Jump to content

Open Letter To Pgi: Why You're Having Such Trouble Balancing Mwo


721 replies to this topic

Poll: Open Letter To Pgi: Why You're Having Such Trouble Balancing Mwo (285 member(s) have cast votes)

Do you think the discussed features should be added to the test server after 12v12 is in the live game?

  1. Yes, yes, a thousand times yes! (235 votes [82.46%] - View)

    Percentage of vote: 82.46%

  2. Nah, I agree with Paul, the game is great as is. (26 votes [9.12%] - View)

    Percentage of vote: 9.12%

  3. I don't really care. (24 votes [8.42%] - View)

    Percentage of vote: 8.42%

Vote Guests cannot vote

#381 Aullido

    Member

  • PipPip
  • 32 posts

Posted 11 July 2013 - 07:32 PM

I play hardcore WWII aircraft simulators. On those games we have cone of fire but we don't mind because that actually help us to hit. What really matters is convergence. Our weapons are fixed and you must try to shoot at the right range for maximum effect. If you are very close, you aim with only one wing, knowing that half your fire power will go wasted, but otherwise you won't hit at all. Any shoot outside convergence must be compensated.

Now picture the situation were at close range fighting a Wounded Mech, but not all you weapons will be equally effective, even same weapons due their allocation. With fixed convergence you must decide which weapon will hit a location (No cone of fire, not randomness) but not all your weapons can hit right there or even at all. Grouping will matter a lot, because you won't wasted ammo or heat and a mix of short and long range weapons is better. You will need to thing before you shoot as Mechwarrior is intended to be.

Would be necessary further adjust damage, cycle time and armor to balance the game. The important thing is we get ride off boating.

#382 Angel of Annihilation

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • The Infernal
  • The Infernal
  • 8,881 posts

Posted 11 July 2013 - 08:16 PM

The simple answer. Because the players realy have no clue about balance so PGI will never be able to statisfy you all (or me likely).

#383 Boyinleaves

    Member

  • PipPipPip
  • 86 posts

Posted 11 July 2013 - 09:20 PM

Great post DarkJaguar, some other fantastic points throughout this thread too.

I've only been playing this game for a couple of months, but I can stand behind the goals of the OP 100%.

While I can't speak as to the correct values or implementation of any specific numbers or system, be it TT inspired or something devised by PGI, I can certainly appreciate that transferring a turn based system to real time does create its own series of problems. I hope that most players want something that is in the spirit of MechWarrior in preference of having TT numbers purely for nostalgia value, but just from the points made about the 10 second turns, I think it would improve the game greatly to have the heaviest, most powerful single shot weapons have longer recycle times. For example, a 10 second recycle time for PPCs, AC/20s and Gauss rifles would strengthen the differences between weight classes, as taking a shot with one of these weapons at a light travelling at full throttle would be a risk, and if you missed, the light mech would escape unscathed.

Anything that improves the differences between weight classes and makes each weight class viable and have a distinct role to play would be a positive change in my book.

To all those who are declaring that ACs must be single fire weapons: this is not the standard meaning of AC either in RL definition, or in BT; although in BT lore, there are some ACs that are single fire weapons (though these are usually too high a caliber to mount on a 'Mech), most are multi-shot, burst fire weapons, and the single roll to hit in the TT rules is an abstraction that serves the purpose of making the game work efficiently. That said, a single fire AC/20 could certainly have a place in MW:O, but I would imagine that such a weapon needs to either have a long recycle time, cause a great deal of recoil, or have a slight cone of fire to balance it out (or a combination of these factors). The other variants of ACs that are available in BT canon just need to be available, and have their own strengths and weaknesses.

The beauty of the customisation options in BT is that you can set up your 'Mechs for a variety of roles, or even to fill more than one role in a less effective manner. What will make MW:O better in my humble opinion, is offering more options for specific roles to be useful in the game.

#384 Aim64C

    Member

  • PipPipPipPipPipPipPip
  • 967 posts

Posted 11 July 2013 - 10:22 PM

I can agree with key portions of this... but at the same time - there are differences in the tabletop values.

Lasers don't apply all of their damage, for example.

Further, tabletop allows you to choose which weapons to shoot at your leisure at a target that comes into range/view within your turn. If a mech tries to dodge between buildings in the city - you can walk a block and dump your weapons into it as if it were standing still.

