46 replies to this topic

[Zoid]

I've noticed this a bit when running a double UAC 10 Warhawk (with jam reduction) and chalked it up to coincidence, but I definitely think there's a bug now. Every time one UAC jammed, both of them did, so I went into the testing grounds.

Results after a little testing: 8 out of 10 times when one jammed, BOTH did. I shouldn't have to tell you how astronomically low a probability that is. These should be independent events, so both of them jamming at the same time is a 1% probability (since the WHK has jam chance reduction). Having it happen almost every single time has got to be a bug.

EDIT:
I ran some additional tests and I believe it's an issue with the jam chance reduction quirk. I spent 600 rounds on a Dire Wolf with 2 UAC10s and got ZERO double jams, which is actually lower than expected. I then did a CTF-3D with the 30% jam reduction (for UAC 5s) and got 6 out of 12 times they jammed, they both jammed. I then loaded up a WHK again with the TC removed and 720 rounds, which after 63 double taps had FIVE double jams and ZERO single jams. I reran this test and got 2 single jams and 6 double jams. So to recap the data:

Dire Wolf (no jam reduction quirk): 600 rounds spent, ZERO double jams.
CTF-3D (30% jam reduction quirk): 300 rounds spent, 6 out of 12 times they jammed, both did.
WHK (30% jam reduction quirk): 1440 rounds spent, 2 single jams, 11 double jams.



I think this data shows it's pretty obvious that the jam reduction quirk is making both jam if one does, as the WHK was roughly 120 double taps. That follows just fine with a 10% chance.

There is an Enforcer and a Jagermech variant that can both mount more than 2 (and have the jam reduction quirk), I'd bet this would be extremely apparent on them. Unfortunately I do not own these 'mechs.

Edited by Zoid, 20 April 2016 - 06:44 AM.

[Champion of Khorne Lord of Blood]

Zoid, on 19 April 2016 - 06:56 PM, said:

I've noticed this a bit when running a double UAC 10 Warhawk (with jam reduction) and chalked it up to coincidence, but I definitely think there's a bug now. Every time one UAC jammed, both of them did, so I went into the testing grounds.

Results after a little testing: 8 out of 10 times when one jammed, BOTH did. I shouldn't have to tell you how astronomically low a probability that is. These should be independent events, so both of them jamming at the same time is a 1% probability (since the WHK has jam chance reduction). Having it happen almost every single time has got to be a bug.

So did you do any testing with firing them separately 10 times each in a back and fourth?

I do notice UACs do tend to usually have 2 jam up at the same time, my triple UAC5 Ebon Jaguar always seems to have its arm guns jam up at the same time... usually after about 2 shots right when the enemies are pushing.

Edited by Dakota1000, 19 April 2016 - 07:01 PM.

[Spheroid]

more data required.

[Wintersdark]

Dakota1000, on 19 April 2016 - 06:59 PM, said:

So did you do any testing with firing them separately 10 times each in a back and fourth?

I do notice UACs do tend to usually have 2 jam up at the same time, my triple UAC5 Ebon Jaguar always seems to have its arm guns jam up at the same time... usually after about 2 shots right when the enemies are pushing.

I'm going to go test this right now, however...

There is an existing bug that may be related. Using multiple Gauss Rifles (3+) sometimes, when you lose one, it'll still try to charge that one and an existing one, rather than the two surviving rifles. I understand this is a different mechanic, but... I'm suspicious.


Anyways, UAC10's on a Warhawk. TO THE TESTING GROUNDS! FOR SCIENCE!

Edited by Wintersdark, 19 April 2016 - 07:04 PM.

[Champion of Khorne Lord of Blood]

Wintersdark, on 19 April 2016 - 07:04 PM, said:

I'm going to go test this right now, however...

There is an existing bug that may be related. Using multiple Gauss Rifles (3+) sometimes, when you lose one, it'll still try to charge that one and an existing one, rather than the two surviving rifles. I understand this is a different mechanic, but... I'm suspicious.


Anyways, UAC10's on a Warhawk. TO THE TESTING GROUNDS! FOR SCIENCE!

With the way the testing grounds and live servers differ I got to wonder about the results we get from them sometimes.

[Wintersdark]

Oops.

