Larian Banner: Baldur's Gate Patch 9
Previous Thread
Next Thread
Print Thread
Page 2 of 8 1 2 3 4 5 6 7 8
dwig #759795 26/02/21 03:05 AM
Joined: Jan 2021
member
Offline
member
Joined: Jan 2021
GoG Galaxy froze up at 92%, I killed it and reopened it (usually works when it freezes) and it restarted the download from scratch, got to 3% and froze again... Can someone tell me what patch version # this is so that I can go to GoG.com and make sure it is there for me to manually download? I'm showing they have version 4.1.90.6165 as the latest, but I don't think that is correct.

Joined: Dec 2020
R
stranger
Offline
stranger
R
Joined: Dec 2020
I read the patch notes and there was no clear indication that they fixed the elusiveness of companion scenes during camp (e.g. Weave scene from Gale, Astarion star scenes, etc); In past iterations of the game, many players reported not being able to properly trigger/having difficulty in triggering companion specific scenes when calling it a night at camp; A lot of save scumming, camp spamming, and excessive experimentation was required from players in past patches as the game seems to only trigger companion scenes under very particular instances which made the chunk of the side story miss-able. It was also possible to finish early access by only using camp a couple of times, so this was reported by a number of players as they were missing a lot of story beats that would have made their playthroughs richer in quality.

I know the patch was only out a couple of hours ago, but would the speed runners among us care to share their experience on this past issue and advise whether this was addressed by Larian in patch 4? I believe they also removed the Tiefling camp party romance scenes during a hotfix last year; Was it already added back with better cinematics?

Im trying to avoid getting burnt out by the game during EA so I ideally would want to make a new PT only when this problem is addressed by Larian.

#759801 26/02/21 03:41 AM
Joined: Dec 2020
R
stranger
Offline
stranger
R
Joined: Dec 2020
I read the patch notes and there was no clear indication that they fixed the elusiveness of companion scenes during camp (e.g. Weave scene from Gale, Astarion star scenes, etc); In past iterations of the game, many players reported not being able to properly trigger/having difficulty in triggering companion specific scenes when calling it a night at camp; A lot of save scumming, camp spamming, and excessive experimentation was required from players in past patches as the game seems to only trigger companion scenes under very particular instances which made the chunk of the side story miss-able. It was also possible to finish early access by only using camp a couple of times, so this was reported by a number of players as they were missing a lot of story beats that would have made their playthroughs richer in quality.

I know the patch was only out a couple of hours ago, but would the speed runners among us care to share their experience on this past issue and advise whether this was addressed by Larian in patch 4? I believe they also removed the Tiefling camp party romance scenes during a hotfix last year; Was it already added back with better cinematics?

Im trying to avoid getting burnt out by the game during EA so I ideally would want to make a new PT only when this problem is addressed by Larian.

dwig #759810 26/02/21 04:02 AM
Joined: Dec 2020
veteran
Offline
veteran
Joined: Dec 2020
I'm at 8 - 10 MBS with steam and i have pretty terrible internet

Last edited by Boblawblah; 26/02/21 04:03 AM.
#759820 26/02/21 05:15 AM
Joined: Jan 2021
addict
Offline
addict
Joined: Jan 2021
I started playing through patch 4 today to document dice rolls to see what weighted dice distribution looks like.

However, I noticed I was having more fun in combat. Things felt more free, more fluid. I noticed something different with high ground on the stairs of the deck of the nautiloid. So I chose to test out what I was seeing, once I got to a location with more verticality.

Here are the findings grin

For a baseline of understanding. Here is how backstab appears and is documented in game.
Backstab UI
Backstab in the Combat Log

Notice that Advantage is clearly displayed with the roll.

Now notice the UI when shooting an arrow from low ground
Smallest of Low Ground
Now look at the attack roll in the combat log grin Disadvantage was not applied.
The Combat Log

Comparing this to higher verticality for a sanity check. Higher verticality is working as expected.
Low Ground UI
Result in the Combat Log

It's really cool, it looks like the issue with the smallest inflection of low ground causing disadvantage has been not part of the game's long term design. There is the start to a fix in Patch 4, but the UI isn't ready yet. I'm ecstatic, I'm now looking forward to future combat. So I chose to do even more testing to see if the change is consistent.

