Distant Worlds 2 - State of the Game and Update Roadmap

[zgrssd]

I notice on the roadmap there is no mention of improving the AI, do the devs have any thoughts on this? The AI is really quite weak, and may even be bugged in the sense that AI empires seem to cripple their own economy over-building troops.

Any work improoving automtion is work improoving hte AI. Since both systems are exactly the same.

[maggiecow]

I notice on the roadmap there is no mention of improving the AI, do the devs have any thoughts on this? The AI is really quite weak, and may even be bugged in the sense that AI empires seem to cripple their own economy over-building troops.

Any work improoving automtion is work improoving hte AI. Since both systems are exactly the same.

Not according to this thread, old pal:

https://www.matrixgames.com/forums/view ... 1&t=383035

Which would suggest player automation settings are better than whatever the AI is doing.

[zgrssd]

I notice on the roadmap there is no mention of improving the AI, do the devs have any thoughts on this? The AI is really quite weak, and may even be bugged in the sense that AI empires seem to cripple their own economy over-building troops.

Any work improoving automtion is work improoving hte AI. Since both systems are exactly the same.

Not according to this thread, old pal:

https://www.matrixgames.com/forums/view ... 1&t=383035

Which would suggest player automation settings are better than whatever the AI is doing.

Where exactly in those 23 posts does it say the thing you think it says?

[maggiecow]

Any work improoving automtion is work improoving hte AI. Since both systems are exactly the same.

Not according to this thread, old pal:

https://www.matrixgames.com/forums/view ... 1&t=383035

Which would suggest player automation settings are better than whatever the AI is doing.

Where exactly in those 23 posts does it say the thing you think it says?

It's literally the conclusion in the first post. Player AI on full automation beats the game AI on very hard, consistently. AI makes stupid mistakes, particularly with the economy, that the player automation doesn't. You literally only had to read one post. A long post. But one post.

From past experience I know you hate admitting you're wrong, but you said the enemy AI is the same as player automation AI and that post strongly suggests it isn't. Have you some proof that enemy AI is the same as player automation AI? Anything but a hunch? Any thing?

[zgrssd]

Not according to this thread, old pal:

https://www.matrixgames.com/forums/view ... 1&t=383035

Which would suggest player automation settings are better than whatever the AI is doing.

Where exactly in those 23 posts does it say the thing you think it says?

It's literally the conclusion in the first post. Player AI on full automation beats the game AI on very hard, consistently. AI makes stupid mistakes, particularly with the economy, that the player automation doesn't. You literally only had to read one post. A long post. But one post.

From past experience I know you hate admitting you're wrong, but you said the enemy AI is the same as player automation AI and that post strongly suggests it isn't. Have you some proof that enemy AI is the same as player automation AI? Anything but a hunch? Any thing?

He changed the entire starting conditions, by overwriting designs and focussing on building operations with longterm payoff.
As usual, compounding effects are in play.

The whole example he tried to make is overshadowed by all the things he did to not rely on Automation.
This only prooved that there is some room to optimize some designs and targets.

[maggiecow]

Where exactly in those 23 posts does it say the thing you think it says?

It's literally the conclusion in the first post. Player AI on full automation beats the game AI on very hard, consistently. AI makes stupid mistakes, particularly with the economy, that the player automation doesn't. You literally only had to read one post. A long post. But one post.

From past experience I know you hate admitting you're wrong, but you said the enemy AI is the same as player automation AI and that post strongly suggests it isn't. Have you some proof that enemy AI is the same as player automation AI? Anything but a hunch? Any thing?

He changed the entire starting conditions, by overwriting designs and focussing on building operations with longterm payoff.
As usual, compounding effects are in play.

The whole example he tried to make is overshadowed by all the things he did to not rely on Automation.
This only prooved that there is some room to optimize some designs and targets.

Answer the question....

[zgrssd]

It's literally the conclusion in the first post. Player AI on full automation beats the game AI on very hard, consistently. AI makes stupid mistakes, particularly with the economy, that the player automation doesn't. You literally only had to read one post. A long post. But one post.

From past experience I know you hate admitting you're wrong, but you said the enemy AI is the same as player automation AI and that post strongly suggests it isn't. Have you some proof that enemy AI is the same as player automation AI? Anything but a hunch? Any thing?

