37 replies to this topic

[MustrumRidcully]

When discussing different weapons, and particularly energy weapons vs ballistics, one thing that is often coming up is how much tonnage you actually need to invest - ballistics are expensive up front, but are cooler, so you don't need extra heat sinks, but if you equip a few energy weapons, you most likely will load up a lot of extra DHS.

So, this is a thread that tries to figure out how much tonnage you actually need to invest for certain goals.

Based on my previous work on weapon efficiency, I wanted to present a bit more immediately understandeable figures. Instead of trying to determine the efficiency (damage/tonnage) of weapons, I simply give the tonnage you need to invest to run certain weapons, if you want to achieve certain comparable goals on the battlefield.

I calculate how many tonnage you need to invest on the weapon(s), the heat sinks and ammo to run them, with specific goals in mind. The goals are determined by:
- How long do I want to shoot before I have to cool off?
- How much damage do I want to inflict (at minimum, sometimes you can'T help but deal a bit more) until then?
- And how often do I want to repeat it (before I run out of ammo)

The specific scenarios in these charts are always designed so that they represent the same total damage inflicted, and the main difference is the time - a short time would represent a "bursty" approach to damage that cannot be sustained and requires early retreat, while a long time represents a sustained firefight.

You will sometimes see that the tonnage requirement is the same for every engagement time - that represents usually that the weapon is not hot enough to actually provide additional heat sinks to achieve the same result (that doesn't mean that the mech won't be hotter in the longer scenario at the end, it just won't be so hot that it would need more heat sinks to avoid shutting down before the alloted time is over). Since the total damage is the same for the scenarios in one chart, the ammo requirement remains constant.

Graphs
The first scenario here is a bit lower on damage, which seems more suitable to expect from light and medium mechs - but see the tonnage itself to see if that would work out.

Total Damage After Engagements: 480


As you can see by the tonnage figures, the low range weapons are very economical, and usually, you're better off with energy weapons.


Total Damage after Engagements: 720
The next chart requires a bit more damage output delivered.


At longer ranges, it really becomes more interesting to run ballistic weapons, unless you expect to only ever run with short burst weapons.

Stuff not modeled
There is of course stuff this can't model - How well you aim, for example. But if you know that (you can use your own profile to figure that out), you can start comparing - if two weapons are close, pick the one you're better with, if one is worse, but you're really a lot better with it, it imght still be worth it using that weapon (assuming you can't improve your resuilts with the other).

An aspect that is often talked about is - how important is single damage projectiles vs beams. These charts don't answer this, but if you, for example, figure out how better the one is over the other independent of tonnage requirements, you can also use the tonnage requirements to figure out if that advantage is negated or even strengthened. Or alternatively, you can go back to the calculations the charts are based on and instead of using the full damage value, you could use one modified by your damage efficiency (comparing the theoretical damage you could have dealt with the actual damage inflicted) with a weapon.

EDIT:
Noesis also mentions an important aspect - the model doesn't care how long it wil ltake between engagements to go back to 0 heat and repeat it. I think this is sufficiently modelled by having variable engagement lengths, but you or he might disagree. It's obvious from the charts that the longer the engagement is to last (and an engagement after which you cannot fully recover includes the variant where you simply have to last longer than expeted), the worse the situation becomes for energy weapons.

Edited by MustrumRidcully, 20 January 2014 - 06:34 AM.

[FupDup]

If I'm reading your gigantic graphs correctly, this means that lasers are better for very short engagements (and ranges) but ballistics become better for prolonged fights and can deal higher damage over time?


Isn't BT supposed to be the other way around, with energy as sustainable and ballistics as the big punch that doesn't last long?

[Deathlike]

FupDup, on 18 January 2014 - 07:50 PM, said:

If I'm reading your gigantic graphs correctly, this means that lasers are better for very short engagements (and ranges) but ballistics become better for prolonged fights and can deal higher damage over time?

Ironically, that sounds about right.

[Corvus Antaka]

FupDup, on 18 January 2014 - 07:50 PM, said:

If I'm reading your gigantic graphs correctly, this means that lasers are better for very short engagements (and ranges) but ballistics become better for prolonged fights and can deal higher damage over time?


Isn't BT supposed to be the other way around, with energy as sustainable and ballistics as the big punch that doesn't last long?

yup.

ballistics are: longer range, lesser heat, no true ammo concerns & pinpoint dmg

energy weapons PPC meets this criteria highly efficently as well. thus the obvious ballistic/ppc combo. the others spread damage or run quite hot.

cut ballistic ranges to proper values, 270/540 for ac/20, etc, and with only the gauss, erppc and erlarge as top sniping guns you'll probably see more brawlers start pushing forward as more viable mechs again.

