20 replies to this topic

[lordkrike]

I am an unemployed mathematician and I have decided to teach myself Python. As a fun project, tonight I decided to script a simulation to compute the DPS and Rate of Fire of the UAC/5.

What's that, you say? It's just on Smurfy's mechlab? Well, sure, if you want to fire it in single-shot mode. But the question I was interested in is how much dps the UAC/5 does when fired in its double-tap mode, on average, including time for jams.

So what I did was simulated someone pulling the trigger on their UAC/5 for as fast as possible for 1.5 million hours of gametime and averaged the results. The results, rounded to the nearest thousandth:

Total shots taken: 86,292,395
Average damage/s: 4.794
Average time between shots (including jam time): 0.959s
Average time between jams (including jam time): 7.299s

So it turns out that the UAC/5 does marginally increased average DPS (about 5.5% more) if fired as fast as possible as opposed to single shot mode. On average, you're going to jam on the 6th shot.

There were also a few people around here suggesting that the UAC/5's jam rate be adjusted to 20% (as opposed to the 25% it has right now), so I simulated that as well.

Average damage/s: 5.357
Average time between shots (including jam time): 1.071s
Average time between jams (including jam time): 8.400s

Which is a 17.8% damage increase! A little bit goes a long way.

For extra silliness, a 15% jam rate does 6.027 DPS, which is a 32.6% damage increase

I screwed up! I only was only adding 4 seconds for a jam, but I forgot that your gun has to finish cooling down on top of the 4 second wait for a total of 5.1 seconds! Sorry! The revised numbers:

Total shots taken: 74999561
Average damage/s: 4.167
Average time between shots (including jam time): 1.200s
Average time between jams (including jam time): 8.402s (this accurately reflects the increased 1.1s between shots)

So that's actually 8.33% weaker on average over time!

For a 20% jam chance:

Average DPS: 4.736
Average time between shots (including jam time): 1.056s
Average time between jams (including jam time): 9.499s

For a 15% jam chance:

Average DPS: 5.442
Average time between shots (including jam time): 0.919
Average time between jams (including jam time): 11.337

Sorry, I'm human and make mistakes.

If you'd like to see my script, you can download it here.

tl;dr don't pull that trigger!

Edited by lordkrike, 10 January 2013 - 07:18 PM.

[nungunz]

I'd also take heat generation into account.

Good work none-the-less.

What compiler do you use, by the way?

Edited by nungunz, 10 January 2013 - 06:44 PM.

[lordkrike]

Python 2.7 32-bit.

I was originally playing around with numpy and scipy, which is why they are imported in that script. I will go ahead and remove the import statements.

The average heat per second is the same as the shots per second, since each shot generates one heat. And the peak heat generation can't get over ~1.82 for a UAC/5 anyway. I didn't want to deal with heat since heat dissipation can vary so much.

[General Taskeen]

Could you also test a UAC/5, without double-shot mechanic, and instead 2x better fire rate, with jam probability?

So Theoretically

(5 Damage, 0.85 Cool Down, 25 or 20% Jam Rate, 5.8DPS)

[Deamhan]

5/.85 = 5.88dps

If you consider a jam rate of 20%

5.88 - 20% = 4.704

[lordkrike]

General Taskeen, on 10 January 2013 - 07:00 PM, said:

Could you also test a UAC/5, without double-shot mechanic, and instead 2x better fire rate, with jam probability?

So Theoretically

(5 Damage, 0.85 Cool Down, 25 or 20% Jam Rate, 5.8DPS)

Sure!

I simulated an AC that has a .85s cooldown, does 5 damage per hit, and each shot has a 25% chance to jam. On the occurrence of a jam, 4 seconds is added to the .85 second cooldown. The results were... disappointing. It has twice as many chances to jam, since the way the UAC/5 is now, it can only potentially jam on every other shot.

