Jump to content

Weapon Balance and Heat System - the Current State (2012/10/30)


150 replies to this topic

#1 MustrumRidcully

    Member

  • PipPipPipPipPipPipPipPipPipPipPip
  • Legendary Founder
  • Legendary Founder
  • 10,644 posts

Posted 30 October 2012 - 12:36 PM

During the Closed Beta, there were quite a few threads about balance, and in a large part it revolves around the heat system. Those old threads are unfortunately now gone (or "archived" with currently no way of reading them). So I decided to start a new thread devoted to the topic. This post will probably be long and contain charts and all. So for starters, a TL;DR overview.

TL;DR Overview
Balance has yet to be achieved. There are some weapons that are very difficult to use for little benefit, and some that are very easy to use for much benefit.
If you try to apply various metrics to gather the effectiveness or efficiency of weapons, you will realize that low-heat weapons are simply the superior way to achieve a decent damage output.

As a side effect of the current system, trial mechs or rather all stock configurations are particularly weak - they are usually overgunned and undersinked, e.g. they carry a lot of weapons that,if they really wanted to use them to their full ability, wolud need a lot of heat sinks. These mechs do not play in any way or form like they may have in the original system they were designed for.

The best approach to fixing this is to reduce the ratio between heat generation and heat dissipation, for example by reducing heat values, particularly for high heat weapons - don't be afraid - this doesn't mean heat will be unimportant or meaningless and will not require management. It just means that you will have more viable choices in how to build a mech, and you will still need to make major decisions on how hot or cool you want the mech to be, and hace meaningful trade-offs between sustained damage, burst damage and alpha strike damage.

If you don't like the details but like to see the charts, skip the parts up to the pretty pictures.

PS: If you worry that lowering the heat for some weapons could make heat management to easy, I have something for you that would maintain heat management as important even for people that build heat neutral mechs. In essence: Lower the heat capacity, have a real heat scale.

The Long Version
This is a long version. Pour yourself a coffee or something. It may contain mathematical models and charts based on them.

Going real time with Battletech - PGI Style
The first thing PGI basically did when they implemented MW:Os weapon systems was to take all table top stats - weight, crits, damage, heat. They just made one major addition. The table top game is a turn based board game, it doesn't run on CPU cycles of a few millisesconds, it runs on turns that represent each 10 seconds and take the "human" CPU sometimes minutes to resolve (and this doesn't even consider the decision making process made by the players).
Now, 10 second turns are a nice abstraction, but for a real time game, just taking those table top stats and saying every weapon shoots only every 10 seconds is a bit lame. 10 seconds is just too long to be reasonable.

So PGI gave weapons firing rates. The rates changed over the course of the Beta - most weapons probably have a recycle time of around 4 seconds, but some have notable lower values.

But what PGI when they did this was that they kept the table top stats for damage and heat as a "per shot" value, and still assumed that heat sinks would dissipate 1 heat over 10 seconds (but of course, a bit of heat every time unit).

Unfortunately, what they never seemed to realize was that this approach has a fundemental flaw - it violates a core constraint of the original battletech rules - and that was that each weapon could fire only once per turn (exception: Ultra AUto-Cannons, but those got special rules and adjustments to cover that advantage). Why is that such an important constrant?
Let's compare two extremes - the Gauss Rifle and the PPC.
THe Gauss Rifle weighs 15 tons and produces 15 damage and 1 heat per turn, and it needs some ammo. The PPC weighs 7 tons and produces 10 damage and 10 heat per turn. In the table top, 2 tons of ammo would be generally sufficient, so let's put the Gauss weight at effectively 18 tons - if you spend 18 tons, in weapon, ammo and heat sinks, you can add 15 damage per turn to a mech without changing how fast he overheats. For the PPC, the same value is about 17 tons. Now, we realize already that the Gauss seems like a good deal here, but in the table top game, that was deemed okay, because the Gauss Rifle was on some level meant as an upgrade, a new tech level.
But so, basically, 18 tons add 15 damage to your damage output in 10 seconds, and 17 tons add 10 damage to your damage output in 10 seconds, without changing how quickly he overheats.
But what happens if you allow to fire a weapon twice per turn?
If you already have a Gauss Rifle, all you need to do is add 1 heat sink - and maybe some extra ammo, if you want. So, let's say, 3 tons for 15 damage.
But if you have a PPC, you need to add 10 heat sinks. So 10 tons for 10 damage.
This highlights 2 things:
1) Low heat weapons benefit more than high heat weapns from increased rate of fire.
2) To improve your damage output, it is generally cheaper to add more heat sinks and ammo to utilize your maximum rate of fire, than adding a full new weapon system.

This means that original balancing assumptions between wepaons are no longer true, and that original assumptions on mech builds are no longer true. You really don't need to equip a mech with 3 PPCs and 28 heat sinks - 2 PPCs and 35 heat sinks will deliver you more damage if you consider the sustainability.

This is the core flaw of the MWO weapon system, in a nut shell.
There is of course more. For example - the heat system lacks the details of the Table Top heat scale - there is just the big "endpoint" - shutdown. But all the points in between, where you lost speed and suffered hit penalties - there is nothing trying to represent this. Which in turn means that while heat can overall be very punishing, alpha strikes itself are not that risky at all. But that's just a secondary concern, really.

