38 replies to this topic

[MustrumRidcully]

It's time for an update on my weapon efficiency charts.
If you're not familiar with how they work:

Spoiler

Overview
Basically, I made excel sheets that calculate how much ammo and heat sinks you need to use a specific weapon ("boated") to deal a particular total damage result, within a certain time frame, a certain number of engagements. The basic idea comes from a Closed Beta concept someone suggested, "Targeted Engagement Time". The assumption is - you don't need to fire forever with any weapon, you only need to fire as long as there is a target around too shoot at. Which can end depending on whether it retreats or is destroyed.
A weapon is more efficient than another weapon if you need less weight to achieve the same result.

Meaning of different Engagement Times
The different time frames represent the length of a single engagement - for example, a sniper might take something like 5 seconds and then get to cool off, while a brawler that cant just move into cover might take 15 seconds to kill or be killed.

How to calculate the Efficiency Stat
The efficiency stat is then the actual damage such a build would generate in the alloted time, divided by the weight used for the weapons, heat sinks and ammo required. Engine heat sinks are not part of this weight - I assume a 250 rated engine (that means for double heat sink mechs, you get the equivalent of 20 free single heat sinks, and with standard heat sinks, you get the usual 10 free single heat sinks).

Plotting the Results
I then plot these values on a chart, from lowest to longest range weapon. (Though I still seperate SRMs and LRMs from the rest, since they are more difficult to compare since they spread damage far more than the other weapons.)

Caveats
Caveat Range
I don't track range directly. My assumption is, if all things are going well, then the efficiency should drop with range. How much they need to drop is something that I haven't determined yet, and might be for someone else to research.
Caveat Gauss
I only calculate the charge-up cycle in form of an increase in the overall cooldown of the Gauss Rifle. I don't track the specific changed "firing rythm" of Gauss.
Caveat Ghost Heat
I don't track Ghost Heat either. If I were to do it, I would track it assuming that people avoid the ghost heat and instead chain-fire when they go beyond the max alpha, which would also create a special firing rythm. But I don't feel like implementing that in Excel. Maybe if I get bored enough* I write a program for this, but that is unlikely.

More Information
A more detailed explanation can be found in this post from the archive.

The Charts
I created two charts each for 3 different engagement scenarios. The engagement scenarios in one chart basically describe builds that can achieve similar DPS. This gives you an idea how more difficult it becomes to sustain your damage over time. The total damage over all engagements is roughly constant.

The most left figure in chart usually implies something suitable for a sniper that fires only for a few seconds and then gets back into cover for safety and cooling, and has to do this more often to get his kills. The most right figure usually implies somethnig suitable for a brawler that stays engaged in combat for a long time, and will usually not do that all that often, because an engagement might usually already lead to either him or the enemy dying.

The two charts then are divided into a chart with a relatively high damage goal only the heavier mechs can hope to achieve (because only they can equip enough weapons), and one with a lower damage goal that might be feasible for lights and mediums.

High Damage Chart
Here is a chart for 3 high damage output volumes.
(Something achievable with 2 AC/20s, 3 PPcs or 6 Medium Lasers)


Low Damage Chart
And here is one for lower damage output:
(Something achievable with 2 PPCs or 4 Medium Lasers, or 1 AC/20)


Please not that I can't guarantee that the MG and Flamer values are accurate, I might stll have a mistake there.


Analysis
It is notable that in general, the low range bracket weapons are more efficient than the long range bracke tweapons, but there is definitely not a steady drop off with range.
Ballistics seem to gain notable efficiency the longer you use them, quickly surpassing energy weapons which seem pretty steady.

Interesting is that AC/5 and LBX10-X are very similar in their efficiency, but I am still ignoring pinpoint precision vs spread in these calculations, so the LB10-X might indeed be weaker from an efficiency point of view.
As one might expect, the UAC/5 is now less efficient than the AC/5 when not used in double shot mode, but more when used in it. (The damage calcaluation does take the jam rate into account, so the efficiency can be far higher, until it considerably dropped during jam times, leading to the average in the chart. Depending on how well you can manage your jams and mitigate them with retreats, you can get better results)

Spoiler

*)or enthusaistic enough. But I don't feel like the work would be worth it right now, it's not like the devs use these statistics.
EDIT: I corrected a mistake in the laser statistics, the LL used a different weapon damage stat and the beam duration was not calculated correctly.

Edited by MustrumRidcully, 18 October 2013 - 12:39 AM.

[Karl Streiger]

MustrumRidcully, on 11 October 2013 - 12:48 AM, said:

Caveat Range:
I don't track range directly. My assumption is, if all things are going well, then the efficiency should drop with range. How much they need to drop is something that I haven't determined yet, and might be for someone else to research.

I just made a quick example - and calculated inflicted damage of a weapon type with the hits i achieved.

