39 replies to this topic

[Phra]

1Sascha, on 21 October 2013 - 03:38 AM, said:

Nope.. even with 3UAC on my Jäger, I'll just stand there with just my... errr... I'm left weaponless due to excessive jamming. Used to run that build before the re-nerf and it was pretty effective. But these days, I'd rather run 4x AC5 on it.

I dunno, haven't played at 15% jam rate, but every time I try to swap to ACs on my Ilya, I feel like I'm missing a huge chunk of my dmg, but than again, I have 3 MLs to fall back upon for the duration of the unjam.

[JigglyMoobs]

It's funny. I have the same problem with UAC5s on my Jager, but not on my Atlas. Maybe the UAC5 + Jager combination is somehow broken?

[Bhael Fire]

I think all they need to do is make the UAC/5 only double fire when the button is double clicked. They can keep the random chance of jamming and it won't be as annoying since the player can actively chose to take a chance on it jamming to get an extra shot off during cooldown.

Right now, since the double-fire mode is always on, it can jam on the first shot way too easily, which leads to frustration.

In other words, it should function exactly like an AC/5 until the button is double clicked.

[D04S02B04]

FinsT, on 21 October 2013 - 06:15 AM, said:

Nope, it is not. I made in-game testing of it few weeks ago. I believe your numbers are wrong: 1st, in-game testing clearly shows the jam chance is not 20%; 2nd, because of that and also because DPS - i believe, - should be calculated as "total damage" / "total time" - and not the way you do it, - i believe that actual average DPS of UAC/5 is 4.016 or so (and not 5.4583 as you claim).

Wanna know more? I've put my results into a detailed post here. Enjoy.

I done up the Math as well and I see where I gone wrong with my assumption in doing up the probability table. I can't see ur post in the forum btw.

What was the base that you used as a function of Total Shots / Total Time? I used 19.5 seconds as it was a common multiple of 1.5sec and 6.5 sec.

[Huntsman]

If you're brawling, you want the AC5. If you're fighting via direct fire support and you're in and out of cover, take the UAC.

The risk no longer is worth the potential reward with the UAC in a state of continuous fire using the double shot mode. In a short burst though, the UAC will on average deal out better damage than any other weapon, and if you're utilizing cover based fighting, you're out of the line of fire while the UAC is unjamming.

My 2 cents...where you would use a gauss, use the UAC instead. The UAC's range is almost as good, and its DPS is much better than the gauss while in DS mode....and again, when you're not in DS mode, you're not exposing yourself to fire. Best of both worlds.

Edited by Huntsman, 21 October 2013 - 10:03 PM.

[FinsT]

D04S02B04, on 21 October 2013 - 09:40 PM, said:

I done up the Math as well and I see where I gone wrong with my assumption in doing up the probability table. I can't see ur post in the forum btw.

What was the base that you used as a function of Total Shots / Total Time? I used 19.5 seconds as it was a common multiple of 1.5sec and 6.5 sec.

6.5? What 6.5? When it jams, it just sits jammed for 5.0 seconds and then it's ready to fire again. That was my impression with the weapon...

And, well, if you can't see my post through the link... Well... I guess i'll drop the thing here. It's a little bit big, though. /gulps

Anyways. %)

________________________________________________________________________

AC/5 or UAC/5: the answer

...

Details and math.

1. Stats.

Most stats are the same:
- damage per shot (5),
- heat (1),
- projectile speed (1300),
- impulse (0.030),
- range (almost: AC/5 has range = 600, max range = 1700; UAC/5 has range = 620, max range = 1800),
- amount of ammo per ton (30) ( - since last patch).
Other stats are a bit different:
- weight (AC/5 = 8 tons, UAC/5 = 9 tons),
- slots (AC/5 = 4 slots, UAC/5 = 5 slots).
And of course, the main difference:
- AC/5 has plain 1.5s cooldown, while UAC/5 also has 1.5s cooldown, but: "can be shoot during cooldown but has 20% chance to jam. If jamed it needs 5.0s to unjam". That's what Smurfy's data says. Testing reveals it's not entirely correct (to say the least). See "Tests done" section at the end of this post.

