4.10 OOS Combat balance

[BigIngrimmsch]

I am wondering about the OOS balance problem, since i dont want to play the game with a not working OOS.
Does someone know if a solution is coming soon?
I could not find any information about that and it is a major problem for the X4 universe to function.

greetings

[paiku]

The only response that i remember is in this topic. But i don't know if they aknowledged the OOS bug with capitals.

[XTC0R]

I would so like to know this. OOS is a topic since release and we all know it will never be perfect. But we are far away from this point.

[baerdesklopapies]

You can try to play a earlyer version of the game.
In Steam: go to the properties of X4. In the Beta Tab, you can select as earlyest version the 3.2.
Not shure about the GOG/Linux versions.

o7

[Xenon_Slayer]

Hi all,

I just wanted to address this briefly before the weekend. Just as with the 'invincible station' issue, I'm going to have to be a bit vague as this is something that's still being addressed and it's difficult to say when it will be up for testing.

First off, thanks for the reports and saves which highlighted some of these issues. As people have mentioned, OOS balancing is tricky and never going to be perfect but (somewhat thankfully) in this case, there were some clear technical issues affecting some very specific situations. Once solved, things already looked much better, especially for the case of capships not dealing the expected damage.

There are a number of other improvements being looked at, and we still welcome any feedback AND saves showing odd situations, just so we can verify we also remedy those.

Cheers,
Owen

[BigIngrimmsch]

Thanks for the statement about the topic.

It seems to be a strange core issue.
XL/L ships are only dealing damage with there main guns. M ships are doing nothing, stations are doing nothing. S ships are killing everything.

I send 20 M ships on a weak SCA destroyer, no one dealt any damage you can watch them for like 20 mins nothing happens.
S ships are flying around my station and not getting damaged. I jump in, 5 seconds later they are dead.

The whole universe development depends on those fights, which makes it really strange right now.


greetings

[Gavrushka]

I'd been away from the game for a while, and wonder if the OOS combat issues are the main reason the economy has grown so sluggish in my game. If battles go unresolved, essentially raging in perpetuity, then the demand for replacement bits also drops off.

Having said that, I sent a fleet of destroyers, carriers and auxiliaries up against a Khaak station in Getsu Fune, thinking no more of it until the video messages 'we're taking hits' started flashing up, and I realised the massive fleet were getting wacked without doing any damage to the station. Withdrawing them would have incurred huge losses, and I have little money for replacements, so I teleported into one of the ships in the sector... I like to think my mere presence inspired the fleet to turn the tide, but perhaps it was more the balancing issues being discussed in this thread. - From 'gonna lose my whole fleet without doing 1% of damage' it went to a good old 'kick in the Khaak's Jacksey' within moments.

But the upshot for me is, I guess my game is populated with faction capital ships locked in stalemate situations because there is no demand for anything much in my game. (Since 4.1 dropped.)

[Lord Crc]

I just wanted to address this briefly before the weekend.

Thanks for the update, much appreciated! Hope we get a new build soon, this is a showstopper.

First off, thanks for the reports and saves which highlighted some of these issues.

IMHO Egosoft really should invest in automated testing, so blatant showstoppers like these don't make it out the door. Given you already have support for different game starts and Linux support, I can't imagine this being too difficult.

We're a small shop which make software with tons of "switches". Manual testing is hard due to combinatorics, and we had bugs constantly ruining the day. After adding a suite with some automated tests of basic functionality things improved a lot.

If you have automated testing, well then improve it.

[BigIngrimmsch]

That should not becomme a rage entry to Egosoft how to do things.
Just pointing out there is a serious OOS problem right now.

[paiku]

I'd been away from the game for a while, and wonder if the OOS combat issues are the main reason the economy has grown so sluggish in my game. If battles go unresolved, essentially raging in perpetuity, then the demand for replacement bits also drops off.

Having said that, I sent a fleet of destroyers, carriers and auxiliaries up against a Khaak station in Getsu Fune, thinking no more of it until the video messages 'we're taking hits' started flashing up, and I realised the massive fleet were getting wacked without doing any damage to the station. Withdrawing them would have incurred huge losses, and I have little money for replacements, so I teleported into one of the ships in the sector... I like to think my mere presence inspired the fleet to turn the tide, but perhaps it was more the balancing issues being discussed in this thread. - From 'gonna lose my whole fleet without doing 1% of damage' it went to a good old 'kick in the Khaak's Jacksey' within moments.

But the upshot for me is, I guess my game is populated with faction capital ships locked in stalemate situations because there is no demand for anything much in my game. (Since 4.1 dropped.)

In my 4.10 game there is a high demand for ships. Stations kill alot of ships, so faction rebuild them. The economy is booming in my game.

[Alan Phipps]

"IMHO Egosoft really should invest in automated testing, so blatant showstoppers like these don't make it out the door."

I'd be fascinated to know how that would be practical and effective for a complex strategy/simulation game such as X4. An automated alternative to live public beta testing - hmmm.

[LameFox]

These issues were coming up in beta anyway, so I do not think it is a matter of needing to automate anything (automate what? how?) but it just needed more time before going live to figure out the specifics.