Right. Update: With 6 tons of ammo, I was only able to get my UAC10's to jam twice.

Neither time jammed both. Testing again.

Dakota1000, on 19 April 2016 - 07:13 PM, said:

With the way the testing grounds and live servers differ I got to wonder about the results we get from them sometimes.

I did the testing grounds once, and it was so hard to get a jam I just went live.

Whole match, 3 kills, only got a jam after I lost a UAC

Clearly, the jam quirk works.

Edited by Wintersdark, 19 April 2016 - 07:23 PM.

[Champion of Khorne Lord of Blood]

Wintersdark, on 19 April 2016 - 07:24 PM, said:

Oops.

Right. Update: With 6 tons of ammo, I was only able to get my UAC10's to jam twice.

Neither time jammed both. Testing again.


I did the testing grounds once, and it was so hard to get a jam I just went live.

Whole match, 3 kills, only got a jam after I lost a UAC

Clearly, the jam quirk works.

Did you try removing your Blessing of RNGesus ring, it grants +5 luck?

[Wintersdark]

Three more matches.

When firing with both UAC10's in the same weapon group, they jammed together every single time. Going to start taking video and submit a formal bug report.

It does NOT happen when firing only one UAC10, or when firing both cannons at the same time on separate weapon groups.

[xXBagheeraXx]

Ive actually noticed this piloting my 3 and 4 uac5 maulers. If I chain fire them THEY CAN FIRE FOREVER...I can easly dump 10+ tons of ammo overboard in a matter of minutes...but If i group fire them Jam city. SO my tactic now is to chain fire for a few shots then start group firing while the cannons are staggered...then its usualy only 1 or 2 guns jamming at a time, rather than having all 3 or 4 jam at once.

[Y E O N N E]

I maintain my suspicion that UAC jam chance is not a purely random event depending only on a uniform probability. I am all but certain there is some code in there that causes them to jam more frequently under certain conditions, some of those conditions being double tapped at the wrong point during the cool-down or having multiple UACs in operation.

[Wintersdark]

Yeonne Greene, on 19 April 2016 - 08:10 PM, said:

I maintain my suspicion that UAC jam chance is not a purely random event depending only on a uniform probability. I am all but certain there is some code in there that causes them to jam more frequently under certain conditions, some of those conditions being double tapped at the wrong point during the cool-down or having multiple UACs in operation.

That's extremely doubtful. It's like the whole Onyxia's Breath thing. People like to see patterns when there aren't any.

There is definitely an issue with group fired UAC's jamming, but that would definitely be a bug, and not deliberate. PGI just doesn't do those sorts of things - code wierd edge cases deliberately to mess with people.

I've done a LOT of UAC testing for various things, and I've never found any correlation to when in the cycle you doubletap (that I've deliberately tested) or other such things.

[xXBagheeraXx]

Serously hop in some multiple uac armed IS mechs like the king crab or mauler...ITs fairly obvious then...I had a game a few weeks ago where the first dam double tap in my mauler 1p with 4 Uac5's jammed all 4 guns....ALL FOUR....I can see 1 or two but all 4 at once? cmon man...

[Y E O N N E]

Wintersdark, on 19 April 2016 - 08:15 PM, said:

That's extremely doubtful. It's like the whole Onyxia's Breath thing. People like to see patterns when there aren't any.

There is definitely an issue with group fired UAC's jamming, but that would definitely be a bug, and not deliberate. PGI just doesn't do those sorts of things - code wierd edge cases deliberately to mess with people.

I've done a LOT of UAC testing for various things, and I've never found any correlation to when in the cycle you doubletap (that I've deliberately tested) or other such things.

I never implied that what is going with UACs on is deliberate, only that it's a thing.

But, they always seem to jam up when I depress the fire button just after the mid-point. However, if I watch the fire group boxes next to my reticle, they indicate clear before the cool-down is complete, and if I pull the trigger at that point they will almost never jam.

[Champion of Khorne Lord of Blood]

So for now if we are UAC boating we should keep all our UACs in separate groups, to avoid the issue while it gets fixed. That might make my UAC5 Jaguar seem a bit less disappointing.