2. DPS over (nearly) infinite time (1000 shots).
AC/5 = 3.33
UAC/5 = 4.0
- The latter is calculated by me using facts i found during in-game testing of the weapon, and assuming "non-jam cooldown" per shot being 0.75s (for simplicity; shouldn't affect the outcome even while in reality it's "pairs" of shots, because ~half of shots on which jam happens will be 1st shot of the pair, and the other half - 2nd shot of the pair). Math is: 5000 (damage, total) / ( 11.65% * 1000(shots fired) * 5s + 88.35% * 1000(short fired) * 0.75s ) = 4.016.

3. DPS over 2.25 seconds time frame
AC/5 = 4.44
UAC/5 = 6.8 average (actually 2.2 ... 8.9 - varies; the former is for when it jams after 1st shot, the latter is for when it doesn't jam for 3 shots in a row (1st ... 3rd shots).
- this kind of DPS is important for punching any standing mech, assuming it will go into very nearby cover once under heavy fire, and/or we need to. 2.25 seconds is close to average time one can keep firing at a target in such situations, and is "fair" average considering AC/5 fixed cooldown.
Note. It is useful to figure out chances here. Based on my tests and with a bit of probabilities' theory,
- chance UAC/5 will do exactly 2.2 DPS ( = single shot which jams the weapon) over 2.25 seconds: ~11.65%;
- chance UAC/5 will do exactly 4.4 DPS ( = 2 shots) over 2.25 seconds: ~10.29%;
- chance UAC/5 will do exactly 6.7 DPS ( = 3 shots) over 2.25 seconds: ~9.09%
- chance UAS/5 will do exactly 8.9 DPS ( = 4 shots, maximum possible during this time frame) = 68.97% (this is total for when 4th shot jams, and when it doesn't)
This (roughly) approximates to average DPS of 6.8, as shown above.

4. DPS over 8 seconds time frame
AC/5 = 3.75
UAC/5 = average 4.6 (actually 1.25 ... 7.5 - varies; the former is double jams on 1st shots, probability of which is ~1%; the latter is no jams whatsoever for 11 shots in a row, probability of which is 25.6%) - this kind of DPS is improtant for an average-length brawl. Longer than 8 seconds encounters will have DPS figures of both weapons being increasingly similar, as it becomes increasingly unlikely for UAC/5 to either jam all the time, or not to jam all the time.
- it is challenging to calculate average UAC/5 DPS here, considering all possible jam cases. Single jam early may end before 8 seconds are over, and more shots then may be fired. Single jam being "late" during 8s does not lower DPS over 8s as much as a jam "in the middle" of it. So, as my limited math knowledge allows, the best guess i can do is this: average must be not too far from to 4.6 DPS (which i put above). If i am mistaken with it, then hopefully not more than by 0.2DPS.

5. Supressing fire reliability (percentage of time a weapon is mauling the target, 1000 shots fired)
AC/5 = 100%
UAC/5 = 72.2%
- the latter is calculated by me using following assumtion: the target is considered "supressed" whenever it is hit by the weapon, and whenever it was hit by the weapon less than 1.5 seconds ago. Math is: (total time supressed) / (total time needed to fire 1000 rounds, see p.2) = ( 88.35% * 1000 * 0.75s + 11.65% * 1000 * 1.5s ) / 1160s = (662.6s + 174.75s) / 1160s = 72.2%.

That's pretty much it. Numbers here define conclusions both above and below.