He changed the entire starting conditions, by overwriting designs and focussing on building operations with longterm payoff.
As usual, compounding effects are in play.

The whole example he tried to make is overshadowed by all the things he did to not rely on Automation.
This only prooved that there is some room to optimize some designs and targets.

Answer the question....

No, I have no proof the AI's are the same.
But neither do you have any proof they are different.

Making them the same means less work, so I think laziness favors my position slightly here.

[maggiecow]

He changed the entire starting conditions, by overwriting designs and focussing on building operations with longterm payoff.
As usual, compounding effects are in play.

The whole example he tried to make is overshadowed by all the things he did to not rely on Automation.
This only prooved that there is some room to optimize some designs and targets.

Answer the question....

No, I have no proof the AI's are the same.
But neither do you have any proof they are different.

Making them the same means less work, so I think laziness favors my position slightly here.

You'd think the XML profiles would be the definitive numbers, but it's pretty apparent from the modding section that there's something going on under the hood that bypasses them, where hard-coded results are used instead. I suspect you're overestimating how well this has all been coded. I mean, two and a half months of patching would back that up.

And regardless, the game needs better enemy AI. The automation seems to be getting there, but the enemy AIs still aren't putting up a fight. Are you ever struggling? I'd be surprised.

[AKicebear]

TLDR: We are prioritizing Critical Issues, Performance Issues and Gameplay issues in that order.

Hi Erik - I'm heartened to see the positive feedback towards recent patches, especially the engine upgrade. Looking through your list of priorities, it seems many could be considered addressed at least partially, maybe even some fully resolved (?).

It's been about three months since this comprehensive State of the Game post. Is it possible to provide an update - where do you consider yourselves on the top two of Critical Issues and Performance, that are rightfully a priority? Have new Gameplay issues replaced some of the old ones you've had a chance to tackle already?

I haven't played the game since April, so won't comment on what you've changed, but just note that I am really looking forward to coming back when retrofitting across multiple designs/hull types is fixed - my personal pet peeve. And before/after then, looking forward to when part of the game are de obfuscated for modders - this is one game that I can see motivating me to tinker again, maybe even learn something new, after a decade or so away.

Look forward to any update you can provide!

[Thuran]

+1
An updated roadmap would definitely be nice.

[Bleek]

+1
An updated roadmap would definitely be nice.

I believe Erik mentioned on the Steam forum that one is coming after the next release goes from beta to live.

[Erik Rutins]

Yes, once the 1.0.5.x series becomes official, I expect we'll prepare an updated state of the game and roadmap.

[azeios]

Yes, once the 1.0.5.x series becomes official, I expect we'll prepare an updated state of the game and roadmap.

I'd be really excited to see this! Just to add - I hope you are not disheartened by some of the negative feedback. While I think the game in its current form has quite a few flaws I really believe the potential is there for this to become my favourite 4X game with some fairly minor changes, many of which have been suggested on these forums elsewhere.

[OrnluWolfjarl]

Agreed, this is a fun game so far. With some more QoL changes and AI fixes, I think most of the playerbase will also be happy.

Personally, I'm waiting until we can upgrade different designs of the same hull separately to dive back in, but I still find myself itching for more and coming back to play a few hours per week.

[Jorgen_CAB]

Personally, I'm waiting until we can upgrade different designs of the same hull separately to dive back in, but I still find myself itching for more and coming back to play a few hours per week.

You know that this is sort of possible in the game right now... as long as you only have one of each class of ships in a specific fleet. I use this all the time. You can select which ship (of a specific class) in a fleet template should upgrade to and they do that automatically.

If this perhaps is enough for you or not I don't know, but this works good for me until they add a way to have multiple ships of the same class in a fleet at the same time, if they ever will.

[AKicebear]

You know that this is sort of possible in the game right now... as long as you only have one of each class of ships in a specific fleet. I use this all the time. You can select which ship (of a specific class) in a fleet template should upgrade to and they do that automatically.

If this perhaps is enough for you or not I don't know, but this works good for me until they add a way to have multiple ships of the same class in a fleet at the same time, if they ever will.

Will an upgraded design (e.g. to v2) automatically be used by the fleet that was using v1? Or do I need to through and manually select the new design for retrofitting?

[zgrssd]

You know that this is sort of possible in the game right now... as long as you only have one of each class of ships in a specific fleet. I use this all the time. You can select which ship (of a specific class) in a fleet template should upgrade to and they do that automatically.