As I thought -- primary weapons got a really high rating - of course lasers have the most reduchtion - and in case of emergency secondary weapon is really disillusioning how low the damage is - even when you hit the target (30%)

That means most primary weapon shots were fired within optimum range. Only the PPC with 72% damage - is an exception. That means i fired that weapon often at ranges were it only dealt hardly damage.

Strange things:

Spoiler

Interesting is that ER-PPC, AC 2 and AC 5 had a positive discrepance between optimal damage (i dealt more damage with ER-PPC - than i hit my targets - 105% ...means out of 10 shots i dealt 105 damage instead of 100...maybe some missed shots are calculated in here...really strange is the AC 2 with 180%!!!!

So in case of my - few values - it is not necessary to use range modifers.

That in mind - that players know what weapons they mount, and what this weapons demand...the range and drop off damage can be ignored - at least also because of the few open spaces were extreme range can be used for good effect.

a factor for SRMs = 75%
a factor for LRMS = 90%
a factor for Lasers = 30-80% (not representive i think)
other ballistic weapons and ER PPC > 95 to 180 %
PPC 75%?

---

demand.
i really hope other players take a look in there weapon stats= columns are:.
damage / hit / base damage (LLAS 9, PPC 10....)
of each weapon....

Edited by Karl Streiger, 11 October 2013 - 01:44 AM.

[Rascula]

Damn fine work there, excellent stuff.

[MustrumRidcully]

I see, you're volunteering to be someone else, Karl. Haxcellent.

Theories regarding the over 100 % damage values:
1) Your guns are trained by a German soccer trainer, and are taught to give 110 %.
2) You triggered a lot of ammo explosions with them. (It's what some people ammo explosion damage is tracked as damage you inflicted, but I don't know if it's actually the case.)
3) Crits now deal extra damage against internal structure, maybe you managed to land a lot of crits?
4) Maybe there is a secret splash radius set for some weapons.
5) There is a bug in the damage tracking.
6) There is a bug in the actual application of damage.

The AC/2 value is really weird IMO. 5-10% difference I could see with crit bonuses and ammo explosions. 80 %? Whoah!

Edited by MustrumRidcully, 11 October 2013 - 02:32 AM.

[Karl Streiger]

MustrumRidcully, on 11 October 2013 - 02:27 AM, said:

The AC/2 value is really weird IMO. 5-10% difference I could see with crit bonuses and ammo explosions. 80 %? Whoah!

Yep - i only had a test run - and i'm pretty sure i dealt more damage to the area but the enemy...but the stats say something different - and i have developed the Swarm AC 2 ...if the shot miss it travells to the next target.
Was on Tourmalin Desert (althoug i can remember that i got AC 2 hits from the right flank - although the enemy was at the left????)

Edited by Karl Streiger, 11 October 2013 - 02:40 AM.

[Kaspirikay]

So if I'm reading this correctly, Small lasers are the best weapons in terms of efficiency?

[Karl Streiger]

Kaspirikay, on 11 October 2013 - 03:34 AM, said:

So if I'm reading this correctly, Small lasers are the best weapons in terms of efficiency?

Based on Weight Heat and Damage - if you are able to close to your enemy yes.

Of course even 1,000,000 Small Laser won't help you to fight a Mech with a single MLAS at 270m

Edited by Karl Streiger, 11 October 2013 - 03:36 AM.

[Aym]

That AC/2 number is very interesting, might be an otherwise unknown bug.

[Appogee]

Kaspirikay, on 11 October 2013 - 03:34 AM, said:

So if I'm reading this correctly, Small lasers are the best weapons in terms of efficiency?

In practice, while you're getting in close enough to use them, you probably take a LOT of damage.

I keep trying them and then removing them in favour of MLs. In my experience, being able to hit from further away, and staying alive longer because I'm being hit less, significantly improves the effectiveness of MLs vs SLs.

[William Mountbank]

When I have mechs that focus on a few long range weapons but still have energy slots free (Stalker, Quickdraw etc...), I find the SL to be very effective and heat efficient as a backup close range weapon for when the action comes to close, though I'm not sure I would want to run them as a primary weapon on anything other than a maxed engine Jenner.

[Skunk Wolf]

Ammo explosion damage and extra crit damage is factored into the damage calculation in our profiles.

Does your data take these variables into account?

Your playstyle can swing these horribly.

Lets say you are an Atlas pilot with two LB10X's who dual drops with his friend running the same. Your style is the common one of sitting behind a hill and waiting five minutes into the match (plinking with a little LRM sometimes) to engage and mop up after the lighter mechs stripped the armour off of the other team, easily one shotting their CT's with the dual LB10X's or getting lots of ammo explosions.

If you are a heavy or medium pilot that is consistently on the front line, you are only doing armour damage before you are CT alpha'd to death by multiple mechs, so no ammo explosion or crit damage bonus.