Where disadvantage is applied
Verticality Test 1

Where the player gets a normal attack roll
Verticality Test 2

I then went on to test if some changes to the UI were already in the game
Testing with right-clicking off the spell to reset the UI

Note where I am standing for this attack preview, and compare to Verticality Test 2
EDIT: Forgot to add this to the original post
The attack in Verticality Test 2 is performed at a lower area from where I am standing.

this last section that came up in the testing is an indication, that high ground was always intended to have a verticality threshold to apply Disadvantage/Advantage.

I'm now keeping an open mind as I play through patch 4. Combat has been more enjoyable and fluid without being forced to avoid disadvantage on ranged attacks every battle. I'm going to use patch 4 to reset my opinion on how I feel about High/Low Ground Advantage/Disadvantage in Baldur's Gate 3.

This change may seem small, but is has made a good change to combat. This is a step in the right direction and I hope the UI will be patched soon to reflect these changes!

Originally Posted by DragonSnooz
Edit: I've chosen to do a thorough review of distributions for rolls shown in the UI as Low Ground, but get rolled at normal. And rolls shown as normal that the combat log rolls as high ground.

Here is the data for "UI lowground & Combat Log Normal":
I used the bugbear assassin and scorching ray to test.
UI for Scorching Ray, Gale's position relative to bugbear
Scorching Ray 1
Scorching Ray 2
Scorching Ray 3

I did use different sections of the bugbear, stomach, head, shoulders, etc. The UI and Combat Log were consistent in what they showed.
Here is the data after 156 scorching rays on the bugbear
Average: 7.8
Standard Deviation: 2.5
All frequencies were within two standard deviations.

The important takeaway is the shape of the histogram, it doesn't match what we would expect for disadvantage, it is a lot closer to what we would expect with normal attacks.

The surprises are how low 6, 7, 15, and 18 are but it would also make sense to weight the dice against four outcomes, to increase the change of 10 and 11. (Their frequencies were still within a normal distribution.)
***
Here is the review of attack rolls where "UI doesn't not show High Ground or any advantage, but the combat log has awarded advantage on the attack roll".
The frequency is in line with what we would expect to see with attacks rolled with advantage.
(This was done at the save file back at the Dank Crypt, N=42)

The combat log is showing what is actually happening in game.

Last edited by DragonSnooz; 28/02/21 12:55 AM.
Joined: Jun 2020
veteran
Offline
veteran
Joined: Jun 2020
I can't confirm for certain, but I don't suspect it has been, otherwise they would surely have mentioned it as something notably more important than telling us in three separate places that they fixed how a goblin pees.

Joined: Dec 2020
old hand
Offline
old hand
Joined: Dec 2020
This is actually pretty major. But I can see why the devs didn't mention any of this in the patch notes, because attempting to describe this at all would have been confusing as hell. If I'm reading this post right, your last picture is showing that the attack would be at disadvantage from where you're standing in said picture, but the attack was performed at a slightly lower angle in the 'Verticality Test 2' picture, and the actual attack roll from there didn't have disadvantage like the other picture said it was supposed to have?

I hope there is something Larian can implement in the UI to show the player where they can consider moving with their remaining movement points in order to remove disadvantage/gain advantage with ranged attacks. If this is here to stay, we're going to need to be informed better than to risk using our remaining movement points moving to an area that only looks like it might be beneficial to the naked eye, only to find out that it really isn't.

Last edited by Saito Hikari; 26/02/21 07:06 AM.
Joined: Jan 2021
addict
Offline
addict
Joined: Jan 2021
Originally Posted by Saito Hikari
This is actually pretty major. But I can see why the devs didn't mention any of this in the patch notes, because attempting to describe this at all would have been confusing as hell.
It was a challenge writing the original post, it really needed screen captures to make any sense.