The fact is that combat, in general, is a fast and chaotic thing. The 'thinking' part of it is long before the shooting starts, and most of what happens when the bullets start to fly is training and fragmented plans stitched together through half-functioning communication. You use cover and concealment to move.

There are certainly balance issues with the game - but at the same time, a lot of one-on-one mech engagements shouldn't last more than about 20 seconds before one or the other is dead or has successfully broken contact. Within the first five seconds of the encounter, you should be committing to the engagement or breaking contact. Within ten you should be moving, and within fifteen you should be seeing results and completed by twenty.

The weapons and armor can be balanced around that. The difficulty is balancing weapons with respect to each other so that different weapons convey different functions and advantages/disadvantages. Heat doesn't translate to a real-time environment the same way it does in tabletop. You can fire your weapons with a "spike" in heat, and move normally your next turn (as if nothing happened). Translating heat effects to a real-time environment straight from tabletop, literally, causes problems, as things between turns don't really affect your mech (aside from damage taken). However, in a real-time environment, those things have immediate effect.

Further - something like an AC20 is 'balanced' the way it is because it deals 20 damage to one location. That weapon, in tabletop, alone, removes over a ton of armor from one location in one turn.

Spreading that out over ten second averages radically changes the practical utility of the weapon. It's nice to believe that spreadsheets can lead to balance - but that's like saying the Large Pulse Laser is competitive with the current PPC. Even the 'ease of use' of the hitscan elements of the pulse laser make it generally inferior and less desirable than the PPC. That's because the laser distributes its damage over time while the PPC delivers its damage up-front.

Introducing different brands/models of autocannon classes would be nice - but that would also require each one to be individually balanced with very different stats - because if they all use the same overall DOT averages because they fall in the same class - only a couple will really be considered competitive weapons.

I would like to see different 'brands' or 'models' of weapons - but the reality is that if you start messing with damage, heat, duration, DOT, etc - you're necessarily going to have to start changing weight, criticals, ranges, etc to really be able to make the weapons competitive options against each other. Especially with ballistics - where altering how damage is applied puts them in heavy competition with much lighter and more manageable laser weaponry.

And we've already seen what a mess all that individual balancing creates.

#385 AndyHill

    Member

  • PipPipPipPipPipPip
  • 396 posts

Posted 12 July 2013 - 05:06 AM

View PostMister Blastman, on 10 July 2013 - 07:37 AM, said:


Putting a system of convergence into MWO is NOT complex. Making it based off of heat, or giving us true heat penalties is NOT complex.

I present to you complex:

Falcon

Please, don't EVER use the word complex to describe MWO again or anything they could put into the game. For the record I fly both of the above sims frequently and the complexity become second nature after a while.


I had to comment to this, because I also like to fly simulators that tend to be relatively complex by nature. I don't even disagree with you in general, but I'd like to point out that complexity in itself is usually a clearly bad thing. Modern jet fighters try to be as simple to use as possible, they just operate in such difficult environments and modern air combat requires the use of many different sensors and weapons, which inevitably leads to a high level of complexity. Most importantly, in simulator games the complexity comes with associated rewards (which is realism, weapon employment effectiveness and interesting and rewarding gameplay to name a few). I would say that's positive complexity in many ways.

An opposite example would be the shutdown system in MWO. You have to press 'o' before shutdowwn to override it for a given time as opposed to simply toggling overheat possibility on or off before a match (how it would be done in a Viper, think about throttle governor for example). Furthermore - and even worse - the 'p' key changes function arbitrarily depending on which state of the process you are in. If you for example press 'p' just as the 'mech starts to power up again, it shuts down. That's simply poor UI design and unnecessary complexity with no rewards included whatsoever.

TLDR: complexity can be a good or bad thing in game design, depending on the associated positive and negative sides.

#386 MustrumRidcully

    Member

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

Posted 12 July 2013 - 06:04 AM

View PostAim64C, on 11 July 2013 - 10:22 PM, said:

I can agree with key portions of this... but at the same time - there are differences in the tabletop values.

Lasers don't apply all of their damage, for example.

Further, tabletop allows you to choose which weapons to shoot at your leisure at a target that comes into range/view within your turn. If a mech tries to dodge between buildings in the city - you can walk a block and dump your weapons into it as if it were standing still.

