are assumed to have instant FTL like you are using now, I assume.  A player race could use one of several methods to communicate between systems.  No inter-system comms at all, like my current campaign.  Light speed comms between system, like Starfire, with buoys at the jump points assumed to be using small jump-capable torpedoes.  And instant comms, like you are using.  In a multi-player race, with different player races using different comms methods, it might introduce interesting situations, although the instant FTL comms using race would have a significant strategic advantage. 

For the carriers, I find two reloads to be useful in engaging the swarm, as they tend to use large numbers of small ships and this gives my carriers the ability to deal with them effectively with multiple strikes.  The recent engagements have strongly motivated me to up my support ship game.  My flagship, the Andromeda, an 80,000 ton dreadnought, survived a close encounter with the swarm, but was almost gutted by acid damage after the battle.  The ship's crew managed to repair a lot of systems, but ran out of maintenance supplies.  And my carriers ran out of missiles before they ran out of targets.  I failed to effectively coordinate my attack with the support units, and my UNREP ships were slow and ungainly, meaning it took time to catch up with the fleet and were hideously vulnerable to attack by the enemy. 

[Steve Walmsley (OP)]

<snip>

It's the first time I have done comms this way. Usually I assume I can transit using the stable jump network, but not outside it. This approach does make life easier for large campaigns.

I used to build carriers with two reloads too, but I realised the battle was often won or loss bt the time I needed the second reload, so I decided for this campaign to go with a single reload, and use colliers if necessary.

This does raise an interesting facet of campaign play style.  The NPR's and Spoilers are assumed to have instant FTL like you are using now, I assume.  A player race could use one of several methods to communicate between systems.  No inter-system comms at all, like my current campaign.  Light speed comms between system, like Starfire, with buoys at the jump points assumed to be using small jump-capable torpedoes.  And instant comms, like you are using.  In a multi-player race, with different player races using different comms methods, it might introduce interesting situations, although the instant FTL comms using race would have a significant strategic advantage. 

For the carriers, I find two reloads to be useful in engaging the swarm, as they tend to use large numbers of small ships and this gives my carriers the ability to deal with them effectively with multiple strikes.  The recent engagements have strongly motivated me to up my support ship game.  My flagship, the Andromeda, an 80,000 ton dreadnought, survived a close encounter with the swarm, but was almost gutted by acid damage after the battle.  The ship's crew managed to repair a lot of systems, but ran out of maintenance supplies.  And my carriers ran out of missiles before they ran out of targets.  I failed to effectively coordinate my attack with the support units, and my UNREP ships were slow and ungainly, meaning it took time to catch up with the fleet and were hideously vulnerable to attack by the enemy.

I really need to catch up with your campaign and others. I've been too busy playing and developing

I'm currently working on adding the ability to drop a large, very high tech and fully developed NPR with many systems into the middle of an existing campaign, without any immediate connection between the NPR systems and the existing systems.

[Garfunkel]

That sounds amazing!

And also, what an epic string of updates!

[doodle_sm]

Empire of the Stars has been informative and a joy to read. I never thought I'd be so invested in characters like Kurosawa Rai and Suzuki Saya.

[nuclearslurpee]

Finally caught up on this proper epic. Very cool to see the amount of ground invasions of alien homeworlds, I'll admit I expected the required tonnages to be in the several millions but it seems that Steve has gone all-in on heavy armor and genetically-enhanced super-soldiers to minimize the transport capacity required - so it's cool to see that approach work very well!

I do note that it seems like logistics continue to be the biggest challenge in mounting a war against a peer polity, and the AI still doesn't seem to present a great tactical or operational challenge, although I may be drawing too many conclusions from the most recent conquest of the Zharovians who were quite dispersed and worn-down from their other war. The Scimitar, I suspect, would have been a stiff out if they even had technological parity with the Empire, after all, but that was earlier on. I guess perhaps the next AI challenge for Steve will be how to get the AI to control their forces without dispersing so much, so that even in the midst of one war they remain a difficult target for the opportunistic human players!