[Monky]

Part of this is because ballistics have to go into the super high ammo counts to be considered viable, eating up tonnage and space that could otherwise be dedicated to heatsinks. In order to bother with ballistics in a serious manner, you're sacrificing loads of tonnage and crits, and therefore unable to use energy as anything other than a momentary backup or you risk overheating. Some exceptions due to poptarting do exist of course.

The real question is whether this is acceptable from a balance and game enjoyment pov. Personally, I'd like tonnage/crits to loosen up a bit for ballistics as it allows more chassis to use a wider variety of builds and makes more possibilities viable. It isn't terrible as is however, mainly it's just jump sniping that prevents most builds from being at or near the top of the pile, if it wasn't for easily accessible cover almost anywhere that you can pop out of from any location by jump jetting up over it, you'd have a lot more going on in the mix of mechs out there.

[Toong]

FupDup, on 18 January 2014 - 07:50 PM, said:

If I'm reading your gigantic graphs correctly, this means that lasers are better for very short engagements (and ranges) but ballistics become better for prolonged fights and can deal higher damage over time?


Isn't BT supposed to be the other way around, with energy as sustainable and ballistics as the big punch that doesn't last long?

Yes, but people in MWO typically carry triple the amount of ammo of most "vanilla" BT 'mechs. In short engagements energy weapons win out, and once heat becomes an issue ballistics come to the forefront, it's true. But if you allow for more time, energy weapons will win out again as ballistic weapons run out of ammo. Energy weapons can literally fire forever, assuming they aren't destroyed; you just have to wait for the guns to fall silent. Whether or not you'll still be alive to fire them once the guns run dry, though, is a separate argument.

[Kyle Reece]

Mmm, so I've read through the OP at least 6 or 7 times. Some of that with the help of vodka and I'm still not getting this.

Can you do a worked example of exactly what you did to get these numbers? Maybe doing the Small Laser and the ERPPC so both ends of the scale are represented?

[MustrumRidcully]

Kyle Reece, on 19 January 2014 - 03:03 AM, said:

Mmm, so I've read through the OP at least 6 or 7 times. Some of that with the help of vodka and I'm still not getting this.

Can you do a worked example of exactly what you did to get these numbers? Maybe doing the Small Laser and the ERPPC so both ends of the scale are represented?

Okay, keep in mind that you usually work the other way around - build a mech and then try to figure out how long it will be able to fire and what its damage output is to be.

You want to equip a mech with small lasers and be able to fire for 10 seconds, dealing 80 damage in that time. You also ant to repeat that 6 times, but that's not so important, as small lasers don't run out of ammo.

A Small Laser deals 3 damage and produces 2 heat per shot. A small Laser fires every 2.25 seconds for 0.75 seconds, meaning its total cycle time is 3 seconds.
In 10 seconds, that means you can fire it 4 times. (1st time at 0 seconds, 2nd time at 3 seconds, 3rd time at 6 seconds, and 4th time at 9 seconds).
4 shots would deliver only 12 damage, so you need additional small lasers. 6 Small Lasers would give you only 72 damage, which is a bit too little, so you need 7 Small Lasers. (This means you deal a bit more than 80 damage.)
7 Small Lasers weigh 3.5 tons.
4 shots with 7 small lasers produce a total of 4 * 7 * 2 = 56 heat.
Your heat threshold without additional double heat sinks and only the 10 base engine sinks is 50. Over 10 seconds, 10 engine double heat sinks dissipate 20 heat, so your net heat at the end would be 56-20 = 36,, which is below your heat threshold. So we don't need any additional double heat sinks, and the total weight investmens is only 3.5 tons.



Let's say we now want to build a mech with ER PPCs. We aim to deal 60 damage in 5 seconds.
Within 5 seconds, we can fire an ER PPC only twice (at 0 seconds and 4 seconds). So we need three ER PPCs.
The 3 ER PPCs will produce 90 heat in 5 seconds. Our heat threshold is only 50 if we just take only 10 engine DHS, and we could dissipate only 10 heat in that time, so 10 DHS is not enough - we need more.
After careful evaluation, we figure out that we need 18 additional non-engine DHS (giving us 18 * 1.4 extra heat threshold and 18 * 0.14 dissipation per second). This gives us a heat threshold of 75.2, and a total dissipation of 4.52 heat/second. After 5 seconds, we're at 67.4, but at 4 seconds (immediately after the second salvo) we're at 71.92 heat - so we can avoid overheating the entire time.