[Chuck Jager]

Yeonne Greene, on 19 April 2016 - 08:10 PM, said:

I maintain my suspicion that UAC jam chance is not a purely random event depending only on a uniform probability. I am all but certain there is some code in there that causes them to jam more frequently under certain conditions, some of those conditions being double tapped at the wrong point during the cool-down or having multiple UACs in operation.

It used to "exploit ably" be this way, and folks figured out the timing to never get jams. I think pgi has the code set in a way to create less processing clutter during game time. Recently I have set my macros to be at the point that the second shot can go off and I get far less jams than by finger tapping. A good starting point is about half of cooldown for uac 5s or right when the stream ends for clan uacs.

Finger tapping seems to cause the most jams for whatever reason. Maybe the macro imitation key press bypasses the "sniffer" to register the jam chance. I do like getting that second shot in asap.

This could all be placebo effect, but it makes me feel better. I do know that the testing grounds are not the same as in game.

[Jackal Noble]

This issue isn't new, even made a thread about it as well lol, but it's hella good to spread more awareness for sure. Annoyed me enough on my Uac5 ebon jag that I went to 10s for a while. Funny enough it actually seems that targeting computers compound the issue.

[Zoid]

Wintersdark, on 19 April 2016 - 08:15 PM, said:

That's extremely doubtful. It's like the whole Onyxia's Breath thing. People like to see patterns when there aren't any.

There is definitely an issue with group fired UAC's jamming, but that would definitely be a bug, and not deliberate. PGI just doesn't do those sorts of things - code wierd edge cases deliberately to mess with people.

I've done a LOT of UAC testing for various things, and I've never found any correlation to when in the cycle you doubletap (that I've deliberately tested) or other such things.

Except this is something with a definite probability you can calculate. For my case, 8 out of 10 times they jammed, they jammed together. 6 times in a row at first. Since with the jam quirk they have a 10% chance of jamming and they should be independent events, the chance of this happening is 0.1^6, or 1 in a million.

I'll look up the actual distribution equation later for something happening N times given M chances if needed, but this really should suffice. If any of my code was giving these results on such a simple test I'd check it immediately.

[Wintersdark]

Zoid, on 20 April 2016 - 04:51 AM, said:

Except this is something with a definite probability you can calculate. For my case, 8 out of 10 times they jammed, they jammed together. 6 times in a row at first. Since with the jam quirk they have a 10% chance of jamming and they should be independent events, the chance of this happening is 0.1^6, or 1 in a million.

I'll look up the actual distribution equation later for something happening N times given M chances if needed, but this really should suffice. If any of my code was giving these results on such a simple test I'd check it immediately.

In my post you quoted:

Quote

There is definitely an issue with group fired UAC's jamming, but that would definitely be a bug, and not deliberate. PGI just doesn't do those sorts of things - code wierd edge cases deliberately to mess with people.

I'm just saying it's not designed to do that, it's a bug, it's testable and provable. Repeatedly getting simultaneous jams on group fired uac's on a reduced jam mech is EXTREMELY unlikely, but yet it's very common (it's the norm in fact) in livefire tests.

I only get that result in live though, not the testing ground.

My post to Ms. Greene, however, was about the rest (timing impacting jamming and other factors) - I've tested those and couldn't find a statistical correlation between timing and jam frequency. It's well within a reasonable margin of error.

But the group fired uac's thing is way beyond that.

[Zoid]

Wintersdark, on 20 April 2016 - 05:50 AM, said:

In my post you quoted:



I'm just saying it's not designed to do that, it's a bug, it's testable and provable. Repeatedly getting simultaneous jams on group fired uac's on a reduced jam mech is EXTREMELY unlikely, but yet it's very common (it's the norm in fact) in livefire tests.

I only get that result in live though, not the testing ground.

My post to Ms. Greene, however, was about the rest (timing impacting jamming and other factors) - I've tested those and couldn't find a statistical correlation between timing and jam frequency. It's well within a reasonable margin of error.

But the group fired uac's thing is way beyond that.

I misread your post, my bad. Anyway, see edit to the original post. I believe the issue lies with the jam chance reduction quirk.

[Wintersdark]

Zoid, on 20 April 2016 - 06:46 AM, said:

I misread your post, my bad. Anyway, see edit to the original post. I believe the issue lies with the jam chance reduction quirk.

easy enough to test. Will bust out a hunchback after work and give it a go.