On communications, I use light-speed in systems and instant comms across jump points, usually with a rule that a jump point must have ships or comms buoys on one or each side to facilitate those communications. This latter rule makes it possible for the aliens to cut off communications in times of war, or for a survey ship to go mysteriously missing in a system if comms have not yet been established. The one exception is my Duranium Legion setting where I allow stabilized jump points to function this way instead, for historical reasons. To me, it doesn't make a huge difference in gameplay (unless comms are cut off), since the light-speed delay is usually much, much less than how long it will take any ships to get there in response, but it does make a difference in roleplay opportunities.

@Kurt: your idea of courier-based communications is very interesting. What is the "physics" justification for this over light-speed comms? I think it could be managed in a campaign of this size if you use real ships for the job, combined with the "Send Message" order. It would add some micro but probably not too much if your roleplay goal is to prioritize independent action by commanders.

[Steve Walmsley (OP)]

I do note that it seems like logistics continue to be the biggest challenge in mounting a war against a peer polity, and the AI still doesn't seem to present a great tactical or operational challenge, although I may be drawing too many conclusions from the most recent conquest of the Zharovians who were quite dispersed and worn-down from their other war.

The problem for the Zharovians was they had the worst possible design combination against my designs. They were entirely energy-based and had slower ships and shorter-ranged weapons. My main problem was running out of MSP. The Bloodclaw were more dangerous and if they had been at full strength, instead of worn down by two separate wars, they could have been formidable. Based on what I found when I salvaged their home world bases, I suspect the Yamatos would have been quickly destroyed if they had attacked them at full strength.

I now have two alien home worlds occupied and it is a matter of time before the third one. The Capellans have a large fleet and could be a problem, but I just launched my first 120,000 ton Dreadnought, which is much more powerful than the Yamatos. Once I assimilate the Zharovians and eventually the Bloodclaw, the Empire will be hard to stop without a significant new challenge.

Because of the need for that challenge I started working on the generation of large, pre-generated NPR Empires with tens or even hundreds of systems, complete with established colonies and distributed forces. They will be assigned a 'number of transits' distance for exploration from their home world, then generate all those systems and distribute a preset total of terraforming, installations, etc to the potential colony sites in that area, with more assets to the best sites, but generally smaller colonies further out. When generating systems, they will not connect to any systems that existed before the race was created, but they can connect to their own systems.

Initially my plan was to add a new, very large alien Empire to my existing game, but I will more likely create a new game with pre-generated NPR empires.

I'm also looking at expanding the spoiler races to create larger threats. So far its only ideas, but I am considering having a minority of Aether Rifts start much larger (instant 'Eye of Terror'), multiple swarms appearing in known space simultaneously or within a short period (Tyranid Invasion) and some form of Tomb Worlds (Necrons) where aliens emerge from large underground facilities and begin expanding as they activate more forces. This would either be a variation of precursors, or a new threat entirely, probably with some key installation on each tomb world that can be destroyed to prevent further activation. I know these are all 40K-themed, but so were the original spoilers in many cases.

Once all that is done, I'll probably start a new campaign.

[nakorkren]

Because of the need for that challenge I started working on the generation of large, pre-generated NPR Empires with tens or even hundreds of systems, complete with established colonies and distributed forces. They will be assigned a 'number of transits' distance for exploration from their home world, then generate all those systems and distribute a preset total of terraforming, installations, etc to the potential colony sites in that area, with more assets to the best sites, but generally smaller colonies further out. When generating systems, they will not connect to any systems that existed before the race was created, but they can connect to their own systems.

I am EXTREMELY interested in the update to allow the addition of muti-system/planet NPCs! I've always wished we would encounter NPCs that had already substantially colonized (or conquered) multiple major planets, and that it would offer more variety and challenge particularly as the player is usually in that state. Will you be implementing that separately from the current system of scaling newly discovered NPCs based on your own size adjusted by difficulty and a random factor? The way you described it, with NPC extent being entirely in undiscovered systems (which makes sense), makes it sound like that's your intent. Would you consider making it part of the base NPC discovery mechanism? In theory in your current game, once you incorporate the two homeworlds you've captured, the next NPC you find will be massive but entirely contained (initially) to a single colony, which seems weird. Having a super "tall" NPC will also result in them very rapidly depleting their minerals on the homeworld, unless that is also scaled with starting NPC size?

