|
Author |
Thread Statistics | Show CCP posts - 4 post(s) |
Aeon Amadi
A.N.O.N.Y.M.O.U.S. League of Infamy
2150
|
Posted - 2013.07.20 07:46:00 -
[1] - Quote
Edit: Posted in General Discussion as its not necessary a Bug. Yet.
Tests were conducted on a domination match on North American servers (both testers are resident) on Skim Junction.
Precursor:
This is an issue chalked up and reported here: https://forums.dust514.com/default.aspx?g=posts&t=92132&find=unread
The purpose of this thread is to lay down the ground work for a testing base contributing to this phenomenon. As such, we've ruled out two direct concerns previously brought up.
A.) That this is a map error in Spine Cresent. Common belief is that this phenomenon only occurs in a small gap at objective Alpha on the bridge. Videos show this is not the case.
B.) That this is associated with Lag. Being as we were on North American servers, this should not have been the case and even in a high latency environment testing would inevitably come out the same due to testing procedures.
What we concluded is that while the hit box may not grow larger - something CCP states there are mechanics that could possibly occur in this manner - something IS happening to prevent the Scouts adequate access through previously passable locations when they sprint. More so than a Heavy, as we'll get to in a later video.
Test A - Scout with base movement speed
http://m.youtube.com/index?client=mv-google&rdm=ticq73hf#/watch?v=DjRe7iWl22A&feature=plcp
Test B - Scout with Two Enhanced Kin Cats
http://m.youtube.com/index?client=mv-google&rdm=ticq73hf#/watch?v=EQ5kOiqEYG4&feature=plcp
Test C - Scout with Two Militia Armor Plates
http://m.youtube.com/index?client=mv-google&rdm=ticq73hf#/watch?v=IfUOdBn6Qt8&feature=plcp
Test D - Heavy with One Enhanced Kin Cat
Warning: Includes crude speech from recorder due to disbelief as the Heavy's elbows barely clipped through the bumpers.
http://m.youtube.com/index?client=mv-google&rdm=ticq73hf#/channel/UC1DQd-dmONpNZJfO8fnnJGA?feature=guide
In Conclusion
While this does not necessarily prove that a Sprinting Scout is easier to hit due to hitbox increase, it does invariably prove that ALL suits are less agile when sprinting as the collidable objects are very much more noticable. Turning corners, trying to squeeze through tight spaces and running past players/objects is difficult to manage when sprinting because of this oddity, whatever the case may be.
So does sprinting decrease a Scouts survivability? Thats still up for debate. If you guys have any other thoughts on how to more accurately gauge this, feel free to suggest something to us and we'll do what we can. |
Aeon Amadi
A.N.O.N.Y.M.O.U.S. League of Infamy
2151
|
Posted - 2013.07.20 07:54:00 -
[2] - Quote
Xocoyol Zaraoul wrote:I think it might be due to the arms sticking out more during the run animation, might just be me.
Animations usually dont have much affect on collision detection in that manner, but its possible and it could very easily be perceived as an increase in hit box detection. Something we'll test later.
Even so, probably shouldnt be more than a heavy on the outer edges |
Aeon Amadi
A.N.O.N.Y.M.O.U.S. League of Infamy
2152
|
Posted - 2013.07.20 11:09:00 -
[3] - Quote
CommanderBolt wrote:You went to the trouble of setting up all this? Fair play my friend.
I can speak from experience when I thought I have missed somebody when shooting, only for them to try and sprint away and die. It seems magic bullets curve and follow the enemy .... :S
(Happens sporadically but quite rare)
Yes we did, and were willing to go much further provided theres questions to be answered |
Aeon Amadi
A.N.O.N.Y.M.O.U.S. League of Infamy
2153
|
Posted - 2013.07.20 11:56:00 -
[4] - Quote
Nack Jicholson wrote:Why are the links to mobile youtube?
I dont have a PC, had to upload with my phone. |
Aeon Amadi
A.N.O.N.Y.M.O.U.S. League of Infamy
2154
|
Posted - 2013.07.20 12:36:00 -
[5] - Quote
Django Quik wrote:I think I recall a dev at some point noting that hitbox is not the same as the collision detection... so, you could well be correct in that the environmental collision detection on the models increases but not necessarily the hitbox. That's just a possibility - I'm not denying your conclusions, just playing a little devil's advocate.