Where did the mechanics lead us to?
But there are ways to also illustrate this, which I and others went a lot of effort to illustrate.
Before I go into this, just one remark:

This isn't just theorycrafting. In truth, these theories are mostly predated by what we actually see happening during gameplay.

We saw many mechs utilizing the fact that low heat weapons can utilize their ROF much better. It leads us to Jenners with 6 small lasers and the "Slunchback"; A HBK-4P refitted witht the biggest XL Engine and replacing all 8 medium lasers with small lasers, and any spare weight spend on heat sinks.
We see people taking the Catapult K2, ripping out the useless PPCs and equipping 2 Gauss Rifles, because the Gauss Rifle was just outperforming the PPC so much better that all the sacrifices in armor or engine weight are worth it. It just makes a giant difference whether you just have the heat sinks to support two PPCs firing every 10 seconds or the heat sinks to support 2 Gauss Rifles firing every 4 seconds.

The Run Hot or Die Tournament also lead to very similar mech configuraitons - once the game is played competitively and people start optimzing


Illustrating the Balance with Math and Charts
There is one general approach to measure weapon strength - this is generally trying to figure out "how much does it cost me to run this weapon in weight" vs "how much damage does it let me deal this way. Theoretically, one could also use such a system for crit requirements, but this is more rarely done. Though with Inner Sphere heat sinks, it may be an interesting addition.

There are two major ways to figure out the weight, and mostly it is the question - how do we calculate the number of heat sinks for a weapon?

1) Heat Neutrality
We try to put in enough heat sinks so that the weapons heat is perfectly negated.
This is a simple, useful approach, though some thing it's too simple.
Generally, heat neutrality is very useful to know if you are working at modifiying your mech and adding or switching out some weapons. If you have already found a mech configuration that you think is "reasonably hot" or "reasonably cool" (ranging from overcooled to heat neutral to gaining heat slowly up to overheats in the blink of an eye), when you made free some weight, you can simply add the "heat neutral" configuration to the mech and nothing will change about its time to last.
If you replace weapons, you may realize that for the weight of 2 Medium Lasers and their heat sinks, you could also, say, fit another Large Laser - maybe that would give you something you found lacking so far.

2) TET - Targeted Engagement Time
This approach is a bit different from heat neutrality is that you try to figure out - how long do I expect an individual combat/skirmish/brawl (e.g. not the entire match, but one fight in that match) too last, and then find a build that optimizes the damage output in this time frame.
This approach has some very practical applications to a mech or weapon configurations. For example, if you have 2 long range weapons, you may decide that these are mnostly used for sniping, where engagement times are usually low (unless the enemy is not trying to get into cover), but also a few short range weapons which you intend to use in a brawl. So you may try to go for a high TET for the brawling weapons and a low TET for the long range weapons. As long as you don'T use them together, you should get what you want. (Whether it's always the best choice to mix roles like this is another question).

The important aspect of the TET approach is that it has to account for a mech's inherent heat capacity - since that allows you to gain heat without needing to immediately/completely dissipate it. This creates some artefacts - if you really always only calculate the TET for a single weapon, you get unrealistic figures - a single medium laser currently can produce 10 heat over 10 seconds on average, so that the time to overheat is at least 30 (in practice, it's even higher - your heat capacity in MW:O is 30 + heat sinks, not just 30 - and you need to have 10 heat sinks now... So basically, with a single medium laser and 10 sinks, you never overheat.)

So for my calculations, I always used 4 weapons.That can be unrealistic for large weapons, but it's close enough - the real error beings at the low heat levels, not the high ones.

How to Calculate Damage
THere are basically two major damage figures people are interested in:
- Damage Per Shot (or "burst" or in Battletech terms "alpha" damage)
- Damage over Time (or "DPS")
Damage per Shot is an important asset to have - you want to nail weak spots at the enemy mech and a high damage per shot means you can destroy them without firing again.
Damage over Time - You can't one-shot everything and everyone, so you will need to take multiple shots - how much damage over time you can inflict on average gives a hint on how long it will take you.

How to put Damage and Weight together
I have basically chosen 3 approaches in total.

For using heat neutrality, I have used two damage figures.
1) Damage Over Time
2) Damage Over Time + Damage per Shot, weighted against each other. (More precisely, I calculated it as (Damage per SHot + 10 x Damage over Time)/10.

For TET, I just used the damage that a mech would inflict continously firing.

Whatever model I used, I then used the calculated weight against the calculate damage.

What we would like to see when we compare damage efficiency stats of weapons

Generally:
Range has not appeared yet in any of the calculations, but is a major factor in balancing weapons. What we want to see ideally is that weapons with similar range have similar damage efficiency stats, and the longer the range, the lower the damage efficiency. This means there is a meaningful trade-off between damage and range. If you use a short range weapon, you must get close to the enemy first, but if you do, you can outdamage the enemy that relies on long range weapons. This introduces also an interesting tactical element, as each participant will try to fight in his "turf" or range category.