[MustrumRidcully]

Kaspirikay, on 11 October 2013 - 03:34 AM, said:

So if I'm reading this correctly, Small lasers are the best weapons in terms of efficiency?

Yes, which is what it should be, considering its range.

There are some practical limitations, like how many hard points you would need.
The "high damage" version requires 7-9 SLs, most mechs can't equip that many lasers. The Hunchback 4P is pretty much the only one that can install that many. You would also need 9 tons of weapon and external DHS for the 10-second engagement already, for a 15 second engagement, it grows to 19.5.

[MustrumRidcully]

Just to see if my charts could be of any use at all.

Some People claim that ballistics are OP. I am not sure if that's actually true, but let's see.

Targeted Engagement Times: 5, 10, 15
Targeted Engagement Damage: 60, 120, 180
Targeted Engagements: 12, 6, 4

Large Laser (450m range, 900m max range) - Damage/Weight Efficiency: 3.6; 3.4; 4
AC/10 (450m range, 1,350m max range) - Damage/Weight Efficiency: 2.1; 3.8; 5.3
=> Large Laser has an advantage if the engagement lasts 5 seconds, but loses at 10 and 15 seconds.

Medium Laser (270m range, 540m max range) - Damage/Weight Efficiency: 10, 6.3; 6
AC/20 (270m range, 810m max range) - Damage/Weight Efficiency: 2.5, 3.8, 5.2
=> The Medium laser remains superior at all targeted engagement times.
(Note however that you need 4 times as many MLs than you need AC/20s - hard point restrictions can make this impractical.)

AC/5 (620m range, 1600m max range) - Damage/Weight Efficiency: 2.1; 3.9; 6.1
PPC (540m range, 1080m max range) - Damage/Weight Efficiency: 2.9; 2.4; 2.7
=> The PPC is superior to the AC/5 for 5 seconds, but afterwards is quickly surpassed by the AC/5.

So my method might suggest that the general observation is true at long ranges, which might explain why we see a preference for ballistic based assaults and heavies. You quickly run into hard point problems at close range, too - 8 medium lasers are available only to the HBK-4P, for example (and on that mech, the build with 8 medium lasers doesn't satisfy the requirements for the TET.)

Edited by MustrumRidcully, 14 October 2013 - 10:36 AM.

[Kazairl]

MustrumRidcully, on 14 October 2013 - 01:02 AM, said:

Just to see if my charts could be of any use at all.

Some People claim that ballistics are OP. I am not sure if that's actually true, but let's see.

Targeted Engagement Times: 5, 10, 15
Targeted Engagement Damage: 60, 120, 180
Targeted Engagements: 12, 6, 4

Large Laser (450m range, 900m max range) - Damage/Weight Efficiency: 2.3; 2.5; 2.9
AC/10 (450m range, 1,350m max range) - Damage/Weight Efficiency: 2.1; 3.7; 5.1
=> Large Laser has an advatage if the engagement lasts about 5 seconds, but yields its advantage to the AC/10 over 10 to 15 second long engagements.

Medium Laser (270m range, 540m max range) - Damage/Weight Efficiency: 10, 7.1, 6.7
AC/20 (270m range, 810m max range) - Damage/Weight Efficiency: 2.4; 3.5; 5
=> The Medium laser remains superior at all targeted engagement times.
(Note however that you need 4 times as many MLs than you need AC/20s - hard point restrictions can make this impractical.)

AC/5 (620m range, 1600m max range) - Damage/Weight Efficiency: 2.1; 3.7; 5.1
PPC (540m range, 1080m max range) - Damage/Weight Efficiency: 2.9; 2.4; 2.7
=> The PPC is superiro to the AC/5 for 5 seconds, but afterwards is quickly surpassed by the AC/5.

So my method might suggest that the general observation is true at long ranges, which might explain why we see a perference for ballistic based assaults and heavies.

Another consideration is this data assumes a default starting heat. But sometimes you will only have 2-5 secs between engagements. In those situations the Ballistics will still be superior.

[MustrumRidcully]

Kazairl, on 14 October 2013 - 04:29 AM, said:

Another consideration is this data assumes a default starting heat. But sometimes you will only have 2-5 secs between engagements. In those situations the Ballistics will still be superior.

I think the 10-15 second engagements length already cover this well enough, to be honest.

If you fight 10 seconds, rest for 5 seconds, and then fire again for 10 seconds, your performance will be close to that of a 15 second engagement.

[Kazairl]

MustrumRidcully, on 14 October 2013 - 05:49 AM, said:

I think the 10-15 second engagements length already cover this well enough, to be honest.

If you fight 10 seconds, rest for 5 seconds, and then fire again for 10 seconds, your performance will be close to that of a 15 second engagement.

Agreed. Although I would be curious to see what the average engagement time was per map.