AC/5 is clearly a winner in supressing fire ability (especially when paired with a 2nd AC/5 and fired one after each other), hammening the enemy often enough to shake him often enough and strong enough to make proper aiming nearly impossible - i heard that the shake is proportional to impulse * damage, and AC/5 has nearly same impulse as AC/2, but 2.5 times higher damage, so it kicks rather well. Also, 72.2% supression ability of UAC/5 is calculated for a 1000 shots; no enemy can stand for that long, of course. Important thing to realize here is that UAC/5 spends almost half of the total time being jammed (roughly (1160 - 663) / 1160 = 42.8% of the time), and during this time, enemy won't be dropped (if the only weapons used are UAC/5s), thus prolonging the firefight; and then if there is a lucky long sequence of no-jamming, - good chances are not all of it will be needed to drop the enemy down. Thus i'd say that real supressing ability of UAC/5 is lower than 72.2% - possibly something close to 60%.

In terms of DPS, UAC/5 is a winner in all sorts of DPS-needed situations on average, by some 20%...50%. However, due to random nature of jamming and long un-jamming time (5.0s), and due to its limited to double-that-of-AC/5 firepower, in case of real bad luck (jams on 1st shot) UAC/5 can do as little as ~40% of AC/5 DPS. Together, these facts define when UAC/5 is useful, and when it is not. Some certain colncusions which can be done out of this - are in the beginning of this post, in "short version" section, and basically are answers to when one clearly needs to use AC/5, and when UAC/5.

end of details and math
____________________________________________________________________________
===================================================================

++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++
Tests done.

Did some UAC/5 testing today (06 october 2013). The testing clearly establishes the following:
- its jam probability is present always - i.e. even on shots made more than 1,5 seconds after a previous shot;
- its jam chance per shot is significantly lower than Smurfy says it is;
- it can't fire more than twice during any given 1.5s period of time, apparently, limiting maximum possible DPS to double of AC/5's DPS.

// This completely invalidates initial conclusions and demands complete rework of this post, which i did. Old version is not saved, because it was utter {Scrap}.

1. clicking like mad, trying to make UAC/5 to fire any faster than when one simply holds its trigger down.
- Result: failure. It can't be fired more often than 2 shots per 1.5s. Period. ><

2. 500 rounds of UAC/5 fired with the fire trigger held down.
- Result: total jams - 58, of which a couple happened on very 1st shot after previous jam. This translates into 58 / 500 = 11.6%. From probabilities' theory, we know that series as large as 500 tries allow to find out the probability of an event quite precisely - plus-minus 1% or less. During this test, i had 2 UAC/5 cannons equipped, and was firing both (to speed up the procss - jams are easy to count).

3. 300 rounds of UAC/5 fired with the fire trigger held down.
- Result: total jams - 35, of which 7 happened on very 1st short after previous jam (which is ~3 more than expected, but this can be "bad luck", since during this sequence, there were 35 "first shots" obviously, and that's statistically low sample to make any conclusion). This test was done having only 1 UAC/5 equipped and firing, to find out whether jam chance depends on number of UAC/5s equipped/firing, or not. Well, it does not: 35 / 300 = ~11.7%. Very good confirmation of jamming chance, too.

Rounding that to ingeger (as well as previous test's result), UAC/5 has 12% jam chance - not 20%. Why then Smurfy is saying 20%, one might wonder. And i have a guess. Perhaps, Smurfy's 20% - is a chance that UAC/5 will jam on either 1st or 2nd shot of its 1.5s cycle. I.e., 20% chance to jam per PAIR of shots (or in other words, chance to jam during its 1.5s "normal" cooldown period) - and the jam can happen on either 1st or 2nd shot, with that. The latter would explain why jam chance is slightly higher than (otherwise expected) 10% per shot. I don't know if this guess is correct or not, but ~11.65% +-0.4% chance to jam per shot is the thing i am pretty sure about - 800 shots fired in total, counting jams, is statistically large enough sample to make such a conclusion. Also, during tests, i got a general "feeling" that the weapon is jamming on 1st shot of a pair nearly half the time, and on 2nd shot of the pair also nearly half the time - i.e. there is no difference on jam chance, apparently, between 1st and 2nd shots of a pair.