Total Shots Taken: 48652786
Average DPS: 2.703
Average cooldown: 1.850
Average time between jams: 7.401

So I lowered the jam chance to 10% and saw what happened. Still not very good.

Average DPS: 4.000
Average cooldown: 1.250
Average time between jams: 12.499

And then I tried the tabletop value of 1/36th... Not crazy, but still probably too good.

Average DPS: 5.203
Average cooldown: 0.961
Average time between jams: 34.632

Edited by lordkrike, 10 January 2013 - 07:32 PM.

[lordkrike]

Another bit: you could remove the chance to jam on the first shot for the alternate UAC/5, so it would be an AC that shoots every .85 seconds, does 5 damage per hit, and has a 25% chance to jam for an additonal 4 seconds on every shot after the first.

A 25% jam rate is still disappointing:


Average damage/s: 3.03035033131
Average time between shots (including jam time): 1.649974245
Average time between jams (including jam time): 8.25013682028

A 1 in 15 (6.667%) jam rate puts it about where the UAC/5 is in single fire mode.


Average damage/s: 4.546
Average time between shots (including jam time): 1.100
Average time between jams (including jam time): 17.601

Keep in mind that long times between jams makes UAC/5s much better in practice than their average DPS suggests, since they basically frontload their damage until they jam. So a UAC/5 with a similar DPS but a longer time between jams will typically fare better in most combat scenarios.

I also realized that it's 1.5 million minutes. Eh. Whatever, still more than long enough.

[nungunz]

You know. Given paired scenarios, if you send me data files on the different situations, I can run a statistical test based on different type one and two tolerances and come up with a significance analysis.

Probably go on the easier theoretical limits from your first post or two before trying to figure in heat.

XD

[Riffleman]

So how does that compare to just the regular AC 5 without jams. What precent higher is the damage. Because If its negligible id switch, they always jam at the worst times.

[VXJaeger]

Imho high-DPS on ballistic weapon is not so bad as in missiles or lazors. Difficulty of aiming correctly evens it efficently.
So even if UAC5 had DPS of 5, combat situation DPS would be much lower.

[Orzorn]

VXJaeger, on 10 January 2013 - 11:41 PM, said:

Imho high-DPS on ballistic weapon is not so bad as in missiles or lazors. Difficulty of aiming correctly evens it efficently.
So even if UAC5 had DPS of 5, combat situation DPS would be much lower.

It doesn't help that the threat of jamming often causes users (at least myself) to not double tap out of fear of losing the weapon right when we most need it. I tend to double tap if shooting at an enemy not looking at me, and also double when ganging up on someone or they're nearly dead, but I don't like double tapping if we're both in a serious spot. Having a UAC lock up and then getting smacked across the face sucks hard.

Edited by Orzorn, 10 January 2013 - 11:44 PM.

[VXJaeger]

Orzorn, on 10 January 2013 - 11:44 PM, said:

Having a UAC lock up and then getting smacked across the face sucks hard.

Jeah. That's why I kicked UAC5s out of my builds and replaced them w/ AC2 or AC5s. At least they don't spend 50% of combat time clearing their jams.

[lordkrike]

VXJaeger, on 10 January 2013 - 11:47 PM, said:

Jeah. That's why I kicked UAC5s out of my builds and replaced them w/ AC2 or AC5s. At least they don't spend 50% of combat time clearing their jams.

If you can take advantage of the increased cooldown but avoid jamming on the UAC/5, it's objectively a better weapon than the AC/2, AC/5 and AC/10.

Riffleman, on 10 January 2013 - 10:09 PM, said:

So how does that compare to just the regular AC 5 without jams. What precent higher is the damage. Because If its negligible id switch, they always jam at the worst times.

The regular AC/5 does 5 damage every 1.7 seconds for 2.941DPS. A UAC/5 fired in single shot mode does 5 damage every 1.1 seconds for 4.455DPS. A UAC/5 fired in automatic mode does approximately 4.167DPS, on average.