So we need 3 PPCs (21 tons) and 18 DHS (18 tons) to enable this build, for a total of 39 tons.

Note that this math doesn't include the possibility of ghost heat raising the heat even further . FOr this particularly scenario, the timing would also work out if we fire a pair of ER PPCs at 0 seconds and a single ER PPC at 0.6 seconds, and still fit it all within the 5 second time frame.

Does this help?

Edited by MustrumRidcully, 19 January 2014 - 05:19 AM.

[MustrumRidcully]

FupDup, on 18 January 2014 - 07:50 PM, said:

If I'm reading your gigantic graphs correctly, this means that lasers are better for very short engagements (and ranges) but ballistics become better for prolonged fights and can deal higher damage over time?


Isn't BT supposed to be the other way around, with energy as sustainable and ballistics as the big punch that doesn't last long?

Keep in mind that this scenario is deliberately designed so that you have exactly the amount of ammo you need, just as you have exactly the heat sinks you need to avoid overheating in the time alloted. I wouldn't try to make any value judgements on whether this is opposed to the table top idea of auto-cannons being non-sustainable in the long run. THis is basically still the same - if for example after 4 180 damage 15 second engagements you actually would need to make a 5th, the AC/20 user would be out of ammo and the ER PPC user could still take a try.

The basic distinction in the table top is also only that ballistics are heat efficient and can sustain fire as long as they have ammo, while energy weapon use can lead to short term bursts, but bursts that can be repeated infinitely.

In MW:O we have hte advantage we know that we're in a 12 vs 12 and everything is brought back to full after that engagement. This makes it possible to estimate how much ammo you need in practice.
(Note that this isn't just important for ammo, it's also important for armour. People like to say that if we had R&R or multiple battles in a row without reload, then ballistics ammo dependency would star advantage would show, but they ignore the fact that you also need repairs after every match, and while it might theoretically sound great to not run into ammo problems, if half your guns and heat sinks have been shot away and you have no armour left for the second combat, it still wouldn't work out well for you - an ammo-depleted Dual AC/20 Jagermech with a single medium laser left might still feel quite threatening to you...)

[Noesis]

MustrumRidcully, on 19 January 2014 - 05:18 AM, said:

Okay, keep in mind that you usually work the other way around - build a mech and then try to figure out how long it will be able to fire and what its damage output is to be.

You want to equip a mech with small lasers and be able to fire for 10 seconds, dealing 80 damage in that time. You also ant to repeat that 6 times, but that's not so important, as small lasers don't run out of ammo.

A Small Laser deals 3 damage and produces 2 heat per shot. A small Laser fires every 2.25 seconds for 0.75 seconds, meaning its total cycle time is 3 seconds.
In 10 seconds, that means you can fire it 4 times. (1st time at 0 seconds, 2nd time at 3 seconds, 3rd time at 6 seconds, and 4th time at 9 seconds).
4 shots would deliver only 12 damage, so you need additional small lasers. 6 Small Lasers would give you only 72 damage, which is a bit too little, so you need 7 Small Lasers. (This means you deal a bit more than 80 damage.)
7 Small Lasers weigh 3.5 tons.
4 shots with 7 small lasers produce a total of 4 * 7 * 2 = 56 heat.
Your heat threshold without additional double heat sinks and only the 10 base engine sinks is 50. Over 10 seconds, 10 engine double heat sinks dissipate 20 heat, so your net heat at the end would be 56-20 = 36,, which is below your heat threshold. So we don't need any additional double heat sinks, and the total weight investmens is only 3.5 tons.

Having a heat cap of 50 and weapons firing 56 means that we cannot fire these weapons as an alpha.

The sustainability "is important" here since heat is effectively the ammo that lasers fire or need to negotiate to be able to fire.

With this model you cannot repeat the process with firing the 7 lasers "twice" without overheating as it has specific targets and isn't addressing the sustainability issue. Something you would not find problematic with AC ballistics.

In fact by comparing the examples given in the original spreadsheet you can see that 2 AC10 would be able to kill 4 targets in the time it takes the equivalent 6 ML to kill 2 targets due to the fact that they would overheat twice in this time. But since the model is based on equivalence to kill targets in a specified time for 1 just below the heat threshold for shutdown for the 6 MLs (with 6 DHS) then by comparison it appears more efficient. In the meantime the AC10 using 3 tons of ammo to kill 4 Mechs but without any problem sustaining fire. In reality the sustainability factor is 50% less then, in fact if you quickly model these two as a comparison in Smurfy you will see that the sustained DPS is 50% less.