4. A long (40 UAC/5 shots) sequence without a single jam - amazingly, i happened in very 1st try, - alphaing UAC/5 paired with AC/5, alpha key held down for the whole duration.
- Result: for the whole duration, both weapons fired synchronously: for each AC/5 shot, there was a pair of UAC/5 shots. This allows to say that UAC/5 maximum DPS ("no jam mode") - is precisely double of AC/5's DPS, not more, not less.
++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++

Fins

[1Sascha]

Bhael Fire, on 21 October 2013 - 05:07 PM, said:

I think all they need to do is make the UAC/5 only double fire when the button is double clicked. They can keep the random chance of jamming and it won't be as annoying since the player can actively chose to take a chance on it jamming to get an extra shot off during cooldown.

Right now, since the double-fire mode is always on, it can jam on the first shot way too easily, which leads to frustration.

In other words, it should function exactly like an AC/5 until the button is double clicked.

Yup.. suggested that one myself in an older thread IIRC.

There seems to be a workaround to achieve this behavior, mind you. I tried it on my Thunderbolt and it worked.

Run 1xAC2 + 1x UAC5.
Put both guns in the same group (group 1, for example), set group to chain.
Put both guns into a different group (group 2), set group to group-fire.

If you fire group 1, the UAC wil fire at the same rate as an AC5 and will never jam. For maximum fire rate (and usual chance of jam) use group 2.


S.

[D04S02B04]

FinsT, on 22 October 2013 - 02:26 AM, said:

... mathematics based on hard testing ...

I accept your concept on total damage / total time. I used a different variation by calculating total shots / total time.

Mechanics
Each time you fire, you have a 1.5 sec cooldown. If it jumps, you spend 5sec unjamming, then it returns to the original state of cooldown it was at, which makes each Jam effectively a 6.5sec CD for 1 shot and 1.5sec for 2 shots when working normally. The 20% jam chance only occurs if you attempt to fire a 2nd shot within the CD period. Therefore my mathematics only counts for doubleshots and assumes u would fire doubleshots for maximum DPS (otherwise u would just pick the AC5...)

Probability Tree Analysis
I did the entire probability tree analysis in Excel given a set of parameters.

First of all establish a common multiple number 19.5.

This is because 1.5 = 0.5x3 and 6.5 = 0.5x13 and the common multiple will be 0.5x13x3 = because both 13 and 3 are prime numbers. The selection of the number will give a range from 0 jams to 3 jams precisely.

Example of Probability Table

UAC5 Probability Table by D04S02B04, on Flickr

I computed all the probabilities and spaced out all the shots with the restriction of time. All probabilities sum up to 1.

Considerations on Calculation
One of the primary issues with the probability table is saying when is a reasonable time to cut off the recording. There are scenarios that at 19.0 secs since the start of firing, it is still possible to fire again and if you consider it as part of firing at 19.5sec it "cheats" the statistics by appearing to fire more shots than it does.

Theoretical Conclusion
Average total Shots Fired (incl. of Jam) : 15.01 over 19.5sec
Average DPS = Avg Shots Fired / Time Interval x 5Dmg = 0.7699 Shots per Second or 3.85 DPS

I did a separate observation of test fire 10 secs and the most common scenario was firing 7 shots with 1 jam (6 successful double shot, 1 jam) which correlates with the figure given.

However statistically speaking, in short firing encounters the UAC5 will typically overperform and underperform because you're attempting less than 100 tries (theoretical 180 total shots) it will be an outlier. I think the correct amount is somewhere below 80 attempts/tries but I did not study statistics so I do not have the precise figures or the theory governing this.

Over the course of the entire match, you will see that the results will be consistent if you stay alive to fire at least 100 attempts. Note as well that the random number generator in computers are NOT random but just appear to be random therefore resulting in statistics that differ upon testing which may explain your result of 11.65% jam rate.