The fact is that combat, in general, is a fast and chaotic thing. The 'thinking' part of it is long before the shooting starts, and most of what happens when the bullets start to fly is training and fragmented plans stitched together through half-functioning communication. You use cover and concealment to move.

There are certainly balance issues with the game - but at the same time, a lot of one-on-one mech engagements shouldn't last more than about 20 seconds before one or the other is dead or has successfully broken contact. Within the first five seconds of the encounter, you should be committing to the engagement or breaking contact. Within ten you should be moving, and within fifteen you should be seeing results and completed by twenty.

The weapons and armor can be balanced around that. The difficulty is balancing weapons with respect to each other so that different weapons convey different functions and advantages/disadvantages. Heat doesn't translate to a real-time environment the same way it does in tabletop. You can fire your weapons with a "spike" in heat, and move normally your next turn (as if nothing happened). Translating heat effects to a real-time environment straight from tabletop, literally, causes problems, as things between turns don't really affect your mech (aside from damage taken). However, in a real-time environment, those things have immediate effect.

Further - something like an AC20 is 'balanced' the way it is because it deals 20 damage to one location. That weapon, in tabletop, alone, removes over a ton of armor from one location in one turn.

Spreading that out over ten second averages radically changes the practical utility of the weapon. It's nice to believe that spreadsheets can lead to balance - but that's like saying the Large Pulse Laser is competitive with the current PPC. Even the 'ease of use' of the hitscan elements of the pulse laser make it generally inferior and less desirable than the PPC. That's because the laser distributes its damage over time while the PPC delivers its damage up-front.

Introducing different brands/models of autocannon classes would be nice - but that would also require each one to be individually balanced with very different stats - because if they all use the same overall DOT averages because they fall in the same class - only a couple will really be considered competitive weapons.

I would like to see different 'brands' or 'models' of weapons - but the reality is that if you start messing with damage, heat, duration, DOT, etc - you're necessarily going to have to start changing weight, criticals, ranges, etc to really be able to make the weapons competitive options against each other. Especially with ballistics - where altering how damage is applied puts them in heavy competition with much lighter and more manageable laser weaponry.

And we've already seen what a mess all that individual balancing creates.

I think of values like rate of fire, damage per hit, heat per shot, weight, criticals and range all as dials. If you for some reason forced to make a 7 ton weapons that costs 3 crits and shoots blue light in some form, you will have to tweak damage, heat and rate of fire.

I think you can use "spreadsheets" for balance - but you can't just make up a spreadsheet and insert a few made up numbers and believe it works. You have to actually figure out the relationships between certain numbers. How valuable is range? You could naively say that a weapon with twice the range is twice as powerful as an otherwise identical weapon, but is that true at all times? (If you think at a typical map, mouse precision, resolution, you will probably realize that 1,500m range and 3,000m range make no big difference, since you simply usually will not be able to hit at that distance.)

The "engagement window" of 20 seconds I think is a pretty big aspect of "balancing by spreadsheet". You must actually know how long an engagement will last. With those AC/40 or Quad PPC, it can turn out that the window is only 12 seconds in some cases. And if you generally talk about a sniper role, you may find that an engagement lasts only 5 seconds (but doesn't end with either side being destroyed).

I made such "spreadsheets" extensively in the past. I've given up because who cares, certainly not PGI, but I think I can show the points where PPCs have an advantage in pure damage output. I can't, however, show you the advantage of pinpoint precision. Maybe if I set my mind on that, I could even figure that one out - As a starting point, you could do a statistical analysis on weapon stats. How many shots were fired for each weapon, how much damage should that have inflicted, and how much did it actually inflict (since laser accuracy is a useless stat, if it wasn't, I'd just take that)? This doesn't account for hits that missed the desired location but still hit the target, but I'd be so bold to claim that this is probably a proportional relationship. But I only access to my weapon stats, and even if I had access to more, all that analysis would be pointless, because PGI implements arbitrary heat penalties for weapon groups instead of engaging in an actual discussion and listening to advice.

#387 StandingCow

    Member

  • PipPipPipPipPipPipPipPipPip
  • Legendary Founder
  • Legendary Founder
  • 3,069 posts
  • LocationUSA

Posted 12 July 2013 - 06:05 AM

Well, I read half the thread (and the OP). No matter what iteration of MW you thought was best, no matter what you think about TT values vs current values... right now, the game balance isn't in a great place.

