46 replies to this topic

[GrimRiver]

I made a post about this sometime ago

It was about how any mech with U-AC jam chance or cooldown actually jams U-AC's
more then mechs without any cooldown or jam chance quirks.

Even more so evident with 2 or more U-AC's fired at the same time.

Cooldowns or jam-chance conflicts with U-AC's RNG jam chance because RNG is still RNG no matter what kind of less % you try to slap on it, all it's doing is adding more number to crunch.

You can't make something less random when it was random in the first place.

[Dino Might]

GrimRiver, on 21 April 2016 - 07:17 PM, said:

I made a post about this sometime ago

It was about how any mech with U-AC jam chance or cooldown actually jams U-AC's
more then mechs without any cooldown or jam chance quirks.

Even more so evident with 2 or more U-AC's fired at the same time.

Cooldowns or jam-chance conflicts with U-AC's RNG jam chance because RNG is still RNG no matter what kind of less % you try to slap on it, all it's doing is adding more number to crunch.

You can't make something less random when it was random in the first place.

I'm not understanding this.

It's fairly simple to reduce the likelihood of a RNG hitting a particular threshhold for your event to happen. I am not sure what you mean by "less random," but you can certainly adjust either standard deviation or mean to affect the outcomes of your RNG. The question is, how does PGI calculate whether a jam occurs or not with the RNG, and how do the quirks affect that RNG?

ETA: example of how you could create a RNG for jamming.

Create uniform random number between 1 and 10 (let's assume whole numbers only for ease of calculation). If the result is 9 or higher, you get a jam. So, you have a 20% jam chance (because the distribution is uniform). Now, generate the random number but always subtract one from the result. Now you only jam on a 10, so you have a 10% jam chance.

If you now open this up to all real numbers in a certain range, you can do a lot with addition or multiplication to affect the probability of a jam from a simple uniform distribution.

Edited by Dino Might, 21 April 2016 - 07:32 PM.

[Kurieg]

Dakota1000, on 20 April 2016 - 09:28 PM, said:

I've been using UAC5s in all cases that I've been mentioning.

I noticed the same thing last night when I put together a dual UAC-5 EBJ-prime. Virtually all the jam events were double jams. Yes, I had them on the same button press (left mouse).

I have not noticed it on the JM6-DD with 3x UAC-5's or a MLR-MX90 with 4x UAC's.

I did run a dual UAC-10 WHK at one point, but I think I mostly had that on 2 different mouse buttons. I have the EBJ's on one mouse button because I have 3 diff. weapon types on it and am too old to run 4 buttons effectively yet.

[Zoid]

It does seem to be certain chassis that have the issue. I saw the same thing on my Ilya Muromets but not on my Atlas.

For those of you who think this is just seeing patterns where there are none, the probabilities are not just low, they are astronomically low for the tests I've run. The Warhawk had a string of eleven double jams. If these were independent events, then when one jams, the other has a 10% chance too. To see eleven double jams in a row is thus 0.1^11=1 in a trillion probability.

To put this in perspective, if 100,000 MechWarriors were double tapping UACs at a rate of once per 5s (which is pretty much max) all day every day, it would take almost 2 years to see this happen, yet I have been able to get it to happen with just a quick solo test. There is definitely a bug somewhere.

[Zoid]

Shamelessly bumping this due to a FW game where it happened 3 times in the space of 30 seconds. It was already a bad game but that just added insult to injury. There is no way this is normal.

What is the proper method for submitting a bug report for this anyway?

[Nicu]

I am having the same problem with the timber wolf using two UAC 10 in the same group. I just tested this in the testing grounds and every time there was a weapon jam, both UAC 10 would be jammed.

[RestosIII]

That is one fine necro you've got there.