Originally Posted by Saito Hikari
If I'm reading this post right, your last picture is showing that the attack would be at disadvantage from where you're standing in said picture, but the attack was performed at a slightly lower angle in the 'Verticality Test 2' picture, and the actual attack roll from there didn't have disadvantage like the other picture said it was supposed to have?

Yes, all across the staircase attacks were rolled at normal not disadvantage. Verticality test 1 is where the game is rolling the attack at disadvantage.

The attack rolls are consistent across testing. What is inconsistent is the UI, the UI has a tendency to show what we experienced in patch 3 but sometimes it will correctly show that disadvantage will not be applied on the attack roll.

Originally Posted by Saito Hikari
I hope there is something Larian can implement in the UI to show the player where they can consider moving with their remaining movement points in order to remove disadvantage/gain advantage with ranged attacks. If this is here to stay, we're going to need to be informed better than risk using our remaining movement points moving to an area that only looks like it might be beneficial to the naked eye, only to find out that it really isn't.

Agreed, it took a bit of testing to find out what verticality it will and won't apply. The staircases were nicely sized to use for testing, I'm wondering how this will work in the Underdark at the moment. Hopefully this is consistent through the rest of patch 4.

Joined: Dec 2020
old hand
Offline
old hand
Joined: Dec 2020
Maybe you could amend this section of the original post like so, to get the point across in a less confusing manner?

I had to review the pictures like 3 times each before I finally understood that what you are saying is that the UI is now showing us inaccurate indications for exactly where the threshold for attacks to gain advantage/disadvantage are. The thresholds may have actually changed in Patch 4 to require a much larger difference in high/low ground in order to gain advantage/disadvantage, but the UI is operating under Patch 3 standards when calculating advantage/disadvantage and chance to hit.

Originally Posted by DragonSnooz
Where disadvantage is applied
Verticality Test 1

Where the player gets a normal attack roll
Verticality Test 2

Note where I am standing for this attack preview, and compare to Verticality Test 2
The attack in Verticality Test 2 is performed at a lower area from where I am standing.

Last edited by Saito Hikari; 26/02/21 06:57 AM.
Joined: Jan 2021
addict
Offline
addict
Joined: Jan 2021
I've edited the post, thank you for the suggestion.

#759861 26/02/21 08:01 AM
Joined: Oct 2014
K
stranger
Offline
stranger
K
Joined: Oct 2014
First impression after trying out , I like the work they did on the viusal, it's really clearer and more beautiful. I did not expect too much because I'm not that sold on druid , I started a new game with two custom chars , a druid and a warlock meant to be 'evilk' and so far simply from the look and feel . I got hooked again.

dwig #759863 26/02/21 08:13 AM
Joined: Jan 2021
member
Offline
member
Joined: Jan 2021
All this time later, I have managed to find a decent work around that has gotten me working (in Vulkan only)... and I now have a question. Has anyone figured out how/where to use the Pod Key?

Joined: Dec 2020
old hand
Offline
old hand
Joined: Dec 2020
Well, I'm going to post an inquiry on Reddit, because this is something interesting enough to investigate for multiple reasons. See if I can get some more useful examples from there, even if there are maybe only a handful of people there today who would actually be any interested in this kind of stuff. I will be doing some testing in the tutorial myself using an archer.

I especially want to figure out if the same applies to the advantage half of this, which I think the tutorial might provide a few good places to test this. It would be really bad if the UI was showing people having advantage in the attack preview but the actual attack is being rolled normally, and it'd be interesting to see if the AI is subject to making the same mistake themselves. (As in, will they know the correct height to get advantage, or will the AI get fooled into simply thinking they are?)

https://www.reddit.com/r/BaldursGate3/comments/lst8ca/there_appears_to_be_an_undocumented_change_to/

Last edited by Saito Hikari; 26/02/21 09:59 AM.
Joined: Dec 2020
old hand
Offline
old hand
Joined: Dec 2020
Okay, I can tell you SOMETHING is happening. And I don't understand WHAT. My mind hurts trying to make sense of this at all.