Heat Neutrality + Weighted Damage Values
Here, Range should suffice. We already made burst potential an advantage in the efficiency chart.

Heat Neutrality + Damage Over Time:
Here, Range is one factor, but we don't account for burst yet. As a general guideline - the higher the damage per shot is for a weapon, the lower we want the efficiency to be.

TET
For TET, we have similar concerns as for Heat Neutrality and Damage Over Time - more range and more burst should mean a lower efficiency.

Weapon Efficiency based on the current stats (as of 30th of Octobre 2012)

Heat Neutrality
This chart contains 2 curves - one based on damage over time (DPS/Weight) and one on the weighted values for damage per shot and damage over time (DPS_ID/Weight)

Posted Image

For double heat sinks:
Spoiler


As we can see - the ballistics still outperform the energy weapons mostly, and the general rule - longer range, lower efficiency - is not observed at all for them.
Energy weapons follow the trend in general, though Small Lasers and Medium Lasers still present notable spikes.

SRMs and LRMs seem to play in an entirely different league.

TET Charts
With single heat sinks.
Posted Image

With Double Heat Sinks
Spoiler


The results are similar to the heat neutrality charts, really.



Solutions to the Imbalance Problem

There are several approaches to the problem of the imbalance resulting from heat:

1) Increase heat sink efficiency by a factor of 2. This has relatively good effects (as we can already see with the DSH charts), but it's not perfect.
2) Adjust the heat produced by weapons so they get closer in efficiency.
3) Take the table top values for damage and heat, and divide them by the weapon's new rate of fire. This has the advantage of also making all trial mechs as good (or bad) as they were in the table top - but it will of course lead to keeping all table top imbalances around. I like the apprpoach in general, but it can only be a starting point.

My Solution
I have one approach that uses mostly 2 and combines it a bit with the 3rd option's general goal of trying to get mechs to perform more closely as they used to. That means a special eye was taken to reduce the heat output. Theoretically, one could also raise all weapons damage output, but that would keep trial mechs hot and unviable - lowering heat to buff a weapon and reducing damage to nerf them was a good approach. Not entirely by accident, I didn't alter the Gauss Rifle one bit here - the other weapons in its role (mostly the PPCs) just get competitive with it.


This solution isn't the one true way. ANd it will not be perfect out of the box. But I believe will already lead to a much better balance and more viable builds than the current system.

Spoiler



Other Challenges
Generally, Double Heat Sinks present a major upgrade in Battletech. They are not entirely without drawbacks due to their crit cost, but they are definitely strong, and they primarily benefit energy weapons.

It can be dangerous to nerf them too hard - future mechs will need them just as much as the current mechs need their sinkle heat sinks to be working properly.

I think one of the most intersting possibilites for "nerfing" them is something that can only be done in MW:O, but wouldn't translate meaningful to the turn based game.

In MW:O, the heat capacity is 30 + heat sinks - if it wasn't, there would never be a way to fire, say, 4 PPCs at once - they produced 40 (currently 36) heat together, instantly overheating you. But this worked in the table top, because the effects of heat were calculated at the end of the turn, and the heat sinks worked in one step.

But in MWO, this capacity could be a way to distinguish DHS and SHS.

Let's keep the capacity at 30 + heat sinks - let's not count the double heat sinks double for the capacity. Heck, one could even lower the base capacity of 30 a bit (say, to 20). The result would be that with DHS, you are more reliant on staggering your fire - you don't fire your 4 PPCs all at once, but you can fire them every time they are off cooldown if you've got enough heat sinks. This would create an interesting dynamic, and it would affect tthe TET efficiency notably (but not affect heat neutrality, which is predicated on the idea to never even gain net heat.) and could be exactly what is needed to make DHS less of a must, and keep the Ballistics viable.

Conclusion
There is a lot to be done, but it isn't hopeless, and I think there are good solutions possible.

Edited by MustrumRidcully, 03 November 2012 - 09:50 AM.


#2 AllGamer

    Member

  • PipPipPipPipPipPip
  • Legendary Founder
  • Legendary Founder
  • 331 posts
  • Locationsomewhere in innersphere

Posted 30 October 2012 - 12:51 PM

That's basically saying, don't follow the table top BattleTech "rules/specs"

it's indeed hard to balance a game of this magnitude, without breaking too many of the original table top specifications while maintaining true to the MW Universe

#3 Lyteros

    Member

  • PipPipPipPipPipPip
  • 456 posts
  • LocationGermany

Posted 30 October 2012 - 12:53 PM

Its not about getting TT balance (hat enough imbalance there), its about balancing out the weapons for MWO.
Right now its mathematically proven that high heat weapons are far weaker then ballistics and LRM.
Adressing this issue will also make trialmechs drastically better.

#4 MustrumRidcully

    Member

  • PipPipPipPipPipPipPipPipPipPipPip
  • Legendary Founder
  • Legendary Founder
  • 10,644 posts

Posted 30 October 2012 - 01:01 PM

View PostAllGamer, on 30 October 2012 - 12:51 PM, said:

That's basically saying, don't follow the table top BattleTech "rules/specs"

it's indeed hard to balance a game of this magnitude, without breaking too many of the original table top specifications while maintaining true to the MW Universe

I prefer to think "follow them smartly". ;) One of my approaches to the solution therefor always was to take the table top values as abstraction of the output of a weapon over 10 seconds, and simply divide all those damage and heat values by the rate of fire.

