23 réponses à ce sujet

[Eric Fagnan]

We have been investigating claims that the health and shield gates are not working properly. After some initial testing, we can confirm that both gates are working correctly, but we will continue to investigate whether there are edge cases that cause them to not work in some situations.

With that said, below is a brief explanation of how the gates work and some factors that could make it appear like they are not working.

We have two damage gates in the game. When your shields hit zero, for a small duration you become invulnerable to damage preventing any health damage from taking place. Similarly, when your health gets below 5% you become invulnerable to damage for a short time. The duration of these gates change based on difficulty and they are listed below:
- Bronze 0.5 seconds
- Silver 0.25 seconds
- Gold 0.1 seconds
- Platinum 0.01 seconds

There is also a cooldown on both gate mechanics, preventing you from becoming invulnerable too often. You can only get the benefits of the shield gate once every 4 seconds, and once every 3 seconds for the health gate. This means that if your shields hit zero on Bronze, for example, you become invulnerable for 0.5 seconds, but if your shields get restored through a Geth Turret, a Volus Shield Boost, Stim Pack, or some other means, and your shields go down again after the 0.5 second window and within 4 seconds you will not get a second window of invulnerability.

None of the health and shield gate values or mechanics have changed since we shipped the game.

The following enemy attacks ignore the player health and shield gate mechanics, which means they can instantly kill you.
- Any sync kill
- Enemy grenades
- Possessed Abomination explosions

Some of our new abilities, like the Volus Shield Boost (on the caster only), and Stim Pack, make the player invulnerable for a short duration after the cast, but this is separate from the shield gate mechanic.

Modifié par Eric Fagnan, 12 octobre 2012 - 07:54 .

[Eric Fagnan]

E71 wrote...

Is there even a point to 0.01 seconds? I mean, come on...

The point is that even a massive single shot of damage won't pass through your shield and hit your health, so you can't die in one shot.

[Eric Fagnan]

Asebstos wrote...

Did the period of invulnerability use to be longer?

Has the ROF on enemy weapons been increased at all?

None of it has changed as far as I'm aware, but we are continuing to investigate.

[Eric Fagnan]

Dunvi wrote...

Thank you for your response! And especially for responding to each of the currently active threads - it really is great to know for certain that the devs are aware of our questions.

Do you know how badly the gates may be affected by lag? I definitely feel like I'm taking damage when I'm expecting a second to drop down, and it usually seems to happen a moment late, so maybe there's lag interactions that are causing the feeling of an overall difficulty spike.

I'd also be curious to know what changes were made to fix the vanguard glitch, just out of curiosity.

The host is responsible for all damage logic, and the clients only simulate it, so I don't believe lag should be a factor.

[Eric Fagnan]

frozzted wrote...

Eric, does this have any relation to using an OPS pack and still dying anyway?

I believe that is a seperate issue and is related to dying on the host before your input gets there.

[Eric Fagnan]

ZombieGambit wrote...

How do you explain this vid?

If you look closely there is absolutely no period of invulnerability from either health or shield gate and his health is already below half while he has a sliver of shields.

Does that mechanic only protect against one hit kills from massive damage single shot attacks or something?

What difficulty is the video? And were you host or client?

[Eric Fagnan]

Shampoohorn wrote...


Thanks for the explanation, Eric.

Are the cooldowns the reason why my drellguards frequently die at the end of a charge?  Or are there some other issues at play, like a lack of invulnerability during charge or no post-charge damage resistance?

I'm not a multiplayer programmer, but I believe the Vanguard charge issues are related to our fixes to the Vanguard glitch. To ensure we fixed the glitch, I believe our programmers changed the way Biotic Charge works so that the host is responsible for more of the checks rather the clients having more authority over the move, and like someone said earlier clients may be allowed to do the move even though on the host they died. This does make the power feel less responsive but it does seem to have fixed the glitches. I can't give more details since I didn't work on that fix.

[Eric Fagnan]

Steppdaddy2 wrote...

Eric, while you are active on the forum- i feel like its important to bring up the issue of a volus teammate locking classes out of all action besides movement and melee. A heavy melee breaks this but I'm not sure if this has been addressed.

Btw, thanks for the clarification.

We are aware of this issue.

[Eric Fagnan]

OuterRim wrote...

Yeh, that is bad. His health clearly dropping even with still a sliver of shields. Something is seriously bugged with it and it sure explains many a mysterious death I've encountered.

The shield GUI is deceiving. We saw this ourselves when we began investigating. The damage gate begins before the shield bar is completely gone (you can hear the shield breaking before the blue bar is gone). But after debugging the code everything was working correctly.

[Eric Fagnan]

PyroByte wrote...

UpirNoir wrote...

Eric Fagnan wrote...

ZombieGambit wrote...

How do you explain this vid?

If you look closely there is absolutely no period of invulnerability from either health or shield gate and his health is already below half while he has a sliver of shields.

Does that mechanic only protect against one hit kills from massive damage single shot attacks or something?