Test 1:

- UI preview shows this as a normal roll.
- Slightly moving the mouse downwards on the enemy model, with my character standing in the SAME SPOT, suddenly previews as an advantage roll???

(Fun fact: The same principle that I imply above could actually be used to great effect in DOS2. If you tried to aim at an enemy but the preview showed that your arrow would be blocked by a narrow pillar or something, you could try ‘fishing’ around the enemy model until you find the pixel perfect angle for the game to let you snipe at them. Even if it was a foot barely sticking out.

This principle made it easier for Dwarves to avoid line of sight from enemy archers, while simultaneously making them the worst archers in theory. But only slightly.)

Test 2:

- I moved slightly downwards to a point where the UI tells me I would not have advantage while attacking.
- The actual attack roll tells me I performed the attack at advantage anyway.

So it's hard to draw any conclusions as to what, at all, actually changed. Only that the attack preview can't really be trusted now when it comes to making attacks from height around the thresh holds. I may need to find a different non-hovering target. But that will have to wait until after I come home from work in another half a day.

I am getting very close to declaring this as a mostly UI bug, but it does feel like something did actually change at the same time. In order for us to even conclude that the actual breakpoints actually did change at all, we'd need to find someone still playing in Patch 3 to make a direct comparison with. I would roll back to the previous version myself if I have to, once you find a suitable location that I can easily access.

...

Well, looks like I'm staying on these forums a bit longer, because chasing down stuff like this is a legitimately interesting change of pace from the usual doomposting arguments. I guess now is as good a time to say this again, and you all know I've never used this phrase in a joking ironic manner.

Programming is hard.

Last edited by Saito Hikari; 26/02/21 09:32 AM.
Joined: Oct 2020
Location: Liberec
veteran
Offline
veteran
Joined: Oct 2020
Location: Liberec
Who the hells is Zelkor? O_o


If my comments bother you, there is nothing easier than telling me to stop.
I mean ... I won't ... but it's easy to say. wink
Joined: Jun 2020
veteran
Offline
veteran
Joined: Jun 2020
If it helps with the investigation (I've got other things I'm focused on right now, and also struggling with... eh... let's call it motivation):

In D:OS2, where you targeted on the enemy made a very important difference, since it measured how far away things were, and line of sight, based on your exact cursor point. Especially for skills that took their damage modifier off how far away your target was - you would often get substantially more damage if you could wrangle your camera/cursor to target the furthest away pixel on your enemy, and often whether you could hit something or not was a fight-the-ui game of seeing IF there was a pixel on the enemy that it would give you a clear line for.

If they have lifted the majority of their code base from that (they have), then it's very likely that it's measuring distance based on your exact cursor position on the model, so in the case where you and your target are close to the right distance/elevation apart, it will change based on where you aim.

Last edited by Niara; 26/02/21 09:29 AM.
Joined: Dec 2020
old hand
Offline
old hand
Joined: Dec 2020
Yeah I’m beginning to suspect that the break points did not actually change.

It seems far more likely the UI hasn’t been entirely honest to us this entire time, maybe even since launch, and it took all the way until today for someone specifically testing the break points to notice.

Joined: Dec 2020
veteran
Offline
veteran
Joined: Dec 2020
Originally Posted by RagnarokCzD
Who the hells is Zelkor? O_o
I guess, they mean Zevlor?


"We are all stories in the end. Just make it a good one."

Doctor Who
Joined: Dec 2020
old hand
Offline
old hand
Joined: Dec 2020
https://www.reddit.com/r/BaldursGate3/comments/lssfm5/so_a_goblin_shoved_my_flaming_sphere_into_the/

... I am not gonna lie, this is actually pretty hilarious.

It feels like Larian turned on the shove button for everything, specifically to get feedback on it. Going to be interesting to see how the wider community reacts to this.

Joined: Jun 2020
addict
Offline
addict
Joined: Jun 2020
Yeah - you can see the improvements each patch - coming along nicely

Page 2 of 8 1 2 3 4 5 6 7 8

Link Copied to Clipboard
Powered by UBB.threads™ PHP Forum Software 7.7.5