PGI has a test server, why not just for the hell of it throw on TT values and recycle times and see what happens?

#388 jeffsw6

    Member

  • PipPipPipPipPipPipPipPip
  • 1,258 posts
  • LocationLouisville, KY (suburbs)

Posted 12 July 2013 - 06:23 AM

View PostAndyHill, on 12 July 2013 - 05:06 AM, said:

TLDR: complexity can be a good or bad thing in game design, depending on the associated positive and negative sides.

Complexity is bad when the person in charge of balance can't understand why MGs aren't useful.

View PostStandingCow, on 12 July 2013 - 06:05 AM, said:

Well, I read half the thread (and the OP). No matter what iteration of MW you thought was best, no matter what you think about TT values vs current values... right now, the game balance isn't in a great place.

PGI has a test server, why not just for the hell of it throw on TT values and recycle times and see what happens?

I'm not opposed to crazy things on the test server, but they aren't even using it for balance. 4 hours a week of test server runs is bug-hunting for 12v12, period.

Also, remember, all you "go back to TT values" people are freaking crazy, because in TT, you cannot aim. If testing it is what it takes to get all of you to stop claiming TT values can work in an FPS, fine! :)

#389 StandingCow

    Member

  • PipPipPipPipPipPipPipPipPip
  • Legendary Founder
  • Legendary Founder
  • 3,069 posts
  • LocationUSA

Posted 12 July 2013 - 06:42 AM

View Postjeffsw6, on 12 July 2013 - 06:23 AM, said:

Complexity is bad when the person in charge of balance can't understand why MGs aren't useful.


I'm not opposed to crazy things on the test server, but they aren't even using it for balance. 4 hours a week of test server runs is bug-hunting for 12v12, period.

Also, remember, all you "go back to TT values" people are freaking crazy, because in TT, you cannot aim. If testing it is what it takes to get all of you to stop claiming TT values can work in an FPS, fine! B)


That's all I am asking for... try it out, shut us up. And not the half assed friends and family TT values, everything. If it works, awesome, if not, people can shut up about it, lol.

#390 DarkJaguar

    Member

  • PipPipPipPipPipPip
  • Big Brother
  • 331 posts

Posted 12 July 2013 - 07:51 AM

View Postjeffsw6, on 12 July 2013 - 06:23 AM, said:

Complexity is bad when the person in charge of balance can't understand why MGs aren't useful.


I'm not opposed to crazy things on the test server, but they aren't even using it for balance. 4 hours a week of test server runs is bug-hunting for 12v12, period.

Also, remember, all you "go back to TT values" people are freaking crazy, because in TT, you cannot aim. If testing it is what it takes to get all of you to stop claiming TT values can work in an FPS, fine! B)


I should start a petition to use this system on the test server. I bet I could prove you wrong about TT values. :blink:

#391 jeffsw6

    Member

  • PipPipPipPipPipPipPipPip
  • 1,258 posts
  • LocationLouisville, KY (suburbs)

Posted 12 July 2013 - 08:13 AM

View PostDarkJaguar, on 12 July 2013 - 07:51 AM, said:

I should start a petition to use this system on the test server. I bet I could prove you wrong about TT values. B)

I'll be the first one to sign your petition. Somehow I do not think the 6 PPC Stalker will be any less deadly without MW:O's doubled armor values. It could almost one-shot an Atlas. Probably could from the back!

#392 stjobe

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • Legendary Founder
  • Legendary Founder
  • 9,498 posts
  • LocationOn your six, chipping away at your rear armour.

Posted 12 July 2013 - 09:26 AM

View Postjeffsw6, on 12 July 2013 - 08:13 AM, said:

I'll be the first one to sign your petition. Somehow I do not think the 6 PPC Stalker will be any less deadly without MW:O's doubled armor values. It could almost one-shot an Atlas. Probably could from the back!

The 6 PPC Stalker would also generate 60 heat if it fired all those PPCs at once, in a system where you auto-shutdown on 30 heat. To avoid an auto-shutdown on an alpha, it'd need 31 SHS (31 tons) or 16 DHS (48 crit slots) - and it could still shut down on a roll of 2-9 on 2d6. Furthermore, the pilot would take damage on a roll of 2-8 on 2d6, and if it foolishly had any ammo, it had a good chance (2-7 on 2d6) to cook that off as well. And if it by some chance avoided all that, it couldn't move at all (-5 MP) and it probably couldn't hit anything (+4 to hit).