But ultimately, the TT balance is not perfect - that's why they introduced battle values to balance individual games. I would have preferred if they would have simply adjusted some stats... Just rename the AC/2 and AC/5 to AC/4 and AC/6 respectively, and upgrade their damage accordingly, and they may be getting somewhere. Diverging from it is

But I don't see having PGI a working battle value system any time soon. It took FASA a lot of effort to get the first implementation right, and they didn't have to fix their net code and implement a match-making algorithm on a server platform in addition to all the rest.

#5 Felbombling

    Member

  • PipPipPipPipPipPipPipPipPip
  • 2,973 posts
  • LocationVancouver, BC

Posted 30 October 2012 - 02:15 PM

Another way to look at it would be that they went overboard with the firing / recycle rates on some weapons, if they indeed intend to stick to the 10 second heat sink cycle. You could easily reign in the Small Laser, for example, by adding another 2.5 seconds into the cooldown or duration... or a combination of both. There is zero reason why a filler weapon should be the wonder weapon it currently is.

#6 Overlord Klein

    Member

  • PipPipPipPipPip
  • 108 posts

Posted 30 October 2012 - 02:17 PM

Wow, TL:DR, but from the looks of it, nice work with all the charts and stuff!

#7 vifoxe

    Member

  • PipPipPipPipPip
  • 101 posts

Posted 30 October 2012 - 02:33 PM

View PostAllGamer, on 30 October 2012 - 12:51 PM, said:

That's basically saying, don't follow the table top BattleTech "rules/specs"

it's indeed hard to balance a game of this magnitude, without breaking too many of the original table top specifications while maintaining true to the MW Universe



[REDACTED] the universe. balance the damn game.

Edited by miSs, 30 October 2012 - 05:01 PM.
circumventing word filter


#8 Squidhead Jax

    Member

  • PipPipPipPipPipPipPipPip
  • 1,434 posts

Posted 30 October 2012 - 03:04 PM

Outstanding analytics as usual, Archchancellor.

As for the general heat system, flavorwise I prefer a big slice of option 1 with 2 applied as needed, though the graphs suggest it wouldn't be much after that.

And as usual, I am flying the flag of keeping DHS properly frosty but lowering their operational cieling.

#9 Vapor Trail

    Member

  • PipPipPipPipPipPipPipPip
  • 1,287 posts
  • LocationNorfolk VA

Posted 30 October 2012 - 03:04 PM

View Postvifoxe, on 30 October 2012 - 02:33 PM, said:



[REDACTED] the universe. balance the damn game.


Ideally, satisfying both should be the goal. And I believe it's possible to do.

#10 Calmon

    Member

  • PipPipPipPipPipPip
  • 392 posts

Posted 30 October 2012 - 03:59 PM

Really great thread! Thanks for doing this.

#11 Abrahms

    Member

  • PipPipPipPipPipPipPipPip
  • 1,478 posts

Posted 30 October 2012 - 07:10 PM

View PostMustrumRidcully, on 30 October 2012 - 12:36 PM, said:

During the Closed Beta, there were quite a few threads about balance, and in a large part it revolves around the heat system. Those old threads are unfortunately now gone (or "archived" with currently no way of reading them). So I decided to start a new thread devoted to the topic. This post will probably be long and contain charts and all. So for starters, a TL;DR overview.

TL;DR Overview
Balance has yet to be achieved. There are some weapons that are very difficult to use for little benefit, and some that are very easy to use for much benefit.
If you try to apply various metrics to gather the effectiveness or efficiency of weapons, you will realize that low-heat weapons are simply the superior way to achieve a decent damage output.

As a side effect of the current system, trial mechs or rather all stock configurations are particularly weak - they are usually overgunned and undersinked, e.g. they carry a lot of weapons that,if they really wanted to use them to their full ability, wolud need a lot of heat sinks. These mechs do not play in any way or form like they may have in the original system they were designed for.

The best approach to fixing this is to reduce the ratio between heat generation and heat dissipation, for example by reducing heat values, particularly for high heat weapons - don't be afraid - this doesn't mean heat will be unimportant or meaningless and will not require management. It just means that you will have more viable choices in how to build a mech, and you will still need to make major decisions on how hot or cool you want the mech to be, and hace meaningful trade-offs between sustained damage, burst damage and alpha strike damage.

If you don't like the details but like to see the charts, skip the parts up to the pretty pictures.

The Long Version
This is a long version. Pour yourself a coffee or something. It may contain mathematical models and charts based on them.

Going real time with Battletech - PGI Style
The first thing PGI basically did when they implemented MW:Os weapon systems was to take all table top stats - weight, crits, damage, heat. They just made one major addition. The table top game is a turn based board game, it doesn't run on CPU cycles of a few millisesconds, it runs on turns that represent each 10 seconds and take the "human" CPU sometimes minutes to resolve (and this doesn't even consider the decision making process made by the players).
Now, 10 second turns are a nice abstraction, but for a real time game, just taking those table top stats and saying every weapon shoots only every 10 seconds is a bit lame. 10 seconds is just too long to be reasonable.