[XTC0R]

"IMHO Egosoft really should invest in automated testing, so blatant showstoppers like these don't make it out the door."

I'd be fascinated to know how that would be practical and effective for a complex strategy/simulation game such as X4. An automated alternative to live public beta testing - hmmm.

I would imagine something like this guy from star citizen is showing in the following minutes in this video:
https://www.youtube.com/watch?app=deskt ... 0#t=82m55s

[Gavrushka]

In my 4.10 game there is a high demand for ships. Stations kill alot of ships, so faction rebuild them. The economy is booming in my game.

Ah, then it's likely I've just saturated the market by building megastations as a precursor to building a shipyard which I'll keep stocked by L freighters on repeat orders for every needed good. - Thing is, I was relying on a good income from my stations in the interim to fund the blueprint purchases...

[Lord Crc]

"IMHO Egosoft really should invest in automated testing, so blatant showstoppers like these don't make it out the door."

I'd be fascinated to know how that would be practical and effective for a complex strategy/simulation game such as X4. An automated alternative to live public beta testing - hmmm.

Well in this case I'm talking about fairly concrete but important issues.

You could easily spin up a large number of games where you have a certain scenario, like a Teladi L fighting a Split L in an otherwise empty universe, to pick something simple. Then average over those N instances (each with different random seed). If the statistics change drastically after a commit (win %, time to kill etc etc), and you didn't expect that... well time to investigate. Obviously you'd need a large collection of scenarios for meaningful coverage.

Obviously this will take some time to run, but you don't need graphics so the simulations don't have to run at normal speed. For most scenarios a super-simple universe should suffice which again would speed up things. This could be done for non-combat logic as well such as miners or station managers.

It seems to me this should have easily caught things like turrets not tracking properly or other similar issues that has plagued the combat in this game.

For all I know they already have something like this in place. In which case great! Except then it needs some more love, since "blatant" issues still slip by.

And I'm not raging against Ego here. I just sense that Ego is struggling with the same that we did with our software. We are also a very small team writing software with non-trivial logic made even more complex by a large number of settings. We would ship new versions with "blatant", serious bugs that we didn't catch because manually testing everything was impossible. Automated testing has been key to prevent us from drowning.

[Sassbarman]

I'd been away from the game for a while, and wonder if the OOS combat issues are the main reason the economy has grown so sluggish in my game. If battles go unresolved, essentially raging in perpetuity, then the demand for replacement bits also drops off.

Having said that, I sent a fleet of destroyers, carriers and auxiliaries up against a Khaak station in Getsu Fune, thinking no more of it until the video messages 'we're taking hits' started flashing up, and I realised the massive fleet were getting wacked without doing any damage to the station. Withdrawing them would have incurred huge losses, and I have little money for replacements, so I teleported into one of the ships in the sector... I like to think my mere presence inspired the fleet to turn the tide, but perhaps it was more the balancing issues being discussed in this thread. - From 'gonna lose my whole fleet without doing 1% of damage' it went to a good old 'kick in the Khaak's Jacksey' within moments.

But the upshot for me is, I guess my game is populated with faction capital ships locked in stalemate situations because there is no demand for anything much in my game. (Since 4.1 dropped.)

“I like to think my mere presence inspired the fleet to turn the tide, but perhaps it was more the balancing issues being discussed in this thread.”

Haha...don’t sell yourself short I think your a great and inspiring leader.

[Gavrushka]

“I like to think my mere presence inspired the fleet to turn the tide, but perhaps it was more the balancing issues being discussed in this thread.”

Haha...don’t sell yourself short I think your a great and inspiring leader.

I'll take that. It sounds like a very plausible explanation...

[Alan Phipps]

@ Lord Crc: "You could easily spin up a large number of games where you have a certain scenario, .."

Ah, that would probably not appeal much to Egosoft since they discovered the hard way and in hindsight that the cheat engine they were using for setting up and accelerating test scenarios was not providing consistently valid gameplay test results during the Alpha and Beta development cycles for X Rebirth. We all know how well that ended up. Hence they have instead stayed with using public beta live testing of actual play builds for their test and feedback for X4.

[Lord Crc]

Ah, that would probably not appeal much to Egosoft since they discovered the hard way and in hindsight that the cheat engine they were using for setting up and accelerating test scenarios was not providing consistently valid gameplay test results during the Alpha and Beta development cycles for X Rebirth. We all know how well that ended up. Hence they have instead stayed with using public beta live testing of actual play builds for their test and feedback for X4.

They shouldn't need a cheat engine. They should use the actual thing, just without graphics and sound. If they can't run it at accelerated pace, the lack of graphics/sound and minimal universes should lead to low resource usage, so multiple instances could run in parallel (e.g. Docker if they need some isolation). If this doesn't produce the same results, then that's a bug somewhere.

Of course for long-term gameplay and universe evolution, I agree this approach would not be the best. This is for focused testing of certain core functionality.

[Eyeklops]

I don't envy the dev team on this topic. Creating a simulation that gives roughly the same results IS vs OOS given all the variables and meets a strict CPU budget is NO easy feat. I'd love to see how that was coded just for the learning experience.