So yes. It would be very much less deadly, and that doesn't even take into account that he couldn't aim all his 60 damage into one pixel, or that it would likely be a bit harder to hit anything at all.

Edited by stjobe, 12 July 2013 - 09:28 AM.


#393 DarkJaguar

    Member

  • PipPipPipPipPipPip
  • Big Brother
  • 331 posts

Posted 12 July 2013 - 10:55 AM

Okay, so I'd like to talk about Engagement profiles and damage windows a bit. In my OP I mentioned that time was a balance factor that is much neglected. Time can be used to further differentiate two weapons that otherwise have the same DPS values. In a bit, after this topic has had some discussion, I'll edit the OP to add it in, below are three charts illustrating Ballistic, Energy, and Missile weapon damage profiles, as well as a PDF with every weapon individually.
Posted Image
Posted Image
Posted Image

PDF of individual weapons

By using different damage profiles, otherwise similar weapons can be VASTLY different, let's take the AC10vPPC for example...
Posted Image
In an engagement less than 5 seconds, the PPC has a slight edge, whereas the AC10 will deal 10 damage more quickly than a PPC over a 10 second span, and for less heat, but it also does it in a 1 second burst. There's advantages and disadvantages to both (AC10s can score a partial hit, PPC is hit or miss), so you end up with two VERY different weapons that have the exact same DPS.

I'll also take requests if anyone would like to see any other side by side comparisons of weapons, I have these charts saved in a way that's very easy for me to do such.

#394 IdjitKicker

    Member

  • Pip
  • Legendary Founder
  • Legendary Founder
  • 11 posts

Posted 12 July 2013 - 11:29 AM

Very Cool DJ !!!

#395 DarkJaguar

    Member

  • PipPipPipPipPipPip
  • Big Brother
  • 331 posts

Posted 12 July 2013 - 11:33 AM

Okay, I added a poll to the OP about having these numbers added to the test server once 12v12 is done. Make your voice heard people! ;)

#396 jeffsw6

    Member

  • PipPipPipPipPipPipPipPip
  • 1,258 posts
  • LocationLouisville, KY (suburbs)

Posted 12 July 2013 - 11:41 AM

View Poststjobe, on 12 July 2013 - 09:26 AM, said:

...and that doesn't even take into account that he couldn't aim all his 60 damage into one pixel, or that it would likely be a bit harder to hit anything at all.

Why wouldn't it be able to aim all its damage onto one pixel?

Are you admitting that TT values can't work with aiming?

#397 DarkJaguar

    Member

  • PipPipPipPipPipPip
  • Big Brother
  • 331 posts

Posted 12 July 2013 - 11:45 AM

View Postjeffsw6, on 12 July 2013 - 11:41 AM, said:

Why wouldn't it be able to aim all its damage onto one pixel?

Are you admitting that TT values can't work with aiming?


No numbers can work with pinpoint aiming, because pinpoint aiming makes high alpha builds the only viable choice.

With realistic aiming, high alpha builds aren't nearly as potent, and there's a better balance between DPS and Alpha.

#398 Orzorn

    Member

  • PipPipPipPipPipPipPipPipPip
  • 3,327 posts
  • LocationComanche, Texas

Posted 12 July 2013 - 11:46 AM

View Postjeffsw6, on 12 July 2013 - 11:41 AM, said:

Are you admitting that TT values can't work with aiming?

They were literally built with the idea in mind that you CAN'T aim them at once location.

#399 Skoaljaw

    Member

  • PipPipPipPipPip
  • Survivor
  • Survivor
  • 126 posts
  • LocationAnywhere, USA

Posted 12 July 2013 - 11:49 AM

bump!

#400 DarkJaguar

    Member

  • PipPipPipPipPipPip
  • Big Brother
  • 331 posts

Posted 12 July 2013 - 11:50 AM

View PostOrzorn, on 12 July 2013 - 11:46 AM, said:

They were literally built with the idea in mind that you CAN'T aim them at once location.


I love your guys' recruiting video Orzorn, who did the editing on it?





20 user(s) are reading this topic

0 members, 20 guests, 0 anonymous users