So PGI gave weapons firing rates. The rates changed over the course of the Beta - most weapons probably have a recycle time of around 4 seconds, but some have notable lower values.

But what PGI when they did this was that they kept the table top stats for damage and heat as a "per shot" value, and still assumed that heat sinks would dissipate 1 heat over 10 seconds (but of course, a bit of heat every time unit).

Unfortunately, what they never seemed to realize was that this approach has a fundemental flaw - it violates a core constraint of the original battletech rules - and that was that each weapon could fire only once per turn (exception: Ultra AUto-Cannons, but those got special rules and adjustments to cover that advantage). Why is that such an important constrant?
Let's compare two extremes - the Gauss Rifle and the PPC.
THe Gauss Rifle weighs 15 tons and produces 15 damage and 1 heat per turn, and it needs some ammo. The PPC weighs 7 tons and produces 10 damage and 10 heat per turn. In the table top, 2 tons of ammo would be generally sufficient, so let's put the Gauss weight at effectively 18 tons - if you spend 18 tons, in weapon, ammo and heat sinks, you can add 15 damage per turn to a mech without changing how fast he overheats. For the PPC, the same value is about 17 tons. Now, we realize already that the Gauss seems like a good deal here, but in the table top game, that was deemed okay, because the Gauss Rifle was on some level meant as an upgrade, a new tech level.
But so, basically, 18 tons add 15 damage to your damage output in 10 seconds, and 17 tons add 10 damage to your damage output in 10 seconds, without changing how quickly he overheats.
But what happens if you allow to fire a weapon twice per turn?
If you already have a Gauss Rifle, all you need to do is add 1 heat sink - and maybe some extra ammo, if you want. So, let's say, 3 tons for 15 damage.
But if you have a PPC, you need to add 10 heat sinks. So 10 tons for 10 damage.
This highlights 2 things:
1) Low heat weapons benefit more than high heat weapns from increased rate of fire.
2) To improve your damage output, it is generally cheaper to add more heat sinks and ammo to utilize your maximum rate of fire, than adding a full new weapon system.

This means that original balancing assumptions between wepaons are no longer true, and that original assumptions on mech builds are no longer true. You really don't need to equip a mech with 3 PPCs and 28 heat sinks - 2 PPCs and 35 heat sinks will deliver you more damage if you consider the sustainability.

This is the core flaw of the MWO weapon system, in a nut shell.
There is of course more. For example - the heat system lacks the details of the Table Top heat scale - there is just the big "endpoint" - shutdown. But all the points in between, where you lost speed and suffered hit penalties - there is nothing trying to represent this. Which in turn means that while heat can overall be very punishing, alpha strikes itself are not that risky at all. But that's just a secondary concern, really.

Where did the mechanics lead us to?
But there are ways to also illustrate this, which I and others went a lot of effort to illustrate.
Before I go into this, just one remark:

This isn't just theorycrafting. In truth, these theories are mostly predated by what we actually see happening during gameplay.

We saw many mechs utilizing the fact that low heat weapons can utilize their ROF much better. It leads us to Jenners with 6 small lasers and the "Slunchback"; A HBK-4P refitted witht the biggest XL Engine and replacing all 8 medium lasers with small lasers, and any spare weight spend on heat sinks.
We see people taking the Catapult K2, ripping out the useless PPCs and equipping 2 Gauss Rifles, because the Gauss Rifle was just outperforming the PPC so much better that all the sacrifices in armor or engine weight are worth it. It just makes a giant difference whether you just have the heat sinks to support two PPCs firing every 10 seconds or the heat sinks to support 2 Gauss Rifles firing every 4 seconds.

The Run Hot or Die Tournament also lead to very similar mech configuraitons - once the game is played competitively and people start optimzing


Illustrating the Balance with Math and Charts
There is one general approach to measure weapon strength - this is generally trying to figure out "how much does it cost me to run this weapon in weight" vs "how much damage does it let me deal this way. Theoretically, one could also use such a system for crit requirements, but this is more rarely done. Though with Inner Sphere heat sinks, it may be an interesting addition.

There are two major ways to figure out the weight, and mostly it is the question - how do we calculate the number of heat sinks for a weapon?

1) Heat Neutrality
We try to put in enough heat sinks so that the weapons heat is perfectly negated.
This is a simple, useful approach, though some thing it's too simple.
Generally, heat neutrality is very useful to know if you are working at modifiying your mech and adding or switching out some weapons. If you have already found a mech configuration that you think is "reasonably hot" or "reasonably cool" (ranging from overcooled to heat neutral to gaining heat slowly up to overheats in the blink of an eye), when you made free some weight, you can simply add the "heat neutral" configuration to the mech and nothing will change about its time to last.
If you replace weapons, you may realize that for the weight of 2 Medium Lasers and their heat sinks, you could also, say, fit another Large Laser - maybe that would give you something you found lacking so far.