Edit: What is DMG?

Edited by Kazairl, 14 October 2013 - 07:50 AM.

[MustrumRidcully]

Kazairl, on 14 October 2013 - 07:42 AM, said:

Agreed. Although I would be curious to see what the average engagement time was per map.

Edit: What is DMG?

DMG = damage

That'S the target damage that a build has to achieve within the targeted engagement time. And then it has to do it multiple times, to account for ammo requirements in a "reasaonble" match.

[Jman5]

Hey Mustrum, I really appreciate the effort that went into this. However, if you want to reach a wider audience, you really should try to figure out a way to explain the graphs better. I spent quite a bit of time staring at the graphs unsure what I was looking at. Perhaps labeling your Y axis might help slightly.

Something interesting to see would be how much the crit weapon efficiencies change in the damage equation when you're hitting internals. Not only do they have increased chance to crit, but the damage values spike.

Edited by Jman5, 14 October 2013 - 08:12 AM.

[MustrumRidcully]

Jman5, on 14 October 2013 - 08:12 AM, said:

Hey Mustrum, I really appreciate the effort that went into this. However, if you want to reach a wider audience, you really should try to figure out a way to explain the graphs better. I spent quite a bit of time staring at the graphs unsure what I was looking at. Perhaps labeling your Y axis might help slightly.

Something interesting to see would be how much the crit weapon efficiencies change in the damage equation when you're hitting internals. Not only do they have increased chance to crit, but the damage values spike.

I updated the charts with some corrections and an additional y-axis title.
Maybe I have to stop trying to also account for missiles and separate them into a different chart to achieve better readability. Stuff gets too small.

I also noticed that the Large Laser data I used was wrong, and then noticed that the beam durations weren't correctly added to the overall cooldown. I'll update my above analysis with newer values.

---

Regarding the comparision between weapons. I used the high-damage efficiency values. Things can change when you go to the low-damage efficiencies.

For example:
Large Laser: 3.6, 5.4, 4.6
AC/10: 2.1, 3.4, 4.8
The AC/10 only gets to exceed the LL performance once you go to 15 seconds.

Medium Laser: 10, 12.9, 10
AC/20: 2.1, 3.6, 4.7

The low damage value scenarios are builds that are more reasonable on a light or medium mech. So as tempting as it might seem to equip AC/10 or AC/20s there, you're definitely better off with lasers. An AC/20 Raven might give you the Hollander Feel and an AC/10 Spider might give you the Urban mech feel, but they are wasteful builds.

---

The difference between a high and low damage builds also implies that weapon balance can be weight or mech dependent. I blame the heat threshold. a heat capacity of 50 (basically the minimum anyone has with 10 engine double heat sinks) is a lot compared to 4 medium lasers, but not so much for 8 medium lasers.
I see it just as another example of how the low dissipation/high heat capacity system makes weapon balance so much harder.

Edited by MustrumRidcully, 14 October 2013 - 10:44 AM.

[keith]

MustrumRidcully, on 14 October 2013 - 01:02 AM, said:

Just to see if my charts could be of any use at all.

Some People claim that ballistics are OP. I am not sure if that's actually true, but let's see.

Targeted Engagement Times: 5, 10, 15
Targeted Engagement Damage: 60, 120, 180
Targeted Engagements: 12, 6, 4

Large Laser (450m range, 900m max range) - Damage/Weight Efficiency: 3.6; 3.4; 4
AC/10 (450m range, 1,350m max range) - Damage/Weight Efficiency: 2.1; 3.8; 5.3
=> Large Laser has an advantage if the engagement lasts 5 seconds, but loses at 10 and 15 seconds.

Medium Laser (270m range, 540m max range) - Damage/Weight Efficiency: 10, 6.3; 6
AC/20 (270m range, 810m max range) - Damage/Weight Efficiency: 2.5, 3.8, 5.2
=> The Medium laser remains superior at all targeted engagement times.
(Note however that you need 4 times as many MLs than you need AC/20s - hard point restrictions can make this impractical.)

AC/5 (620m range, 1600m max range) - Damage/Weight Efficiency: 2.1; 3.9; 6.1
PPC (540m range, 1080m max range) - Damage/Weight Efficiency: 2.9; 2.4; 2.7
=> The PPC is superior to the AC/5 for 5 seconds, but afterwards is quickly surpassed by the AC/5.

So my method might suggest that the general observation is true at long ranges, which might explain why we see a preference for ballistic based assaults and heavies. You quickly run into hard point problems at close range, too - 8 medium lasers are available only to the HBK-4P, for example (and on that mech, the build with 8 medium lasers doesn't satisfy the requirements for the TET.)

reading this pgi kinda did their job. though the energy based weps should be balanced around lasting 15 or so secs. a great quick and easy fix would be tweak the DHS value.