As a result for me the examples don't really compare any realistic example of game play as they are using specific targets that don't demonstrate sustainability and therefore is not a good representation of efficiency at all. This since it doesn't take into account the full heat used component of energy weapons as part of MWO mechanics. And ignores a build up of heat prior to shutdown with further sustained fire as not being relevant to Mech performance as a result.

In fact the lack of additional heat sinks that would be applied to these kind of builds seems underestimated imho and from experience (the equivalence of energy ammo needs) and yet the tonnage for ballistics happily included with overestimated values to some extent as part of the modelling, not the actual required needs for the exercise or test conditions.

As such I can't find these examples given as useful representations of actual efficiency as they ignore various needed aspects of game play with heat management.

[MustrumRidcully]

Noesis, on 19 January 2014 - 06:00 AM, said:

Having a heat cap of 50 and weapons firing 56 means that we cannot fire these weapons as an alpha.

4 shots with 7 small lasers are not one alpha, it are 4 alphas. A single alpha of 7 small lasers is only 14 heat. In the 3 seconds to the second alpha, the mech will have time to dissipate some heat. The maximum heat accumluated over time never exceeds the heat threshold.

Quote

In fact by comparing the examples given in the original spreadsheet you can see that 2 AC10 would be able to kill 4 targets in the time it takes the equivalent 6 ML to kill 2 targets due to the fact that they would overheat twice in this time. But since the model is based on equivalence to kill targets in a specified time for 1 just below the heat threshold for shutdown for the 6 MLs (with 6 DHS) then by comparison it appears more efficient. In the meantime the AC10 using 3 tons of ammo to kill 4 Mechs but without any problem sustaining fire. In reality the sustainability factor is 50% less then, in fact if you quickly model these two as a comparison in Smurfy you will see that the sustained DPS is 50% less.

The Smurfy DPS is atually the one irrelvant, since it calculates how much firepower you can sustain indefinitely. But you never shoot for an infinite amount of time, you shoot for a finite amount of time, and that is exactly what I am modelling.


So apparantly, my explaination didn't help.


Let's be more specific with the small laser example:

7 Small Lasers with 10 DHS.
0 seconds (1st alpha): 14 heat from alpha => 21 damage total
3 seconds (2nd alpha): 14 heat from before, -6 heat dissipated over last 3 seconds, +14 heat from new salvo= 28-6 = 22 heat => 42 damage total
6 seconds (3rd alpha): 22 heat from before, -6 heat dissipated over last 3 seconds, +14 heat from new salvo = 30 heat => 63 damage
9 seconds (4th alpha): 30 heat from before, -6 heat dissipated over last 3 seconds, +14 heat from new salvo = 38 heat => 84 damage total
10 seconds (timer runs out): 38 heat from before, -2 heat dissipated over last second = 36 heat => 84 damage total

Edited by MustrumRidcully, 19 January 2014 - 06:35 AM.

[Noesis]

MustrumRidcully, on 19 January 2014 - 06:27 AM, said:

The maximum heat accumluated over time never exceeds the heat threshold.

Your model still excludes the fractional heat build up during this time and does not account for it in the model as relevant to efficiency. In fact this partial build up of heat that would normally be relevant to a pilot in terms of heat management simply ignored to effecting overall performance as a result. Thus not a realistic model of efficiency.

[MustrumRidcully]

Noesis, on 19 January 2014 - 06:35 AM, said:

Your model still excludes the fractional heat build up during this time and does not account for it in the model as relevant to efficiency. In fact this partial build up of heat that would normally be relevant to a pilot in terms of heat management simply ignored to effecting overall performance as a result. Thus not a realistic model of efficiency.

Did you see my edit with the specific timeline.
I account for every heat build-up and the maximum threshold reached within that timeline, there is no point in that time where the heat threshold is ever exceeded.

[Noesis]

And apologies for reading the heat build up as one alpha, my mistake. But the point remains that there is now 36 heat in to be accounted for. Trying to sustain the "similar" process is now not possible due to this heat being there and the Mech will overheat.

The very fact that for energy builds there is now 36/50 heat used for sustained fire means that the weapons have to be managed more as per that firing arrangement. Since it is not sustainable. This residual heat that you seem to want to simply ignore will be very apparent in a real game. As such your model is flawed imo.

[Noesis]

MustrumRidcully, on 19 January 2014 - 06:37 AM, said:

Did you see my edit with the specific timeline.
I account for every heat build-up and the maximum threshold reached within that timeline, there is no point in that time where the heat threshold is ever exceeded.

No because your model excludes its relevance to MWO at all, preventing this aspect of heat management being relevant to game play. Thus it isn't a realistic model, and therefore it is not a good measure of efficiency as a result. Simply ignoring any partial build up of heat to represent other weapons as more effective is simply a distortion of how the game actually plays with the relevance of how much heat is important to the use of energy weapons.