I just launched my first 120,000 ton Dreadnought, which is much more powerful than the Yamatos.

Can't wait to see the specs in your next (final?) writeup. This has been a campaign for the ages. A very interesting and well researched theme, many NPCs, an overextended player empire engaging in multi-front wars, very large combat ships, and not just one but multiple homeworld ground invasions!

Once all that is done, I'll probably start a new campaign.

At the risk of venturing into "when will the next update be out" territory... Do you intend to release 2.6 before the start of your next campaign, or continue updating it through that campaign?

[Kurt]

<snip>

@Kurt: your idea of courier-based communications is very interesting. What is the "physics" justification for this over light-speed comms? I think it could be managed in a campaign of this size if you use real ships for the job, combined with the "Send Message" order. It would add some micro but probably not too much if your roleplay goal is to prioritize independent action by commanders.

Couriers are for inter-system message transport, in-system comms are handled by the usual light speed transmission, either radio or laser based.  In my campaign I have built small cheap couriers with commercial engines and stationed several in each inhabited/exploited system.  To reduce player overhead, I don't actually move the couriers between systems to carry the messages, instead I plot the course to the destination and note the time involved without actually sending the couriers, as I would then have to send them back to their origin. 

As I noted elsewhere, my campaign is smaller than Steve's, at least currently, and even at its current size it is something of a pain to keep track of who knows what and when. 

Having given the situation some thought, I am probably going to establish a courier network, with jump-capable couriers stationed on each side of jump points, strung out in between systems.  This will effectively allow for near light-speed communications in between systems.  The couriers are small enough and cheap enough that it should be doable, and as commercial ships don't require maintenance checks.

Kurt

[skoormit]

I started working on the generation of large, pre-generated NPR Empires with tens or even hundreds of systems, complete with established colonies and distributed forces. They will be assigned a 'number of transits' distance for exploration from their home world, then generate all those systems and distribute a preset total of terraforming, installations, etc to the potential colony sites in that area, with more assets to the best sites, but generally smaller colonies further out. When generating systems, they will not connect to any systems that existed before the race was created, but they can connect to their own systems.

Initially my plan was to add a new, very large alien Empire to my existing game, but I will more likely create a new game with pre-generated NPR empires.

Please, please, please make it so that NPRs generated in-game can be like this.
I would much rather encounter a new NPR around the time that our exploration frontiers begin to overlap, rather than exactly when my exploration frontier reaches their home system.

Currently, Aurora only generates new systems as they are entered for the first time, which is why non-game-start NPRs are always discovered in their home system.
What if Aurora instead created systems ahead of the explored frontier by a number of hops (possibly a game configuration parameter)?

Those systems would exist unknown to the player (or any NPR) until discovered.
If a new NPR is generated in such a system, Aurora can follow the process you describe above to expand the NPR outside their home system, backfilling some or all of the systems between their home system and the currently explored frontier (and also expanding in the other direction from their home system, away from the explored frontier).

Most of these "exploration buffer" systems won't contain new NPRs, of course.
To avoid unnecessarily increasing calculations during turn processing, Aurora can leave uninhabited, undiscovered systems in stasis--no need to calculate orbital motion (or anything else) on a system that nobody yet knows about.

[nuclearslurpee]

I started working on the generation of large, pre-generated NPR Empires with tens or even hundreds of systems, complete with established colonies and distributed forces. They will be assigned a 'number of transits' distance for exploration from their home world, then generate all those systems and distribute a preset total of terraforming, installations, etc to the potential colony sites in that area, with more assets to the best sites, but generally smaller colonies further out. When generating systems, they will not connect to any systems that existed before the race was created, but they can connect to their own systems.

Initially my plan was to add a new, very large alien Empire to my existing game, but I will more likely create a new game with pre-generated NPR empires.

Please, please, please make it so that NPRs generated in-game can be like this.

I have always thought it's a bit weird to meet "coincidentally" NPRs just when they are beginning to play their own "TN Start". Even if the new NPR generation during a game was changed to cause some number of systems to generate in a chain before generating the NPR home system, that would be much more immersive to me.