Recommended Usage
Load 2 UAC5 and a backup weapon. Fire UAC5 close to cover until weapon jams before firing your secondary weapons and thereby run to cover and wait for unjam. This will allow you to maximise the use of "overperforming" statistics.

[MustrumRidcully]

Quote

This will allow you to maximise the use of "overperforming" statistics.

This is the important part, IMO. The trick is unsuring you use your good streaks and minimize the drawbacks of the bad streaks.

[MeiSooHaityu]

I like the OPs idea in theory, but the jamming can be too easily defeated with a Macro. The Jamming needs to be a system that cannot be circumvented easily.

The current system is ehhh, but saddly I don't think the OPs idea would work as long as Macros exist. If it wasn't for macros, I would definently like the idea.

Maybe if the game looked at firing frequency along with how long you hold down the button.

[1Sascha]

MeiSooHaityu, on 23 October 2013 - 08:10 AM, said:

Maybe if the game looked at firing frequency along with how long you hold down the button.

Yup.. which would also be the logical choice of sorts. I know, I know.. logic has no real place in this game when it comes to nerfs, but hey..

I guess I'm repeating myself here, but: The source states that jams occur due to excessive vibrations of the gun. So the gun should only jam (IMO) if it generates enough vibrations. Which it wouldn't after just a single shot.

Don't get me wrong: I don't want the UAC to be the uber-weapon of the day (it'd just get re-nerfed anyway). All I want is for those silly 1 shot -jam - 1 shot -jam 1shot -jam missions to go away.

How about this: So long as you only fire a single shot, the gun would *never* jam, but make it so that you don't get the full rate of fire initially. Meaning that you couldn't abuse this behavior with a macro. Only after shot one would you get the full rate (call it "spin-up time") and jams could only occur after shot two.
So if you'd try to abuse it by only ever firing that first shot in sequence, you'd only get a regular AC5's cyclic rate.

Which of course could be achieved by another suggestion I've made: Dual-stage trigger. Stage one: Never jams but only gives you an AC5's cyclic rate. Stage two: Full rate of fire but with chance of jamming.

Another suggestion I really like is the "remove jams, but add spin-up-time and excessive heat generation on med to long bursts"-thing.


S.

[MustrumRidcully]

MeiSooHaityu, on 23 October 2013 - 08:10 AM, said:

I like the OPs idea in theory, but the jamming can be too easily defeated with a Macro. The Jamming needs to be a system that cannot be circumvented easily.

The current system is ehhh, but saddly I don't think the OPs idea would work as long as Macros exist. If it wasn't for macros, I would definently like the idea.

Maybe if the game looked at firing frequency along with how long you hold down the button.

Circumventing the jamming with a macro now makes your UAC/5 into an AC/5 that weighs 1 ton more and uses 1 crit slot more.
There is no longer a direct advantage from the macro.

I'd still prefer an in-game switch between double and single shot, though.

[MeiSooHaityu]

MustrumRidcully, on 24 October 2013 - 02:04 AM, said:

Circumventing the jamming with a macro now makes your UAC/5 into an AC/5 that weighs 1 ton more and uses 1 crit slot more.
There is no longer a direct advantage from the macro.

I'd still prefer an in-game switch between double and single shot, though.

Maybe I misunderstood, but it sounded as though the OP wanted the doubletap to be less reliable (more likely to jam) the longer the fire button was held down. I figured if a Macro quickly tapped the fire button, that would defeat the mechanic. That was just how I understood it.

[Vodrin Thales]

MeiSooHaityu, on 24 October 2013 - 08:16 AM, said:

Maybe I misunderstood, but it sounded as though the OP wanted the doubletap to be less reliable (more likely to jam) the longer the fire button was held down. I figured if a Macro quickly tapped the fire button, that would defeat the mechanic. That was just how I understood it.