What difficulty is the video? And were you host or client?

It is Gold difficulty and the person doing this video was hosting (soloing for demonstration purposes) (At least that is how I remember the guy describing it when he first postet it)

Maybe someone could do another quick demo video just so that we don't have to rely on hearsay information about it. I would do it, but I'm having trouble connecting right now (Not that that would be anything new)

I just created a lobby with the same setup (Glacier/Reaper/Gold) and a Husk managed to damage my health while I still had 5% or more shield remaining.

This is the way our shield and health bars are represented visually. There is a delay between the damage value and the way the bar is drawn.

[Eric Fagnan]

EvanKester wrote...

JSLfromBx wrote...

I know what happened Mr Fragman, please hear me out.

I think the problem is not how the shield gates works nows, the problem is how the gates worked BEFORE the patch.

You say that gold gates only last 0.1 sec and yet everyone who played the game extensively saw the gates in actions and 0.1 s is just too short a time for the human brain to notice, furthermore, I don't remember anyone ever noticing a different behavior betwen the gates on bronze and gold.

My theory is this:  BEFORE the patch, the gates were actually buged, bronze, silver and gold, all setting had a gates lasting 0.5 sec of invulnerability and that's the way we played the game for 6 month. Apparently someone fixed it the the changed somehow sneaked past your own internal documentation of changes.

It explain everything very clearly ,that's why your data say the gates are working fine while every player tell you something changed drastically, well going from 0;5 sec of invulnerability to 0.1 sec certainly makes a tons of different versus ennemies with auto weapon.

If anyone else feel my theory is correct, please do quote me to bring it to Mr Fragman attention, I am sure he has means of testing the old buid of the game on an internal server. And then we can get the final word on this.

This actually sounds like a reasonable explanation of the situation. Not sure if that's quite the case though, but it would explain what we're seeing.

...I definitely miss the old shield gate.

Sadly this is not the case. We tested older builds of the game without the recent patch and content and saw no difference.

[Kristen Schanche]

Hey Guys,

I've spent a good chunk of my day today focused on this issue. I've been running various combinations of old patches and old DLC, and none of my testing has turned up any differences in the shield gates and the delay between them on any difficulty.

[Kristen Schanche]

Khranitel wrote...

Thanks for your reaction, Eric.

Eric Fagnan wrote...

The following enemy attacks ignore the player health and shield gate mechanics, which means they can instantly kill you.
- Any sync kill
- Enemy grenades
- Possessed Abomination explosions

So, as I understand, Praetorian beam do not ignore shield/health gates.
Then how we could explain this moment:
www.youtube.com/watch
1:27. He was downed by Praetorian beam, from about half his shields.

Praetorian beams rapidly hit a number of times in succession, similar to the way the players Particle Beam does.  It's not a single damage attack.  You can see by closely watching in that video that he does hit each of the gates, with an adequate amount of time for the 0.01 platinum gate to tick.  All of my tests today show that issuing damage to the player at that rate produces the same result prior to the patch.

[Eric Fagnan]

Khranitel wrote...

Thanks for your reaction, Eric.

Eric Fagnan wrote...

The following enemy attacks ignore the player health and shield gate mechanics, which means they can instantly kill you.
- Any sync kill
- Enemy grenades
- Possessed Abomination explosions

So, as I understand, Praetorian beam do not ignore shield/health gates.
Then how we could explain this moment:
www.youtube.com/watch
1:27. He was downed by Praetorian beam, from about half his shields.

Behind the scenes, beams are like really fast firing weapons, so a few of the hits will hit the gate, but then afterward it will continue to chew away at your defenses.

[Eric Fagnan]

We've looked through basically all the code and data changes over the past couple months and we can't find any unintended changes that would account for the game feeling more difficult, other than the obvious ones listed in the patch notes and all the new content. None of the enemy weapons, firing behavior, AI, or damage systems have been changed other than what was listed in the patch notes.

[Kristen Schanche]

Deaths by suicide generally occur when the thing that killed a player was generated by a source that's no longer available. An example of this would be an Atlas fires a rocket, before the rocket connects with the player, the Atlas is destroyed. The rocket then connects and scores a kill. That can trigger a 'suicide' death in the ticker.

As for the main thrust of the thread, I'm certainly planning to dedicate most of my weekend to multiplayer, I gots me some challenges to get, so I'll be keeping an eye out as well for odd behavior. It's certainly true that live play can find issues that testing doesn't, but that said right now I'm still pretty confident in the tests we've run, and I don't see anything that indicates that the behavior is any different then it's always been. I'm very curious to find out what it is that's causing this perception though.

Modifié par Chris Schanche, 12 octobre 2012 - 10:48 .

[Kristen Schanche]

Farosyrn wrote...

Chris Schanche wrote...

Hey Guys,

I've spent a good chunk of my day today focused on this issue. I've been running various combinations of old patches and old DLC, and none of my testing has turned up any differences in the shield gates and the delay between them on any difficulty.