In either case, it's definitely something that needs some deep dev investigation.
Interesting. Still, increased COLLISION detection shouldnt occur from sprinting, at least in my opinion. If we're sticking with this fast paced theme then weneed to be able to move without outside influence getting in the way. |
Aeon Amadi
A.N.O.N.Y.M.O.U.S. League of Infamy
2164
|
Posted - 2013.07.20 22:44:00 -
[6] - Quote
Iron Wolf Saber wrote:Collision Model =/= Hit box.
Right, we've brought this up in the original post. Either way, its proven that its more difficult to sprint through an area with collidable objects. |
Aeon Amadi
A.N.O.N.Y.M.O.U.S. League of Infamy
2171
|
Posted - 2013.07.21 01:35:00 -
[7] - Quote
Shrapned wrote: hope you CLUELESS idots realize that ALL Hit-Boxes were made to be ALL the SAME NO Matter what Suit your using now. CCP did this with the patch-release of the Commando suit.
ALL hit-boxes are NOW the Same and the Standard is the Assault suit or "Medium" category in the Market tab.
The Heavy suit had there Hit-boxes REDUCED.
The Light suit had them INCREASED.
I always wonder how people become so alphanumerically violent and Ive chalked it up to bad parenting. or schools. dunno.
Either way, calm down and post constructively so that we know who the hell youre talking to, then post the evidence of your theory. kthnxbai. |
Aeon Amadi
A.N.O.N.Y.M.O.U.S. League of Infamy
2171
|
Posted - 2013.07.21 02:18:00 -
[8] - Quote
Quick update with some more findings.
Query: How much room does a Scout need to sprint through a gap? Test Method: Used a compact nanohive (2.0m radius) to judge distance between two LAVs and steadily increased the gap width until the Scout could successfully pass through undetered.
Results: 1.0m clearance is required for sprinting th as little as a 0.5m space for walking. With this we can assume adequate space needed, when sprinting, on either side, is a 25% increase approximately. Sprinting causing a noticable increase in collision occurrences and, if space is tight, will cause the player to 'glide' along collision edges or outright stop if approached head on.
Query: Is there a difference in necessary clearance for a Heavy, as compared to the results fromthe Scout? Test Method: Same instances as previous testing, simply changed suit.
Results: Clearance necessity was equal to that of a Scout and all suits therein. A Heavy cannot clear an object that a Scout cannot and vice versa. Data is inconclusive as results were largely thesame.
Query: Does this phenomenon attribute to the notorious map issue(s) in which the player is caught in 'quicksand'? Test Method: Area in map where player gets 'snared' was located and tested with variety of suits with variety of conditions.
Results: This was found to be almost entirely due to small hills having sides that are too steep, rather than invisible objects colliding with the player. While tests are inconclusive as we were unable to test all 'map snares', we are convinced this is unrelated.
Testers: CeeJ Mantis, Harlequin 13 and mr mustard
|
Aeon Amadi
A.N.O.N.Y.M.O.U.S. League of Infamy
2171
|
Posted - 2013.07.21 02:37:00 -
[9] - Quote
Maken Tosch wrote:I have a quick question.
After re-reading the OP, how were you able to determine that the hit box doesn't increase with the collision box?
We didn't, I carefully suggested that it's still under the maybe category as the only way to test is live fire. We did, however, prove it was not related to a certain map or object and it applies to everything in the environment. |
Aeon Amadi
A.N.O.N.Y.M.O.U.S. League of Infamy
2195
|
Posted - 2013.07.23 00:27:00 -
[10] - Quote
Shotty youshould check ojt my other thread CCP: Very Important that you read this
Think youmight like what you see |
|
Aeon Amadi
A.N.O.N.Y.M.O.U.S. League of Infamy
2196
|
Posted - 2013.07.23 01:20:00 -
[11] - Quote
Updated Report:
Test was conducted with Mr Musturd as the potential scout victim.
Test consisted of using nanohives spread out five meters apart to form a 25 meter track for scout to run. Subject was then fired upon with a Tactical Assault Rifle while aiming down the sight. Subject was asked to walk for preliminary test, then to sprint. Test was conducted at a variety of ranges and angles.