2) TET - Targeted Engagement Time
This approach is a bit different from heat neutrality is that you try to figure out - how long do I expect an individual combat/skirmish/brawl (e.g. not the entire match, but one fight in that match) too last, and then find a build that optimizes the damage output in this time frame.
This approach has some very practical applications to a mech or weapon configurations. For example, if you have 2 long range weapons, you may decide that these are mnostly used for sniping, where engagement times are usually low (unless the enemy is not trying to get into cover), but also a few short range weapons which you intend to use in a brawl. So you may try to go for a high TET for the brawling weapons and a low TET for the long range weapons. As long as you don'T use them together, you should get what you want. (Whether it's always the best choice to mix roles like this is another question).

The important aspect of the TET approach is that it has to account for a mech's inherent heat capacity - since that allows you to gain heat without needing to immediately/completely dissipate it. This creates some artefacts - if you really always only calculate the TET for a single weapon, you get unrealistic figures - a single medium laser currently can produce 10 heat over 10 seconds on average, so that the time to overheat is at least 30 (in practice, it's even higher - your heat capacity in MW:O is 30 + heat sinks, not just 30 - and you need to have 10 heat sinks now... So basically, with a single medium laser and 10 sinks, you never overheat.)

So for my calculations, I always used 4 weapons.That can be unrealistic for large weapons, but it's close enough - the real error beings at the low heat levels, not the high ones.

How to Calculate Damage
THere are basically two major damage figures people are interested in:
- Damage Per Shot (or "burst" or in Battletech terms "alpha" damage)
- Damage over Time (or "DPS")
Damage per Shot is an important asset to have - you want to nail weak spots at the enemy mech and a high damage per shot means you can destroy them without firing again.
Damage over Time - You can't one-shot everything and everyone, so you will need to take multiple shots - how much damage over time you can inflict on average gives a hint on how long it will take you.

How to put Damage and Weight together
I have basically chosen 3 approaches in total.

For using heat neutrality, I have used two damage figures.
1) Damage Over Time
2) Damage Over Time + Damage per Shot, weighted against each other. (More precisely, I calculated it as (Damage per SHot + 10 x Damage over Time)/10.

For TET, I just used the damage that a mech would inflict continously firing.

Whatever model I used, I then used the calculated weight against the calculate damage.

What we would like to see when we compare damage efficiency stats of weapons

Generally:
Range has not appeared yet in any of the calculations, but is a major factor in balancing weapons. What we want to see ideally is that weapons with similar range have similar damage efficiency stats, and the longer the range, the lower the damage efficiency. This means there is a meaningful trade-off between damage and range. If you use a short range weapon, you must get close to the enemy first, but if you do, you can outdamage the enemy that relies on long range weapons. This introduces also an interesting tactical element, as each participant will try to fight in his "turf" or range category.

Heat Neutrality + Weighted Damage Values
Here, Range should suffice. We already made burst potential an advantage in the efficiency chart.

Heat Neutrality + Damage Over Time:
Here, Range is one factor, but we don't account for burst yet. As a general guideline - the higher the damage per shot is for a weapon, the lower we want the efficiency to be.

TET
For TET, we have similar concerns as for Heat Neutrality and Damage Over Time - more range and more burst should mean a lower efficiency.

Weapon Efficiency based on the current stats (as of 30th of Octobre 2012)

Heat Neutrality
This chart contains 2 curves - one based on damage over time (DPS/Weight) and one on the weighted values for damage per shot and damage over time (DPS_ID/Weight)

Posted Image

For double heat sinks:
Spoiler


As we can see - the ballistics still outperform the energy weapons mostly, and the general rule - longer range, lower efficiency - is not observed at all for them.
Energy weapons follow the trend in general, though Small Lasers and Medium Lasers still present notable spikes.

SRMs and LRMs seem to play in an entirely different league.

TET Charts
With single heat sinks.
Posted Image

With Double Heat Sinks
Spoiler


The results are similar to the heat neutrality charts, really.



Solutions to the Imbalance Problem

There are several approaches to the problem of the imbalance resulting from heat:

1) Increase heat sink efficiency by a factor of 2. This has relatively good effects (as we can already see with the DSH charts), but it's not perfect.
2) Adjust the heat produced by weapons so they get closer in efficiency.
3) Take the table top values for damage and heat, and divide them by the weapon's new rate of fire. This has the advantage of also making all trial mechs as good (or bad) as they were in the table top - but it will of course lead to keeping all table top imbalances around. I like the apprpoach in general, but it can only be a starting point.

My Solution
I have one approach that uses mostly 2 and combines it a bit with the 3rd option's general goal of trying to get mechs to perform more closely as they used to. That means a special eye was taken to reduce the heat output. Theoretically, one could also raise all weapons damage output, but that would keep trial mechs hot and unviable - lowering heat to buff a weapon and reducing damage to nerf them was a good approach. Not entirely by accident, I didn't alter the Gauss Rifle one bit here - the other weapons in its role (mostly the PPCs) just get competitive with it.


This solution isn't the one true way. ANd it will not be perfect out of the box. But I believe will already lead to a much better balance and more viable builds than the current system.

Spoiler



Other Challenges
Generally, Double Heat Sinks present a major upgrade in Battletech. They are not entirely without drawbacks due to their crit cost, but they are definitely strong, and they primarily benefit energy weapons.