[MustrumRidcully]

Noesis, on 19 January 2014 - 06:40 AM, said:

And apologies for reading the heat build up as one alpha, my mistake. But the point remains that there is now 36 heat in to be accounted for. Trying to sustain the "similar" process is now not possible due to this heat being there and the Mech will overheat.

The very fact that for energy builds there is now 36/50 heat used for sustained fire means that the weapons have to be managed more as per that firing arrangement. Since it is not sustainable. This residual heat that you seem to want to simply ignore will be very apparent in a real game. As such your model is flawed imo.

If the goal of the model was to have the mech at 0 heat at the end of every engagement, then it would be just a heat neutrality calculation. That's much easier, but also irrelevant in practice.

You have 36 heat, and you dissipate 2 heat per second. You'll need to spend 18 seconds to recove from this, basically a relaxation period. For the 10 second / 80 damage scenario, the relaxiation period varies from 0 (MG, Gauss) to 23 seconds (Medium Laser).

Just like the scenario assumes that you will not have more engagements then you have ammo for, it also assumes you'll have this relaxation time between the engagements. One could devise a model that sets a maximum relaxation period, but that relaxation period would also be arbitrary. I think the settings I use now are sufficiently close to what we want to know. More would be gained from figuring out the impact of accuracy and beam vs projectile, to be honest.

No model can replace reality. Keep in mind also - if we didn't get this model balanced, do you think adding a concern for how fast you can recover from your heat will show that things are actually better balanced?

Edited by MustrumRidcully, 19 January 2014 - 07:01 AM.

[Noesis]

MustrumRidcully, on 19 January 2014 - 06:59 AM, said:

If the goal of the model was to have the mech at 0 heat at the end of every engagement, then it would be just a heat neutrality calculation. That's much easier, but also irrelevant in practice.

You have 36 heat, and you dissipate 2 heat per second. You'll need to spend 18 seconds to recove from this, basically a relaxation period. For the 10 second / 80 damage scenario, the relaxiation period varies from 0 (MG, Gauss) to 23 seconds (Medium Laser).

Just like the scenario assumes that you will not have more engagements then you have ammo for, it also assumes you'll have this relaxation time between the engagements. One could devise a model that sets a maximum relaxation period, but that relaxation period would also be arbitrary. I think the settings I use now are sufficiently close to what we want to know. More would be gained from figuring out the impact of accuracy and beam vs projectile, to be honest.

No model can replace reality. Keep in mind also - if we didn't get this model balanced, do you think adding a concern for how fast you can recover from your heat will show that things are actually better balanced?

Ignoring the relevancy of heat build up as you have for me then does not make it an accurate reflection of weapon efficiency.

Also since heat management is the main balancing component for energy weapons the relevance to a discussion about balance as a result makes this model practically useless as it does not incorporate all the appropriated balancing mechanics relevant to MWO that a pilot needs to apply.

[Kyle Reece]

Mmm, I think where you were throwing me was in the practical application of the numbers you've produced. The lack of 'mechs that could mount 7 small lasers to even make it worthwhile to make that comparison was leaving me stuck in "what's the point of these numbers?" land.

Now I see where you're coming from it's an interesting comparison. It needs to be tempered with a number of other metrics to tell the full story, but its a good piece of the puzzle.

I'm wondering if we wouldn't have more success in modelling the weapon table if we first work out a set of example engagements. For example; the scout harasser, the long range pop-tart, the dive in brawler and the jack of all trades. That would then drive the variables of engagement style, damage output, cool-down opportunities etc.

[Nik Van Rhijn]

Actually the hunchie 4P can mount 9, the B'master 1G could mount 7 (bit pointless).

[MustrumRidcully]

Noesis, on 19 January 2014 - 07:13 AM, said:

Ignoring the relevancy of heat build up as you have for me then does not make it an accurate reflection of weapon efficiency.

Also since heat management is the main balancing component for energy weapons the relevance to a discussion about balance as a result makes this model practically useless as it does not incorporate all the appropriated balancing mechanics relevant to MWO that a pilot needs to apply.

Do you think that the situation would improve for the weapons that have excessive tonnage requirements now if we also set a limited amount of time needed between the engagements? Do you think there will be a significant qualitative change in the results (e.g. "Oh, look, ER PPC need a lot less tonnage now than AC/5s! OMG, Small Lasers need more tonnage than AC20s for similar results!)

Edited by MustrumRidcully, 19 January 2014 - 09:38 AM.