And we all really appreciate it, thanks!
But, numbers aside, don't you feel like you're dying more easily when you play the game, even if you can't quite put a thumb on it yet?

I've been playing quite a bit at home and that hasn't been my perception no.  I generally play low or no fitness caster type builds though so I'm used to blowing up fast if I step out.

[Kristen Schanche]

JaimasOfRaxis wrote...

The tests are much more obvious when using non-casters.

There's a giant spike in the number of squad wipes I've seen and it's with my usual team, so I know it's not a skill issue. Breached shields now result in instant death, so very clearly, something is off.

I was providing an anecdote from my home play.  Our testing today to find out if anything is happening here has been across a variety of kits with varying health and shield pools, on all difficulties.  I cast a pretty wide net trying to track down any changes here.

[Kristen Schanche]

I've been playing all evening with a number of people, and I'm just not feeling the difference. I'll still be keeping an eye on it, and I will try to run some more tests when I get back into the office on Monday. But we havn't changed anything, there hasn't been a stealth hotfix, and Eric and I have been pretty through in testing the shield and health gate timers.

I spent part of my afternoon going through and stepping back patches one at a time and running time checks, and the gates are giving their expected duration every time. I'm not closed to the idea that there could be an edge case here or certain conditions that cause this, but so far we havn't found it. And I would hope it goes without saying that we have no reason to lie about this. If there's an issue here and we find it, we'll be up front about it.

[Kristen Schanche]

Mabari Owns High Dragon The 2nd wrote...

Chris Schanche wrote...

I've been playing all evening with a number of people, and I'm just not feeling the difference. I'll still be keeping an eye on it, and I will try to run some more tests when I get back into the office on Monday. But we havn't changed anything, there hasn't been a stealth hotfix, and Eric and I have been pretty through in testing the shield and health gate timers.

I spent part of my afternoon going through and stepping back patches one at a time and running time checks, and the gates are giving their expected duration every time. I'm not closed to the idea that there could be an edge case here or certain conditions that cause this, but so far we havn't found it. And I would hope it goes without saying that we have no reason to lie about this. If there's an issue here and we find it, we'll be up front about it.

Are you playing with perfect connections though? Thats the issue. Any sort of lag renders the invulnerability useless.

A very large number of the callouts both in this thread and on the forums are claiming these issues with Host as well.  I'm using my home connection to play at the moment though, standard cable package, so not overly remarkable in either direction.

[Kristen Schanche]

There's a lot of conflicting information in this thread, some people seem to think it's health gate only, some both, some that there isn't an issue and some that there is. We're still keeping an eye on all cases as we play this weekend, but as I've said before, I ran checks against this without this patch, then without the previous patch, then without the patch before that, and the gate timer was the same across the board.

We're not dismissing these concerns, but Eric spent a lot of time going through everything we changed in this patch that could have touched this and couldn't find anything, and I still can't see any evidence that these gates don't work the way they do before the patch.

[Eric Fagnan]

EvanKester wrote...

Brutes have always hit through shield gate, since their melee swipe can actually hit multiple times. There's nothing new about that.

I would have to double check when I'm back in the office but I believe Brutes do hit multiple times or else they wouldn't be much of a threat. 

[Kristen Schanche]

Qui-Gon Glenn wrote...

Are the testers doing there testing on LAN inside the office? With minimal/near perfect lag/latency?

I would like to know if testing is being done... internationally, ie teaming up Bioware Edmonton and Austin testers perhaps. This game is international, and loooooong pings affect the players of this game greatly; I am curious to see if the testers intentionally subject themselves to the same playing conditions as the majority of their player-base or if they test only in lab perfect conditions.

I have been a beta-tester. When I started testing, I didn't have a clue what to look for... I would not have thought to test every power against every enemy in every difficulty back then. The current testing crew obviously is hitting about my rookie average, as they certainly did not test Shadow Strike against all enemies in all conditions, as the admitted bug on Atlas' proves. I notice issues with SS against Scions as well, and it is generally less useable than pre-patch; mostlikely an issue of how the host PC now handles all the logic for that stuff (biotic charge, etc).

Breaking abilities to fix things.

Everything that is reasonable to test within the available time constraints is tested.  This does mean that sometimes things are missed.  The Shadow Strike issue you bring up was caused by a last miniute emergency fix that did not have time for a full QA pass after implementation.  These kinds of things happen during development and they are unfortuante, but rarely reflect on the abilities of the QA teams.

I'm going to reiterate for the last time in this thread, a very large number of users experienceing this issue are reporting it as a host issue, that is not latency or lag related.  While we do run latency checks as well, the reports coming from players suggest we should be able to see this across the board.

Modifié par Chris Schanche, 13 octobre 2012 - 07:54 .

[Eric Fagnan]

Hey everyone, the team does feel the game has become more difficult with the latest content and patch and we think there are a lot of reasons for this, but we don't believe it is the health/shield gate issue. We will continue to investigate the gates to see if we've missed any edge cases, but at this time we are working on other solutions to the jump in difficulty.