Results proving hitbox expansion were negative, as firing to the sides of the scout from a forward angle showed no width expansion what so ever and even shots that would be considered 'grazing' do not register. Firing at the scout from a side angle however revealed what could be considered 'hit box expansion' to the untrained eye.
I believe what is actually occuring is that the hit box lag (the hit box being behind the subject, rather than directlt on him) is actually increasing with speed. Hit markers were not showing when firing directly on the target as he sprinted, but 2-3m behind him proved best results. The distance of the hit box correlates to the speed of the model, with higher speeds expanding the gap between the model and hit box rather than the hit box itself.
Because of this, there is more leeway for inaccurate shooters who arent turning fast enough to land shots directly on target but BEHIND the person, or even doing so intentionally. Turning corners also shows this to be true, especially at high speeds.
This, unfortunately, is something entirely unavoidable as it is associated directly to the latency - not of the client, but the server providing the information.
Feel free to make comments, but the only way to solve this issue is to somehow decrease the gap between model and hit box when moving at high speeds. While it may do more harm than good for the game asa whole, the only way I can see of doing this is to actually place the hit box INFRONT of the model when moving at higher speeds.
tl;dr - The faster you go, the farther your hit box lags behind you. When that occurs, the shooter doesnt have to aim directly on you and if he did so he would miss because your hitbox isnt able to catch up with your model.
IMPORTANT EDIT: This DOES NOT answer the increased collision detection as mentioned in the original post. I feel this is anentirely separate issue. |
Aeon Amadi
A.N.O.N.Y.M.O.U.S. League of Infamy
2204
|
Posted - 2013.07.24 03:51:00 -
[12] - Quote
Shotty GoBang wrote:Aeon Amadi wrote:Updated Report: Test was conducted with Mr Musturd as the potential scout victim. Hi Aeon, These aren't the results I had expected to see, but I cannot argue with your method. Your test results explain why we're taking damage after reaching cover. Thanks for clearing this up, and great job once again. So the sudden decrease to scout survivability in 1.2 resulted primarily from a performance improvement that boosted hit detection, right? I doubt CCP will rollback a performance improvement, so I wonder if they'll give us back our smaller hitbox now . - Shotty GoBang
As Ive previously explained to others outside ofthis forum I have no idea how it could be fixed because all its really doing is rewarding players who arent aiming on target, but in all actuality missing by a degree associated with how fastthe target is moving. Basically, youre hitting because youre missing, whether that be intentional or not. Ive noticed a massivve jump in my KDR since discovering this though, landing 5-6 KDR games rather frequently. |
Aeon Amadi
A.N.O.N.Y.M.O.U.S. League of Infamy
2211
|
Posted - 2013.07.24 17:35:00 -
[13] - Quote
Shotty GoBang wrote:Aeon Amadi wrote:Ive noticed a massivve jump in my KDR since discovering this though, landing 5-6 KDR games rather frequently. Hey Aeon, When we last spoke (July 8th), you had switched from Scout to medium frame following the initial Uprising nerf(s). Are you back to running Scout? - Shotty GoBang
I graduated to Assault at the beginning of Uprising because of a number of reasons.
1- Getting a working scout requires INTENSIVE skill specialization as even using basic kinetic catalyzers requires Biotic Upgrades 5. I disagree with this as it severely limits early scout builds.
2- Im convinced that there is nothing that a scout can do that someone else cant do better and have actually challenged some CPM members and other players in Skype to give examples. Minmatar Logi is far better suited for hacking, Caldari Assault far better suited for CQC etc. There are a plethora of reasons for this and one being that profile dampeners areuseless to visual marking.
3- I love the Gallente and when they announced the assault variant I fell in love.
4- Im an armor tanker at heart.
I had originally gone scout out of a tester's standpoint, playing the most skill intensive role to submit feedback as I had previously done with Heavies and Logistics when they were underpowered. However, now that the game is live, Ive focused my efforts on what I specialize in first and testing second, though I sympathize with Scouts because I understand the hardships.