If this perhaps is enough for you or not I don't know, but this works good for me until they add a way to have multiple ships of the same class in a fleet at the same time, if they ever will.

Will an upgraded design (e.g. to v2) automatically be used by the fleet that was using v1? Or do I need to through and manually select the new design for retrofitting?

I think so.
You need to have Automatic Retrofitting for the design on.
And I think there was a Checkbox in the Fleet Template Screen too, that needed to be checked.

The only limit is that you can not have more then 1 Model per Role per Fleet.
And the game behaves poorly if you have multiple Models per Role on the Designer side.

[Jorgen_CAB]

You know that this is sort of possible in the game right now... as long as you only have one of each class of ships in a specific fleet. I use this all the time. You can select which ship (of a specific class) in a fleet template should upgrade to and they do that automatically.

If this perhaps is enough for you or not I don't know, but this works good for me until they add a way to have multiple ships of the same class in a fleet at the same time, if they ever will.

Will an upgraded design (e.g. to v2) automatically be used by the fleet that was using v1? Or do I need to through and manually select the new design for retrofitting?

I think so.
You need to have Automatic Retrofitting for the design on.
And I think there was a Checkbox in the Fleet Template Screen too, that needed to be checked.

The only limit is that you can not have more then 1 Model per Role per Fleet.
And the game behaves poorly if you have multiple Models per Role on the Designer side.

As long as you are manually designing ships all is fine with the designer, don't allow the game to design warships if you want multiple designs roles in a class of ships, only if you ever have one design per class.

Your fleets will automatically update each class to the ship type you specify in the ship template. This works just fine too. Just make sure it is not set to update to the latest design.

[OrnluWolfjarl]

Personally, I'm waiting until we can upgrade different designs of the same hull separately to dive back in, but I still find myself itching for more and coming back to play a few hours per week.

You know that this is sort of possible in the game right now... as long as you only have one of each class of ships in a specific fleet. I use this all the time. You can select which ship (of a specific class) in a fleet template should upgrade to and they do that automatically.

If this perhaps is enough for you or not I don't know, but this works good for me until they add a way to have multiple ships of the same class in a fleet at the same time, if they ever will.

Unfortunately, it doesn't really work well with the way I play. I particularly like using large numbers of destroyers of different designs, accompanied by a few larger ships, also with specialized designs. Building designs and fleets according to my plans and needs is half the fun for me. If I want to do that with fleet templates, I'll have to use a big bunch of really small and different fleets, and that's hell to manage. Or compromise and use "regular" fleets, which gets boring for me, hence why I'm taking a break from DW2 currently.

It's my hope they'll implement this change, because it seems a lot of players are feeling it's sorely missing. If they don't implement it soon, then I'll guess I'll have to adjust the way I play, but I feel it would be a shame.

[Jorgen_CAB]

Personally, I'm waiting until we can upgrade different designs of the same hull separately to dive back in, but I still find myself itching for more and coming back to play a few hours per week.

You know that this is sort of possible in the game right now... as long as you only have one of each class of ships in a specific fleet. I use this all the time. You can select which ship (of a specific class) in a fleet template should upgrade to and they do that automatically.

If this perhaps is enough for you or not I don't know, but this works good for me until they add a way to have multiple ships of the same class in a fleet at the same time, if they ever will.

Unfortunately, it doesn't really work well with the way I play. I particularly like using large numbers of destroyers of different designs, accompanied by a few larger ships, also with specialized designs. Building designs and fleets according to my plans and needs is half the fun for me. If I want to do that with fleet templates, I'll have to use a big bunch of really small and different fleets, and that's hell to manage. Or compromise and use "regular" fleets, which gets boring for me, hence why I'm taking a break from DW2 currently.

It's my hope they'll implement this change, because it seems a lot of players are feeling it's sorely missing. If they don't implement it soon, then I'll guess I'll have to adjust the way I play, but I feel it would be a shame.

I agree with you to a degree... it have limitations. Especially if you just want one big fleet. But even then you probably can find a design in each class to do what you want in that fleet for the most part unless you just want all of them to be destroyers for some reason.

I tend to run smaller fleets and view them more as small task-groups or squadrons than anything else and for me that works fine until they change this to become more dynamic and free.