It can be dangerous to nerf them too hard - future mechs will need them just as much as the current mechs need their sinkle heat sinks to be working properly.

I think one of the most intersting possibilites for "nerfing" them is something that can only be done in MW:O, but wouldn't translate meaningful to the turn based game.

In MW:O, the heat capacity is 30 + heat sinks - if it wasn't, there would never be a way to fire, say, 4 PPCs at once - they produced 40 (currently 36) heat together, instantly overheating you. But this worked in the table top, because the effects of heat were calculated at the end of the turn, and the heat sinks worked in one step.

But in MWO, this capacity could be a way to distinguish DHS and SHS.

Let's keep the capacity at 30 + heat sinks - let's not count the double heat sinks double for the capacity. Heck, one could even lower the base capacity of 30 a bit (say, to 20). The result would be that with DHS, you are more reliant on staggering your fire - you don't fire your 4 PPCs all at once, but you can fire them every time they are off cooldown if you've got enough heat sinks. This would create an interesting dynamic, and it would affect tthe TET efficiency notably (but not affect heat neutrality, which is predicated on the idea to never even gain net heat.) and could be exactly what is needed to make DHS less of a must, and keep the Ballistics viable.

Conclusion
There is a lot to be done, but it isn't hopeless, and I think there are good solutions possible.


basically

#12 MustrumRidcully

    Member

  • PipPipPipPipPipPipPipPipPipPipPip
  • Legendary Founder
  • Legendary Founder
  • 10,644 posts

Posted 30 October 2012 - 11:53 PM

View PostAbrahms, on 30 October 2012 - 07:10 PM, said:


basically


Now if I could just find a way to put all this information into people's head without having them read so much...

#13 RAM

    Member

  • PipPipPipPipPipPipPipPipPip
  • The Resolute
  • The Resolute
  • 2,018 posts
  • Google+: Link
  • LocationVancouver, BC

Posted 31 October 2012 - 12:24 AM

View PostMustrumRidcully, on 30 October 2012 - 12:36 PM, said:

So PGI gave weapons firing rates. The rates changed over the course of the Beta - most weapons probably have a recycle time of around 4 seconds, but some have notable lower values.
Unfortunately, what they never seemed to realize was that this approach has a fundemental flaw - it violates a core constraint of the original battletech rules - and that was that each weapon could fire only once per turn (exception: Ultra AUto-Cannons, but those got special rules and adjustments to cover that advantage).

False – Battletech has always had rate of fire. PGI did not ‘invent’ anything.


RAM
ELH

#14 Asatruer

    Member

  • PipPipPipPipPipPip
  • Knight Errant
  • 235 posts
  • LocationSeattle

Posted 31 October 2012 - 12:25 AM

View PostMustrumRidcully, on 30 October 2012 - 11:53 PM, said:

Now if I could just find a way to put all this information into people's head without having them read so much...

It really should not be so hard, as it boils down to a fairly simple problem, but the best solution is not quiet as simple as the problem. It would be if option 1 was the best fix, but it is only the easiest, while some hybridization of 2 and 3 would be the best way to go. Even then I would really like to see a less binary overheating or not reason to manage heat, such as TTs gradual increasing severity of heat related issues. That is probably too much of a **** dream though, as implementing a good system that is good at relating to the player what ill effect they are currently suffering would be pretty hard.

In any case, getting the heat/weight/damage ratios sorted out will go a long way to achieve a better semblance of balance, but there will still likely be remaining issues. Such as PPCs still not completely matching a gauss due to firing and projectile flight characteristics, but that is OK. Though it will be interesting to see what the alluded to buff of disrupting targeting will do.

#15 Asatruer

    Member

  • PipPipPipPipPipPip
  • Knight Errant
  • 235 posts
  • LocationSeattle

Posted 31 October 2012 - 12:29 AM

View PostRAM, on 31 October 2012 - 12:24 AM, said:

False – Battletech has always had rate of fire. PGI did not ‘invent’ anything.


RAM
ELH
You have made that claim before, but were not able to back it up then either. Please, demonstrate where in core BattleTech rules can a weapon other than an UAC be fired more than once per turn. If you can, I would love to be able to cite it and use it next time I play BattleTech.

Though I know for a fact that last time I played back in the early 90s, no such thing was possible, thus easily disproving your, "always had" claim.

Edited by Asatruer, 31 October 2012 - 12:31 AM.


#16 MustrumRidcully

    Member

  • PipPipPipPipPipPipPipPipPipPipPip
  • Legendary Founder
  • Legendary Founder
  • 10,644 posts

Posted 31 October 2012 - 01:28 AM

View PostAsatruer, on 31 October 2012 - 12:29 AM, said:

You have made that claim before, but were not able to back it up then either. Please, demonstrate where in core BattleTech rules can a weapon other than an UAC be fired more than once per turn. If you can, I would love to be able to cite it and use it next time I play BattleTech.

Though I know for a fact that last time I played back in the early 90s, no such thing was possible, thus easily disproving your, "always had" claim.