Some of the ideas Ive suggested (despite them being unanimously disapproved of) was to give Scouts a specific role. Originally I suggested Cloaking be restricted to Scouts only to give them a definitely different playstyle over any other class much like Heavies and their use of Heavy Weapons/Dual Light Weapons.
Another suggestion was to reduce the resource cost on Drop Uplinks and make them Scout specific to force incentive on thought behind placement, but there are a number of difficulties with this in retrospect. A. They need to reduce their visibility to pre-uprising astheyre just too easy to find. B. Anyone who specced into them that doesnt run Scout - rather, everyone in PC - would have a conniption fit.
There are a lot kf changes to Scouts needed to make Scouts a viable class and I think a good place to start is to enforce a standpoint that precision must beat profile EVEN WHEN AIMING DOWN THE SIGHTS as theres been times where a Scout was suddenly detected by a Heavy because a Sniper looked in that general direction. Scanners need a rework as even scouts with all level five skills cant get beneath their radar (despite no one using them) from Advanced and up.
Finally, I would like to bring up homogenization of the Scouts between variants as the Gallente has some weird slot layouts between Advanced and Prototype.
Anyway, rant over. Back on the issue of hitbox/collision, Id like to remind you all that there is not an increase in hitbox size, so far as I know. There is a discrepancy with collision while sprinting that I feel needs to be ironed out but all of the issues with thehitbox are still correlating to my theory that speed = increased hitbox lag. |
Aeon Amadi
A.N.O.N.Y.M.O.U.S. League of Infamy
2211
|
Posted - 2013.07.24 17:41:00 -
[14] - Quote
Lucifalic wrote:+ 1 and really great work on these tests. Like others I experienced a serious decline in my scouting survivability and hit detection with my shotty. I was aiming at people but I'll try aiming a bit behind and see if the large holes in their body appear. I hope they sort this out. With speed being my preferred play style I feel pretty crippled in 1.2 but have a hard time leaving my hard level scout.
Bear in mind its on a case by case basis. Theres a lot of variables to consider.
- Target frame. A light frame will be all over the place, a heavy not so much. - Target speed. The faster they move the greater the distance their hitbox is from their model. This will change between anarmor and a shield tanker. - Latency. The server has to catch up with their actions before it sends you the information. If theres poor latency, try to predict where they WERE not where they ARE. - Understand that this only applies when they are moving perpendicularly as opposed to parallel. If theyre running away from you or toward you, just aim at their model and let the game handle the rest. - Most importantly, remember that all of these rules apply to you as well. |
Aeon Amadi
A.N.O.N.Y.M.O.U.S. League of Infamy
2214
|
Posted - 2013.07.24 18:23:00 -
[15] - Quote
Django Quik wrote:Aeon's findings explain why people have been complaining about being hit and killed even after already getting behind cover but this is not an experience new to Uprising; it's something people I know have complained about as far back as Codex.
Its an issue that is well known in FPS games. Google 'hit box lag' and youll see some interesting articles/pictures.
The issue unique to Dust is that its pretty much the only FPS game that has hit box lag associated with variabled speed because... Well... Its the only FPS with variabled speed. I mean, sure, a game like TF2 has roles with different speed values that COULD potentially affect it but none in such a way that Dust does what with kinetic catalyzers and armor plates affecting it, or even different frames with those modules.
Butnow that I think about it.... I need to run one more test to finalize this theory.... Give me a bit guys. |
Aeon Amadi
A.N.O.N.Y.M.O.U.S. League of Infamy
2218
|
Posted - 2013.07.24 19:36:00 -
[16] - Quote
Based on the evidence shown in these two videos (sorry for being a **** poor shot with a tactical - not much experience with it) its safe to assume that increased speed DOES impact hit box lag quite significantly.
Test conducted using similar experiment modifiers as before, with nanohives forming a track. We increased the spacing between them from 5m to 10m to have more of a variable toward angle of fire, with subjects being easier to hit as they ran 'away' with their back facing the shooter. The theory for this being that being as the hit box is behind them, its much easier to hit without the collision box being in the way, but more onthat in a moment.
For this test, being almost entirely about speed, the heavy was loaded down with complex armor plates and the scout used enhanced kinetic catalyzers.