For very short engagement windows, the UAC/5 can still be a good weapon. If you only have (at most) three seconds to hit the target, and then you will be safe for the next 5-6 seconds, bursting from your UAC/5 puts out more damage than a regular AC/5. I'll plot the damage vs time tables so you guys can see them.

[Apoc1138]

VXJaeger, on 10 January 2013 - 11:47 PM, said:

Jeah. That's why I kicked UAC5s out of my builds and replaced them w/ AC2 or AC5s. At least they don't spend 50% of combat time clearing their jams.

I still use UAC's... I just found that in close range combat, the added kick they add for the first few seconds can never be made for by not having jamming in other lower powered weapons like the straight 5 or AC2 - the promised rate of fire of the AC/2 never seemed to actually happen or it couldn't deliver damage consistently enough to counter act the jamming

on my atlas I run UAC's with large lasers or 4 mediums, so that when the AC's do jam you still have decent damage being kicked out, and when they are running you can either try to minimise jamming by letting them cooldown and still get much better results that AC5's, or you can run the risk to get that kill shot in before they jam again

and on my cataphract I run 2 UAC'5 and 2 AC'2 which you can chain fire and almost never jam but still get a higher rate of fire than and more damage output than even quad AC5's, though marginally less pin point damage when alpha firing for say sniping, but then if you want to snipe you are better off with dual gauss

[lordkrike]

So, it turns out that I found something interesting when making the damage/time tables for each respective fire mode. The average damage into perpetuity doesn't exactly tell the whole story.

The DPS of the UAC/5 in double fire mode depends critically on the ability to sharply limit engagement time. If you're firing in a combat and you jam at the last moment before you enter cover to disengage, automatic mode does pretty well. On average, for shorter time periods, you can do more damage than in semiautomatic.

We're talking really short time periods here. During an engagement window of 1.1 seconds, the fully-automatic mode does an average of 15.45 damage for a very impressive 14.04DPS, while semi-automatic mode can only to ten damage (for a still impressive 9.09 DPS). Remember that during an engagement time of 1.1 seconds, a UAC/5 gets two chances to either single shot or to burst (one at zero seconds, one at 1.1 seconds). It drops off to slightly below the semiautomatic mode really, really quickly, though. And if you get caught out of cover while your weapon is clearing a jam, it falls into the sink, since if you had fired in semiautomatic mode you could be returning fire for an additional four seconds.

So, in the chart below, SEMI represents the damage curve over time for firing in semiautomatic mode. AUTO represents the damage curve over time for firing in fully-automatic mode. Notice for about up to three or four seconds, auto mode is pretty good, and then it tapers off to meet up with and eventually fall below SEMI mode. AUTO + JAM TIME is just the auto curve shifted over to the right by its average engagement time (which is how long it takes, on average, for the UAC/5 to do that much damage, including time for jams). That last curve tells us a good comparison between the automatic and semi modes when you are out of cover and can't limit your engagement time.



So, if you know you can dive into cover immediately after jamming or know that you'll have <4 seconds to shoot at the enemy and plenty of time to clear the jam, auto mode is fairly good. The problem is you never know exactly when you're going to jam, and if you get caught out of cover while your weapon is jammed, you're both taking more damage and doing less.

Edited by lordkrike, 11 January 2013 - 07:49 AM.

[Agent 0 Fortune]

I don't know what the math you said, but I like me some UAC5, the more the better.
Sometimes I chain them for suppression fire.
Sometimes I do timed burst for sniping
But sometimes I I just like to hold the trigger down and watch the world burn.

[Revo13]

The UAC5 is boss-hoss.

/endthread

[JSparrowist]

I get jams on first and second shot all the time. I dunno what is up, even chaining them I get constant or first shot jams.

AC/5 for me! Eff UAC/5 in the A!

[MavRCK]

Great post!

Mathematics to back up the game is always a good thing!

Thanks!

TL;DR: PGI please fix the UAC5!

[Tennex]

+1