That could be fixed by only allowing fire rate to double, not fire with each press of the trigger.

Overall though I would prefer a solution similar to what MWLL used. Have a heat bar for the gun and have heat increase with every double tap and the gun jams when it overheats. That way the player can control whether or not the gun jams, while allowing increased fire rate that is not completely OP.

[Krubarax]

It would still make the AC5 pointless.
If you can choose to fire the UAC5 "safe"
And then turn up the heat when you need to.

Then it is clearly MUCH better than the AC5.

I still argue that the UAC5 is good as it is right now.
It is a High-risk - High-reward weapon.
Learn to use it. Learn to expect jams. Stay close to cover, pop out, fire, duck back.

As I said; If you do not like it, just usa regular AC5's.

[Howdy Doody]

Spreadsheet Warrior is one of the most fun parts of this game. Great discussion gents!

GB Krubarax, on 25 October 2013 - 02:21 AM, said:

It would still make the AC5 pointless.
If you can choose to fire the UAC5 "safe"
And then turn up the heat when you need to.

Then it is clearly MUCH better than the AC5.

I still argue that the UAC5 is good as it is right now.
It is a High-risk - High-reward weapon.
Learn to use it. Learn to expect jams. Stay close to cover, pop out, fire, duck back.

As I said; If you do not like it, just usa regular AC5's.

This is a really good point that I hadn't really thought of. If the UAC5 worked like the OP has stated (which I think is a GREAT idea) I would never use the AC5. There would need to be more weight/heat added for me to even think about it.

[Voidcrafter]

Howdy Doody, on 25 October 2013 - 03:35 AM, said:

Spreadsheet Warrior is one of the most fun parts of this game. Great discussion gents!



This is a really good point that I hadn't really thought of. If the UAC5 worked like the OP has stated (which I think is a GREAT idea) I would never use the AC5. There would need to be more weight/heat added for me to even think about it.

Yea - at least until that atlai somehow manage to facehugyou and both the UACs just decieded to go on jamming rampage after the first shot(or before it? is it possible still? )
Believe me fella, I've played my back with those weapons before the "adjustment" - then they were high-risk-high-reward weapon, now the high-risk is present, where the "reward" is gimped to almost none.
I still DO think that somehow thier chances of jamming is higher than the described - but you know I can hardly proove it.
5% in the gaming world doesn't mean something will DEFINITELY will happen 5 times of 100 - it only means something, that can happen 20 times in a row or never happen at all.
With that in mind - I can assure you - 20% happens A LOT
I miss their usefulness - even through the PPC hell they were my primary source of damage and I was doing really good...
Now... meeh - it's a death trap :|

[MeiSooHaityu]

Vodrin Thales, on 24 October 2013 - 08:30 AM, said:

That could be fixed by only allowing fire rate to double, not fire with each press of the trigger.

Overall though I would prefer a solution similar to what MWLL used. Have a heat bar for the gun and have heat increase with every double tap and the gun jams when it overheats. That way the player can control whether or not the gun jams, while allowing increased fire rate that is not completely OP.

That sounds like a good idea.

[xMEPHISTOx]

1Sascha, on 21 October 2013 - 02:13 AM, said:

This gun has gone from useless to kinda useful to mostly useless.

Well I would have to disagree. One the most common snipers I see in 12's is the HGN733C w/2PPC and 2 UAC5's. Highly effective. The key is not to hold down fire button but use for quick shots and the occasional double shot which can turn a 30pt shot to a 40 pt shot, for a sniper that's a massive difference.

http:/www.youtube.com/watch?v=56L0Sgpq63w

http:/www.youtube.com/watch?v=owkHxfiKjqU

Edited by xMEPHISTOx, 26 October 2013 - 01:25 AM.

[Bhael Fire]

DOUBLE-SHOT mode should be manually activated by the player, not automatic.

tl;dr:

double click = double fire.