This did spawn a new theory however, one that will take significantly more time to research, in that a projectile striking a targets collision box will cancel out a round. This is largely due to the 'shield flicker' from hitting a target directly but receiving no hit indicator, an issue many Snipers faced pre-chromosome. What we have noticed however is that it will actually prevent the round from continuing on to the hitbox. This is also why targets in CQC may seem like they are being hit (shield flicker/sparks flying off of armor) but not receiving damage.
One thing is for certain though and that is that it is easier to gauge the hitbox placement on a slower target than it is a faster one, needing multiple tests to estimate the placement on the scout.
Heavy Hit Box Lag Test - http://m.youtube.com/?reload=9&rdm=tmj34105#/home
Scout Hit Box Lag Test - http://m.youtube.com/?reload=9&rdm=tmj34105#/home |
Aeon Amadi
A.N.O.N.Y.M.O.U.S. League of Infamy
2221
|
Posted - 2013.07.25 00:08:00 -
[17] - Quote
Vaux Karn wrote:As a sniper I have learne one thing about these so called "magic curving bullets," they are not magic or curving, it is a result of network latency (aka lag). If someone is sprinting, shots lined up perfectly on them will always miss, you have to shoot behind them to hit them. And that applies to all suits, the faster they move the further behind them you have to shoot. Yes this is a problem, but it is not an enlarged hit box, just a displaced hit box. This is not exclusive to snipers either, it applies to all weapons. I am not saying it doesn't need to be fixed, it most definitely does, but you should at least know what the issue is before just assuming. Again, the problem is hitbox displacement due to network latency.
Yeah, pretty muchjust repeated everything that was already said here, no offense |
Aeon Amadi
A.N.O.N.Y.M.O.U.S. League of Infamy
2324
|
Posted - 2013.08.02 08:44:00 -
[18] - Quote
CCP Wolfman wrote:Aeon Amadi wrote:Edit: Posted in General Discussion as its not necessary a Bug. Yet. Tests were conducted on a domination match on North American servers (both testers are resident) on Skim Junction. Precursor: This is an issue chalked up and reported here: https://forums.dust514.com/default.aspx?g=posts&t=92132&find=unreadThe purpose of this thread is to lay down the ground work for a testing base contributing to this phenomenon. As such, we've ruled out two direct concerns previously brought up. A.) That this is a map error in Spine Cresent. Common belief is that this phenomenon only occurs in a small gap at objective Alpha on the bridge. Videos show this is not the case. B.) That this is associated with Lag. Being as we were on North American servers, this should not have been the case and even in a high latency environment testing would inevitably come out the same due to testing procedures. What we concluded is that while the hit box may not grow larger - something CCP states there are not mechanics that could possibly occur in this manner - something IS happening to prevent the Scouts adequate access through previously passable locations when they sprint. More so than a Heavy, as we'll get to in a later video. Test A - Scout with base movement speed http://m.youtube.com/index?client=mv-google&rdm=ticq73hf#/watch?v=DjRe7iWl22A&feature=plcpTest B - Scout with Two Enhanced Kin Cats http://m.youtube.com/index?client=mv-google&rdm=ticq73hf#/watch?v=EQ5kOiqEYG4&feature=plcpTest C - Scout with Two Militia Armor Plates http://m.youtube.com/index?client=mv-google&rdm=ticq73hf#/watch?v=IfUOdBn6Qt8&feature=plcpTest D - Heavy with One Enhanced Kin Cat Warning: Includes crude speech from recorder due to disbelief as the Heavy's elbows barely clipped through the bumpers. http://m.youtube.com/index?client=mv-google&rdm=ticq73hf#/channel/UC1DQd-dmONpNZJfO8fnnJGA?feature=guideIn Conclusion While this does not necessarily prove that a Sprinting Scout is easier to hit due to hitbox increase, it does invariably prove that ALL suits are less agile when sprinting as the collidable objects are very much more noticable. Turning corners, trying to squeeze through tight spaces and running past players/objects is difficult to manage when sprinting because of this oddity, whatever the case may be. So does sprinting decrease a Scouts survivability? Thats still up for debate. If you guys have any other thoughts on how to more accurately gauge this, feel free to suggest something to us and we'll do what we can. Links not working...