I don't want to offend RAM, because he's dedicated to the game and Battletech in general and supportive for the community, but (yes, this is me doing it anyway), I think I have to resort to simply ignore his comments on the topic of game balance and the heat system. Positively said, he thinks extremely different about these topics than I do, and I can't seem to find any other common ground other than "Battletech is fun and so are Mechwarrior games" - which is not sufficicent to discuss about the heat system or weapon balance.

Quote

It really should not be so hard, as it boils down to a fairly simple problem, but the best solution is not quiet as simple as the problem. It would be if option 1 was the best fix, but it is only the easiest, while some hybridization of 2 and 3 would be the best way to go. Even then I would really like to see a less binary overheating or not reason to manage heat, such as TTs gradual increasing severity of heat related issues. That is probably too much of a **** dream though, as implementing a good system that is good at relating to the player what ill effect they are currently suffering would be pretty hard.


It's always dangerous to say things like "But that would be easy to implement", so I will just say: I think it's immenintly feasible to implement a more complex heat scale, and it may just be a question when they get around to do it - provided the will is there. I think it would do the game good and enrich the general gameplay experience considerably.

In fact, if I had a choice, I think a more complex heat scale is far more interesting and engaging then the crit system. I coul do just fine without the crit system and its RNGs stuff and all that. A complex heat scale means there are things to manage in game, that there are more trade-offs to discuss with every shot token. I think that's more valuable for the gameplay experience.

Edited by MustrumRidcully, 31 October 2012 - 01:30 AM.


#17 Indoorsman

    Member

  • PipPipPipPipPipPipPip
  • Elite Founder
  • Elite Founder
  • 792 posts
  • LocationTexas

Posted 31 October 2012 - 02:10 AM

View PostMustrumRidcully, on 30 October 2012 - 12:36 PM, said:

Solutions to the Imbalance Problem

There are several approaches to the problem of the imbalance resulting from heat:

1) Increase heat sink efficiency by a factor of 2. This has relatively good effects (as we can already see with the DSH charts), but it's not perfect.
2) Adjust the heat produced by weapons so they get closer in efficiency.
3) Take the table top values for damage and heat, and divide them by the weapon's new rate of fire. This has the advantage of also making all trial mechs as good (or bad) as they were in the table top - but it will of course lead to keeping all table top imbalances around. I like the apprpoach in general, but it can only be a starting point.


1, bad
2, good
3, bad

Option 1 should be increase heat sink efficiency ~10% at a time, w/option 2 already having been done until the highest heat weapons become useable. Option 3 is just a bad idea :-p

#18 wuselfuzz

    Member

  • PipPipPipPipPip
  • 143 posts

Posted 31 October 2012 - 03:15 AM

I fully support Mustrum's Ideas, since IMO they're very well thought and backed with proper maths. IMO, they'd be the perfect combination of keeping the feel of the original BattleTech weapons and making them work in a real time simulation.

I'd also like to express my dislike for those posts stating that using the TT values for orientation is a bad thing "just because this isn't TT" while not stating any other alternative.

#19 Vapor Trail

    Member

  • PipPipPipPipPipPipPipPip
  • 1,287 posts
  • LocationNorfolk VA

Posted 31 October 2012 - 03:27 AM

View PostAsatruer, on 31 October 2012 - 12:29 AM, said:

You have made that claim before, but were not able to back it up then either. Please, demonstrate where in core BattleTech rules can a weapon other than an UAC be fired more than once per turn. If you can, I would love to be able to cite it and use it next time I play BattleTech.

Though I know for a fact that last time I played back in the early 90s, no such thing was possible, thus easily disproving your, "always had" claim.



Um. Just to be precise. A rate of fire is how many shots per second a weapon has.

Number of shots per unit time.

One shot per round.
That is a rate of fire. So TT had a rate of fire. It just happened to have the same rate of fire for every weapon. That of one shot every ten seconds.

All that's happened for MWO is the rate of fire for each weapon has been unlinked from the others, and increased.

A PPC got a 3.333 shots / 10 seconds, (or 1 shot every 3 seconds) rate of fire.
A Gauss got a 2.5 shots / 10 seconds (or 1 shot every 4 seconds) rate of fire.
And so on.

#20 Draco Argentum

    Member

  • PipPipPipPipPipPipPipPip
  • 1,222 posts

Posted 31 October 2012 - 03:37 AM

As we've been discussing in the beta forum you're exactly right.

Theres a few more principles that should be noted for weapon balance. You mentioned DPS range and alpha.

Unweildliness is another one, having a huge RoF sounds great but it is really unweildly. You must be fireing on cooldown to use the DPS. This means you can't be torso twisting to spread damage. You can't have the enemy walk around a corner either. In a sense this is the opposite of alpha. But not completely, these factors become non linear at high rofs. Below three seconds it really kicks in.

At least two differences, IMO this is critical. Any give pair of weapons should at least two differences. Range, DPS, alpha are the big ones, weaker ones worth half each: contiguous crits, burn duration, cluster shells, projectile speed. By having two or more differences its harder to say that one weapon is definately better than the other.


I'd post my attempt but excel charting functions kicked my butt. I'm still wrestling with it.





1 user(s) are reading this topic

0 members, 1 guests, 0 anonymous users