Interesting. Give me a second, I'll see what I can do. They were uploaded via my phone as it's the only way I could record the game. |
Aeon Amadi
A.N.O.N.Y.M.O.U.S. League of Infamy
2324
|
Posted - 2013.08.02 08:48:00 -
[19] - Quote
All provided links have been replaced. The links used previously were for the mobile version of youtube - they now go to the videos as they are for normal Desktops. |
Aeon Amadi
A.N.O.N.Y.M.O.U.S. League of Infamy
2325
|
Posted - 2013.08.02 08:52:00 -
[20] - Quote
CCP Frame wrote:CCP Wolfman wrote:Links not working...
Same problem here. Video links don't work and I just get YouTube home page instead. UPDATE: Now it works. Thank you for edits!
Hey, thanks for reporting the issue guys - wouldn't have fixed it if Wolfman hadn't said anything |
|
Aeon Amadi
A.N.O.N.Y.M.O.U.S. League of Infamy
2327
|
Posted - 2013.08.02 10:53:00 -
[21] - Quote
CommanderBolt wrote:CCP Wolfman wrote:Ok, this is caused by the collision box changing shape when the player sprints to better fit the mesh. We'll change this so you don't catch on stuff.
It isn't related to any issues with hit detection. It looks like the hit detection issue may be latency related, investigation continues. Thanks for looking into this. I have also noticed issues with sprinting and getting stuck on the 'crows nest'. On the map - Ashland, The rooftop area accessible via ladder in-between objectives B and C, closer to B. That area overlooking the CRU on one side and the B objective on the other. It is as if we get stuck in quick sand or something. I have only had it happen a couple of times but its always in that area and with a scout suit that is sprinting.
There's a couple of areas (Harlequin 13 knows more as to their specific locations) where the player will have a similar effect but I'm almost positive it's from the terrain being too steep in a small area. I'll get with him later tonight and see if he can spot them out so we can submit them. |
Aeon Amadi
A.N.O.N.Y.M.O.U.S. League of Infamy
2327
|
Posted - 2013.08.02 10:56:00 -
[22] - Quote
CCP Wolfman wrote:Ok, this is caused by the collision box changing shape when the player sprints to better fit the mesh. We'll change this so you don't catch on stuff.
It isn't related to any issues with hit detection. It looks like the hit detection issue may be latency related, investigation continues.
Awesome, this will definitely help out with evasion of objects when sprinting at the very least.
My main concern with Hit Detection (particularly with scouts) is that the hit box -does- lag farther behind with increased speed. The faster you go, the farther it lags behind. Both of the videos I provided in post #61 show this in the same location in the same match. So unless there was a latency spike in between the Heavy Hit Box Lag test and the Scout Hit Box Lag test, I'd be more inclined to rule latency out.
That is, network latency. Performance is completely different matter. |
Aeon Amadi
A.N.O.N.Y.M.O.U.S.
2383
|
Posted - 2013.08.06 16:55:00 -
[23] - Quote
THE TRAINSPOTTER wrote:CCP this is what you missing , dedicate and pasionate people
Awwwwhhh, not gonna lie this made me smile
I've actually been working -VERY- hard on going to college for Game Design so I can get my bachelor's degree to hopefully work at CCP one day. It's been my ultimate goal for the past five years and I'm very excited to say that I'm finally making progress toward that end.
I did notice they are hiring QA for Shanghai but I don't have any previous job experience...
Unless you count numerous alpha/beta tests for a variety of games, mods and even being a developer for a text based game.
I dunno - might throw an application in for the ***** and giggles, see what happens xD |
Aeon Amadi
A.N.O.N.Y.M.O.U.S.
2392
|
Posted - 2013.08.07 09:56:00 -
[24] - Quote
hgghyujh wrote:I wonder if this is the cuase of the sticky walls in this game??? I'm so tired of being grab onto by walls.
It is. |
Aeon Amadi
A.N.O.N.Y.M.O.U.S.
2392
|
Posted - 2013.08.07 09:58:00 -
[25] - Quote
Iron Wolf Saber wrote:CCP has been known for hiring from the community. I mean look at CCP Logibro and CCP Mintchip lately.
Right but I'm sure they had some kind of qualifications like... I dunno. A high school diploma...
Of which I do not |
|
|
|